The Not-So-Rapid Blog

Today I almost got fired.

Today IBM support saved my job.

We've had a serious problem at my workplace, and it's quite a tale. I'm not sure if I can ever tell it in public, to be honest. But today I put my neck on the block, and was very relieved to find that IBM did indeed stop the axe from falling.

Tomorrow could be a long day. And there are some murky political aspects to this. But suffice it to say that right now I'm very happy with IBM support. Very happy indeed.

 

Comments (0)
philipstorry February 15th, 2007 23:31:45

Vince Schuurman tells us of Philips' move from Notes and Domino.

I find it difficult to believe that this will save them money, and with a migration of this size tere's probably a considerable risk of the loss of data fidelity - to say nothing of functionality.

Here's the quote I found intriguing.

“We want to make our IT systems as easy for employees to operate as their home PCs.”

That's from Daniel Hatert, their Chief Information Officer.

 

My phone at work isn't as easy to use as my phone at home. When I want to dial someone outside this buiding, I have to put a 9 in front of the number. Unless it's someone in a building we also own, in which case I have to use a "normal" four digit number as always. I'm lucky that my direct dial number actually bears some resemblance to my internal extension number - many of the numbers assigned since don't, for reasons I don't quite understand. The voicemail system is cumbersome - I have to dial it, then give it my extension number, then give it a pin code. Then there's a big long menu system. Oh, and woe betide me if I need to call anyone in another country. That's not going to happenw ithout having to jump through hoops.

Is my employer's phone system suitable for business purposes?

Absolutely. My phone at home is simple because it assumes one user, in one fixed location. My employer's phone system has no such luxury.

 

Very few things in a well run organisation will ever be as simple as they are for a home. The needs are different. Anyone who doubts that should try looking around their desk at work and counting the differences each device brings.

 

It is, I'll admit, a laudible goal to make things easy to use. But comparing the home to the office seems crass. I would be very worried if someone told me that their large organisation could organise their documents as easily as teh average home organises its recipes or its postal mail. The scales, the purposes, the legal aspects and the governance required are just poles apart.

 

I note that Vince updated this entry to say that they're looking at outsourcing. It seems to me that any part of Philipsthere's which makes money - and wbuildingvoice-mailhappenwithoutlaudabletheishes to continue to do so - should probably make their own IT arrangements if they can. Whilst I'm not privy to the plans, everyting I've seen before tells me that these signs point to a plan which will see Philipseverything paying more for less service.

Ah well. At least Sony won't be alone in their shambolic loss-making position in the industry...

Comments (4)
philipstorry February 8th, 2007 14:16:00

So, LotusSphere is bringing us many goodies. Activities, Connections, Quickr, and more.

At the moment, everyone is all loved-up on the LotusSphere high. We've got shiny tech demos, we've got good feelings, we've got everyone (except me) in one place having a great time.

This is cool.

However...

(You knew that was coming, didn't you?)

 

Late last year something happened which rather annoyed me. And I get the feeling it's going to happen again after LotusSphere. At the time, I mentioned to someone that I was going to blog about it - but eventually, I decided to wait until now. It seems far more appropriate now.

In the past, the Lotus community online has thrown little spats about new technologies coming along but not using Domino as the platform. (SIP Gateways, Activities, etc.)

That's got to stop.

I understand why people do it. I'm very attached to Domino myself. It's a great product. It rocks. There's nothing quite like it. But it's not good at some things. It's never going to compete with DB2 for some kinds of data storage/retrieval, for example. And it's not a great servlet server. (Actually, it's not a servlet server at all.)

 

We can have the shiny new functionality now, built on existing stable technologies from elsewhere in IBM's portfolio. Licensed for use in that specific purpose, supported by IBM in case we have any questions or issues, and allowing us new opportunities and capabilities.

Or we can twiddle our thumbs whilst IBM busts a gut trying to shoehorn functionality into Domino, and get the same things much later and probably way behind the market trends.

 

I know which one I'd prefer.

 

When we finally get these great bits of software into our grubby little mitts, can we please concentrate on:

• How it's integrated
• How it's licensed
• How it's supported
• How we can hack it seven ways to Sunday, increasing our knowledge and doing really cool things in the process

Some of these things won't be Domino as we know it. Some of them may require an additional non-Domino server. And on one level, that sucks, But on another, we're so far away from the Microsoft-style licensing/product SNAFU stack that it's laughable. IBM knows that a stack like that isn't good for them. Heck, it's half their counter-FUD argument!

IBM are giving us great things. They might not be perfect first time out of the gate, but if we can keep our conversations out of platform prejudice and in terms of integration, licensing and education/support resources, we're far more likely to keep this great momentum we're acquiring.

 

Sorry to be such a downer whilst everyone else is on a high at the 'Sphere.

 

That's it. I'm done. Back to the partying, the learning, the mooching and the smooching! Go 'Sphere!

 

Comments (2)
philipstorry January 23rd, 2007 22:17:00

I'm not really one for prognostications. I am rarely found in amongst the crystal balls, the runes and the pigeon entrails.

But I am willing to make one very important prediction.

Notes 8.0 will be the most important software release this year. It may even be the most important software release of this decade.

 

"What the heck is Phil smoking?", you're thinking. But I'm not smoking anything, and I am deadly serious about this.

 

And the reason is simple - Microsoft Office.

Microsoft Office is at a dangerous crossroads. It's just changed its interface for its core programs. That means more training, more helpdesk calls, and ultimately more costs. I doubt it will be picked up as quickly in corporates as Microsoft would like.

And at the same time, Microsoft is changing the default file formats. Oh dear. Remember Office 97? A decade later, we're in for the same "could you please resend that?" chaos as we had last time around.

 

In short, Microsoft has made a huge gamble with Office.

 

And that's where Notes steps in. If you're going to change the UI and the file format, why stick with Office at all? Notes 8.0 includes the IBM Productivity Tools - a word processor, spreadsheet and presentations package that read/write ODF and run within the Notes 8.0 client. Not to mention that the old idea that the Notes client looks ugly has been thrown out of the window - look at the screenshots of the productivity tools. They're running in Notes. Gorgeous, isn't it?

Yeah. It even surprised me.

If Notes 8.0's productivity tools handle legacy Microsoft Office formats well , then this is a serious threat to Microsoft. Frankly, it's turned up at EXACTLY the wrong time for them, and demonstrates to the whole IT industry - no, wait, the whole world! - that you now have credible alternatives to Office.

 

On that level, it doesn't matter how successful Notes 8.0 is. Simply by setting the example, it will become the most important piece of software of the year, and quite possibly the decade. The changes that Notes 8.0 signifies with this simple move are immense.

 

I've not yet seen Notes 8.0. A quick note to IBM - this needs to have excellent filters to/from the legacy Microsoft Office formats. Just ODF alone isn't enough. And thank you. Thank you, oh thank you, thank you.  Notes 8.0 looks the part, and has the tools people need. In ten year's time, 2007 could be looked back on as the year that the Microsoft Office monopoly was shattered.

Notes 8.0 won't do that by itself. The excellent work being done on ODF will help, the large changes to Microsoft Office will help, and the continued work of Sun, OpenOffice.org and others will help.

But Notes 8.0 is the right product, in the right place, at the right time.

If I were Microsoft, I'd be needing new underwear right now...

 

 

Comments (0)
philipstorry January 23rd, 2007 13:58:00

I bumped into an old friend the other day. I hadn't seen him for a couple of years - last I'd heard, he was an IT manager.

Now he's a SharePoint Consultant.

As we chatted, the conversation turned geeky. He said Notes was dead and that Exchange had millions more seats, I asked him to point to his data. We changed subject. I said that SharePoint scales badly, he said to point to my data. (Fair point, I haven't tried running it on a zSeries yet...)

But I had to ask him one business question - licensing costs. If his solution requires SQL Server for data storage, Exchange for email, Office for data input - how do the customers react to that?

"Oh, that's great for us. The licensing costs make our development costs look better. If it's £20,000 of licenses, another £40,000 for development looks reasonable."

The blasé way in which he said it shocked me, I must admit.

 

I can't begrudge him a living - he's an old friend, and has a baby due in March. (It was a most informative conversation, as last time I saw him he was newly single and having to sell a joint-owned flat...)

But it did make me wonder. Although he talked about the cool stuff he can do with the next version of SharePoint, we didn't get round to discussing how clients took to having to upgrade (Office|Exchange|SQL Server) to use them - both of us realised, suddenly, that we were standing in the middle of the pavement on a dark, windy and wet night and that we had people to go home and see.

Exchange didn't kill Notes. SharePoint never seemed likely to, as far as I was concerned - it had too high a set of costs surrounding it. Perhaps I was wrong on that front?

 

Comments (8)
philipstorry January 16th, 2007 13:53:00

I am, occasionally, a type nut.

That is to say, I like typography. Well, the typeface end of it, anyway. A good typeface is a wonderful thing to behold.

I've got a large collection of fonts at home - a hundred or so that I like very much, and which have followed me around from PC to PC at home. (For those that aren't aware, a font is a subset of a typeface - a specific rendering of a style or size within the typeface family.)

But recently, I've been using Bitstream's Vera quite a bit. Especially on my work's PC.

Vera has been donated to the world by Bitstream, free to copy and redistribute. OpenOffice.org's Suite comes with it, and it's used in recent versions of Gnome and KDE and therefore is available in most places.

It's a nice crisp set of fonts. Bitstream did a great job in providing not only Serif and Sans Serif versions, but also a Sans Mono, which is an ideal programming / terminal font, owing to its clear distinctions between O and 0, and 1 and l.

(That was a capital o, a zero, a one, and then a lowercase l. If you weren't sure, then you shouldn't program in whatever font you're using right now!)

 

My entire Windows theme now uses Vera, as does Notes and a few other apps I have. Switching fonts for your UI is not a light step to take, but I've found it to be quite pleasant with Vera. It's certainly worth a try for those that are curious.

 

Vera is not the best font I've ever seen. (Fonts are things you use for specific purposes in an ideal world, so I don't really have a "best font".) But it is very nice for general purpose usage, especially in user interfaces on computer screens. It also looks pretty good when printed, too.

 

Anyone who can't see where this is leading - get to the back of the class!

 

I've checked the screenshots of Hannover that Mary Beth has provided on her blog. And I saw Arial. Or at least, that's what it looks like.

BLAST! Arial's fine on computer screens, but not so hot when printed. Whilst I don't hate Arial, I do find it jarring at larger point sizes, and dislike a couple of points in its design. And Arial is only available on Windows, which is my biggest beef - I'd like to be using the same font across all platforms, if possible. Hence my current preference for Vera...

 

And anyway, we've lived with a slightly sucky font performance from Notes for years now! Let's get this fixed!

 

This is, of course, a legacy issue. A problem that came in from a time when there was a client for Windows, OS/2, AIX/Solaris/HP-UX and Mac. Each of those four main platforms (I'm lumping the Unixes as one for simplicity) had its own font systems, none of which were quite the same. Even where they were similar, the differences were big enough to see with the naked eye, so Notes had to do something to retain a consistent look across platforms.

With four main platforms to serve, Iris had an innovative solution - they created "virtual fonts" for Notes. When you work in Notes, you've probably noticed that the default fonts for editing in Rich Text are simply called Default Sans Serif, Default Serif, and Default Monospace. (And Default User Interface and Default Multilingual, but they're not used so much.)

 

Notes basically picks a font on each platform to act as these fonts. On Windows, it's a 9 point Arial for the Default Sans Serif, for instance. The idea is that on each platform, you should get a similar looking font at a similar weight - thus freeing developers (and users) from having to think about multi-platform font issues, and allowing them to get on with putting text on the screen rather than wondering how it might look on another platform. Leave just a little padding to account for any tiny weight differences, and you'll be fine...

I'll give Notes its credit here. In a time where the idea of cross-p latform compatibility was a joke, this was actually a pretty good system. Remember, back then the closest we had to a multi-platform file format as good old plain text files. And even then, UNIX like to use a different line separator... In a time when the idea of data looking the same on any platform was a mere pipedream for many, Notes was doing its best to provide it. And doing far better than most others, too!

Then came the web. And the death of UNIX on the client side. And the death of the Mac as a platform. And OS /2 didn't so much as die, but barely had a chance to live anyway.

And we found ourselves with just two clients. Windows, and a slightly less loved Mac version which was always along late to the party.

But the font system remained unchanged. It still served. Silently. Unappreciated. Occasionally maligned by those that had cause to notice it.

 

Things have moved on, though. For instance, we're now up to Web 3.0, with Web 4.0 scheduled to ship early next February. (Venture capital willing, that is.)

And cross-platform typography is improving. Other projects outside of Notes have discovered a need to have fonts represented properly everywhere they go. Bitstream's donation of Vera is an excellent example of this - originally intended for the Gnome project, OpenOffice.org grabbed it too, and that's certainly why the PC I'm typing this on now has the font available.

So if OpenOffice can use the same font across multiple OSes, why can't Notes? Especially as it's running inside Eclipse...

 

Actually, I can think of a couple of reasons.

 

The first is that different platforms can have different numbers of dots-per-inch (DPI) in their interfaces, which makes things complicated. The number of dots-per-inch assumed by the display driver naturally has a huge effect on the size of everything being displayed, especially vector-based information like type.

Windows has a standard DPI of 96, moving to 120 if you use Large Fonts. The Mac, being slinky and elegant, has a DPI of 72. And X Windows - well, who knows? The most common setting is 96, which happens to match Windows. But it could be anything.

Differing DPIs are the kinds of complications that the Notes system was designed to work around originally, and some kind of compensation may be necessary - in Eclipse or in Notes itself - when switching from platform to platform. This doesn't change from the current situation, and a standard all-platform font like Vera would be an improvement but will not solve this problem. The big question is how that affects prioritising this kind of change for the Notes developers...

 

The second problem is the lack of a unicode version of Vera. That will be trickier to work around, really. However, the default Unicode Font will probably be different on most platforms regardless.

In an ideal world, IBM would simply make the investment and buy a unicode version of Vera from Bitstream to redistribute with Notes. Actually, that's not true - in an ideal world, IBM would ask Bitstream nicely if they would donate a unicode version, both Serif and Sans-Serif, in the same manner, ands Bitstream would be generous yet again.

That would rock, roll, and quite probably twist a bit too. However, I doubt it has even occurred to IBM, and Bitstream have already been more than generous with Vera anyway, so it's probably not going to happen.

So IBM might have to decide whether they want to either buy in a good unicode font for the purpose, or just use whatever they can find on each platform and lump it. Personally, I can live with unicode looking a little different from platform to platform for a while, providing that the same font is used for non-unicode (western) text across each platform. However, IBM are a global company and might not feel the same. Just as with the DPI problem, this will affect how high a priority IBM places this issue.

&nb sp;

Because of those two problems, I don't expect IBM to do anything about this anytime soon. Frankly, I'd be quite surprised to see anything turn up in R9, let alone get pushed into R8 given how late into the development cycle they are right now. Which is a shame - a good cross-platform client should have good cross-platform fonts to work with.

 

Still, we can hope. And I doubt anyone has really thought about this since the R4.6 days, when they began consolidation towards Windows-only Notes clients.

 

So by getting this off my chest and into my blog, I'm hoping someone bright in IBM will start thinking about this. We're not quite at the multi-platform client stage anyway, so there's no great rush. It's a nice-to-have, a bit of polish on the chrome that is Eclipse.

 

The font doesn't have to be Vera, by the way. If IBM have some other nice font family in mind, that would be cool too. But Vera's very nice. I think Notes should shack up with her. I also think that you should all go and fetch a copy of Vera, as you might like it. Especially the coders, who will like the monospaced version!

Here's hoping...

 

Comments (2)
philipstorry December 15th, 2006 15:26:00

One of the user-based bugbears in Notes has been, for a long time, the Out Of Office agent.

Users just don't understand why it works the way it does, and even if you explain why it runs that way, they're not going to be impressed.

(Why does it wait hours to respond? Scalability. Imagine running an agent every time an email is delivered. That's easy. Now imagine doing it for 10,000 mailboxes on one server. Domino can host that many mailboxes, easily, on the right hardware. It can't run that many agents concurrently. The current out of office system is designed for customisation, flexibility and scalability. Not for speed of execution.)

Heck, the other day, I was fairly surprised when, after explaining briefly that it runs that way because the product is designed for such scalability, a new user just srhugged and said "makes sense - thanks". Fair bowled me over, that did!

 

But Domino 8 changes this. In Domino 8, the Out Of Office replies can be generated by the router.

Hurrah! Because let's face it, the old system was a legacy one designed in a time when servers didn't have the computational power available to perform complicated Out Of Office calculations every time an email was delivered. Over a year ago, I was talking about how this could be improved, and I'm really glad to see that IBM has people even smarter than me.

 

I'm not going to claim any credit for this advance, you see. Firstly, I don't know exactly how they've implemented this new functionality.

Personally, having had more time to think about it, I've changed my mind on how I'd implement it - rather than build a custom rule, I'd just hard-code the functionality into the router and read the profile document fields for the Out Of Office information. That would be faster, smarter and allow you to expand the future functionality without mucking about banging your head against the limits of your Mail Rules implementation.

But if they've done it via mail rules, that's cool.

 

What's even cooler is that the Out Of Office replies are even smarter than before. In fact, according to the presentation that Craig links to, the Router also stops sending the messages automatically at the end of the Out Of Office period, supports delegation, and can send one message per person or a response to every mail.

Now, that could be one very fancy mail rule they've built there - but it seems to me that it might take a while to evaluate such an @Formula, so perhaps this is hardcoded into the router?

 

The second reason I'm not going to claim I had any great influence over all of this is that, frankly, it was all a bit obvious. Other systems have been working this way for a while now, proving the capability is there in hardware these days. And even to say "great minds think alike" seems a bit rude - if I had a mind that great, I'd be writing code for Domino R8, not blog entries on it!

 

One of the strengths of Domino and Notes has always been that the same code (database system, replication engine, text handling) is used on both server and client. Although R8 is primarily client-driven, the Out Of Office agent in an excellent example of how client improvements are driving server improvements, and how IBM is listening to users. They're doing this the right way, rather than just prettying up the old system or throwing in a hack of some kind.

About the only problem with this is that you have to upgrade your server to get it - but at least you can do that in-place on the same hardware, so I don't see that as a problem really. More business as usual.

 

Despite it being more of a client-facing release, which will see little improvement in the Designer and Administration clients, I'm pretty excited about the Domino R8 server. The fact that it's overshadowed by the R8 client is more a reflection of how much good work has gone into moving to Eclipse, rather than how little work has gone into Domino.

 

I can't wait for R8!

 

Comments (2)
philipstorry December 13th, 2006 12:33:00

I've not managed to write up my next entry on how search is - and should be - affecting everyone. This is mostly because I had to deal with the Little Server That Could.

Everyone that's worked with Domino at more than three sites has a story about the Little Server That Could, or so it often seems.

 

The Little Server That Could is the server which, in circumstances that defy both the odds and the credulity of those hearing the tale, Just Got On With It.

Pentium Pros that were still functioning as main mail gateways when every other server was a Pentium III - because they'd never given enough trouble for anyone to think about replacing them.

Machines that only replaced because a hard disk in their array failed, and it turns out that replacement disks aren't available. And haven't been for three years.

Machines that carried on despite gross misconfigurations below them, just quietly limping along.

 

Today's story is in that last category.

One of our servers decided, last week, to stop backing itself up properly. Which was odd, but I was in Manchester and there wasn't much I could do. It was in a cluster anyway - I made a note to check that all DBs had replicas, just in case, and carried on. After all, it was still working as a server - and the other server was being backed up too. So the risk was acceptable.

 

When I got back down here in London, I wandered by to have a look at my sick little server. First up was the mandatory "eyeball mk 1 check" - are all the cables cabled, the blinking lights blinking, and the whirring things whirring as they should?

Nope. Two of the network cards had lights, indicating that they were plugged in - but weren't blinking. Which means that there's no network activity for them.

So I logged on to get more details.

Hey presto, the problem is obvious. The server has teamed network cards - two teams, in fact. One for cluster replication traffic, one for user traffic.

Both teams had problems.

 

The user traffic team had just lost a network card. It was still there in the machine, listed as an active card - it just wasn't in the team. Odd, but the team still had three cards in it so all was reasonably well.

The replication team had a slightly odder problem, though. Windows had forgotten that the second network card in its team was, well, a network card. It was resolutely convinced that the device was just unknown hardware...

I pinged a few locations - on both networks. And noticed some packet loss.

(At least that explains why my Server Controller connection to this machine had been disconnecting randomly.)

 

So the Domino server was stopped, and everyone using it (both of them - it was out of hours!) failed over to one of the other servers in the cluster.

 

Fixing the user team was simple, but Windows refused to accept that the replication team had a second network card until I dissolved the team and rebooted. The moment the team driver no longer had its sticky little mitts on the card, all became good again.

(Except that I then had install the drivers for the card and rebuild the team from scratch, that is.)

 

As I tried to fix this, I noticed that trying to get documentation from across the network was a bit of a pain. All kinds of interesting errors when using network fileshares, and http requests were sometimes failing if I tried to use a browser. Annoyances that could be worked around, but still ones that slowed me down as I worked.

But what occurs to me is that whilst these protocols were having trouble, this server has - for the past week - carried on serving data to Notes clients without trouble.

In fact, I checked. One complaint that it was slow in our helpdesk. Just one.

Notes and Domino must have been re-sending data quite a bit to accomplish that. Yet not once did they complai n.

 

As it happened, one of the users of this server was still in the building, so I spoke with them once it had been fixed. (I was waiting for a test transaction log backup to complete anyway, so I had the time.)

They said that it had been a tad slow for the past week, but they figured that it was no slower than it was on the network in the morning or evening when everyone was logging off - and that because we'd advertised network maintenance for this weekend, it was probably just a side-effect of that.

They'd certainly been able to work well enough. They had been annoyed a couple of times when it had hung briefly whilst loading something, but that was it.

 

So, there we have it ladies and gentlemen. My mail-02 server from our mail cluster: The little server that could. (And did.)

 

There's the benchmark set. I'm sure someone must have a story of the resilience of Notes/Domino that tops that, so if you do - share 'em, please!

 

Comments (0)
philipstorry November 28th, 2006 22:15:00

Ed Brill pointed at an interesting blog entry from Colin Williams, on a put-down delivered to a hostile Notes user.

In the discussion at Ed's site, talk veered round (briefly) to some people preferring "Mail" to "Memo" when referring to their messages.

As the title of this entry implies, I'm of the opinion that if you can't figure out that an email is a mail is a memo is a message is a note, and probably a few other names that have been used besides, then you're probably in dire need of training before you find all your support calls filed in a folder named "PEBKAC".

I'm happy with the name Memo as it stands. But I thought I'd share a brief story that shows just how important this naming convention is for people in the real business world.

 

When I started with my current organisation, I wanted to get friendly with three groups of people: the smokers, the PAs and the canteen staff.

The smokers know all - it's the gossip pit of any organisation - and you should get down there and be friendly as quickly as you can. It helps to have an ear to the ground that covers most of the organisation...

Getting to know the PAs pays off because when they - or their bosses - call you, you recognise them and they treat you like a human being rather than some inanimate object that must magically solve their problem for them right now.

And if you can't figure out why you need to know your canteen staff/nearest sandwich shop/pub staff well, then you're no friend of mine!

 

One of the PAs had mentioned that she was over quota. In fact, all the PAs seemed to be over quota. So I checked their mailboxes, and was aghast to see huge 6Mb emails everywhere.

I couldn't figure out why, though. I mean, what are all these emails doing, across a set of very diverse subjects, at 6Mb a piece?

 

It turned out that there was a Word document in every single one. And I know that most of you have figured it out just from that one piece of information, but trust me - it gets better.

The Word documents were a standard form. I got hold of a copy of the Word template that they were using, and then went to ask some questions. Here's what I ascertained:

• The form was for Memos between managers
• It has "fields" for: From, To, CC, Subject, Date and Number of Pages.
• It has an organisational logo in the top right hand corner. (Can you guess what made it 6Mb large yet?)

My first priority was, of course, to get them a new copy of the form that didn't have a full-size printing-press-quality copy of the organisation's logo embedded into it. I managed to get the Word template down to about 170Kb and still get good print quality from it - any smaller, and the logo went a bit jagged.

Now that I had an immediate solution, I went back to them and handed it over. I also, of course, asked them why they had such a form.

 

The answer? "We need to send Memos."

I pointed at their open copies of Notes, and guided them to the appropriate button.

There was a moment of awkward silence from the PAs present. Followed by an admission that the Word template had sounded like a good idea when first proposed at a meeting of PAs, but that nobody had quite connected the dots.

 

I asked if the number of pages was needed, and whether the logo was necessary. The number of pages was something they could do without, but there was a little resistance on the logo - until it was pointed out that none of the Memos would ever be published in that format anyway, so it wasn't necessary in any way.

 

Leaving aside the inherent silliness of the situation, I'm left wondering - does it matter that it's called a Memo? As I've always explained it, it's just called a Memo in Notes because Notes has a bit of history, and when if first came out most people weren't used to "electronic mail". But they were used to office memos. Notes was simply mapping a normal working practice onto its own technologies.

If a group of 40 intelligent and well-organised people hadn't even twigged that their "messages" were called "memos", what does that say?

 

Every now and again someone tells me that the Memo name is an anachronism, and we should call it something else. The only really convincing argument I've ever heard for that is consistency - Notes calls it a Memo in the template, yet says that you have new mail when it alerts you and sends outgoing mail on the replicator page. Change that text outside the mail template to read message/messages, and you've solved the problem IMO.

(Especially when a message might be a calendar invite, a to-do or a phone message. Did I have new mail? No, I had a meeting invite. Consistency demands that we call it Mail or Memo in the mail template, but then refer to all new items as Messages outside the template, or we're working against our own argument IMO.)

 

Oh, and just to re-assure everyone, of course I offered to put our logo in the mail template. It was decided that it wouldn't be a good use of my time, in the end. (But hey, I offered. And sometimes, that's what's important.)

 

Comments (0)
philipstorry November 17th, 2006 13:49:00

Ed Brill has some new Hannover Notes 8.0 screenshots!

OK, so that's not really a very good joke. They're Notes 1 screenshots, featuring the eye-gouging Windows 1.0 colour scheme, which wasn't "bettered" until Windows XP shipped. (And Windows 1.0 at least had the excuse of a limited palette to choose colours from!)

So, joking aside, I  do have a comment on these screenshots that I think is pertinent.

Look at the last screenshot - what appears to be a New Replica dialogue box. Isn't it all quite clear?

In fact, I only have one major criticism of the whole dialogue box - the fact that Access Control List option looks like a confusing extraneous option. (Can you even uncheck that? Or did radio buttons function as checkboxes in Windows 1.0 too? I'm not sure, because the first version of Windows I really used was Windows 2.x.) But apart from that one setting, which may be an OS problem anyway, this dialogue box is - for the time - near-perfect.

I add "for the time" because conventions move on, time flies, and stuff happens. But I think this dialogue box stands the test of time in one fairly important regard:

It uses plain language.

Just look at the options for what to replicate:

• New notes only
• Old and new notes

The kind of plain language that Notes needs more of.

 

I know that there's a lot of strings in Notes, and that it'll take a long time - but we need to get more dialogue boxes that ask simple questions like that. One of the genuinely deserved criticisms that Notes gets is that its error messages are cryptic, and that some of its dialogue boxes assume knowledge of Notes that may not be there. Both of these can be handled by using plain language.

I hope Hannover steps towards dealing with this. It won't be something that can be quickly fixed, but it would be nice to see a shift towards this kind of improvement.

Of course, that dialogue box isn't perfect. It shows the source server and database filename, but why? That seems extraneous. It might be useful once in a blue moon for the power user, but they can just try opening it from the right server before they create their replica...

There is always room for improvement. At least Notes dropped that part of the dialogue box!

 

Oh, and did anyone else notice that the message highlighted in the Computer Industry Gossip database was "IBM scares clone makers, sellers"? Now there's a coincidence - back then it was cloners of their hardware, and now it's with cloners of their messaging systems.

 

Comments (0)
philipstorry November 14th, 2006 08:52:00

I post about realising what a nasty old habit I had in my muscle memory, and then the Lotus Notes blogosphere lights up with the proposed and of double-right-click to close windows.

Looking at this, I'm inclined to agree with Mary Beth.

If Hannover can't have double-right click to close a window, I'm prepared to accept that. Actually, I've been prepared to accept that for quite some time. I barely double-right click now, anyway.

In fact, I just checked my options and it's not even turned on - so when I migrated to a new machine a few months ago, I must have not bothered. I stand corrected - I never double-right-click now.

 

The choice here seems simple to me - consistent backwards-compatible behaviour across the two clients (old Win32 binary and new Eclipse-based Hannover) at the expense of other features, or more development of the client at the expense of some inconvenience to the few users that are using an esoteric backwards compatibility option.

Neither is an attractive option. But the clincher, for me, is that this isn't even a default anymore.

Which means that the only people using it are power users or people that have it forced on them by policies.

So, not many people then.

 

There's our choice. Let's be generous and say that there are five million users. Who do we want to serve - the 120 million that didn't even know it was there, or the 5 million using it? (And I still think I'm being more than generous with that figure.)

 

Folks, the point of Hannover is to create the baseline for a new generation of Notes client that's easier to use.

What I find ironic here is that the users that are best placed to change are the ones that are screaming loudest.

 

Seriously, folks. You're literate and skilled. Many of you have adapted through four or more versions of Notes. Can't you just take a hit for the team and let IBM drop this one, in the name of consistency and simplicity?

Sometimes you're just hurting the one you love, even though you think you're protecting them.

 

Comments (5)
philipstorry November 7th, 2006 14:50:00

When we upgraded from R5 to R6.5 here, we enabled soft deletions on our mail databases. It solved a problem that many users had when it came to managing mail - they'd select mails for deletion, but not actually delete them.

Soft deletions were an excellent solution. The biggest problem we had with them, once we'd proved that they worked, was picking a timeout value - the default of 48 hours was felt to be too low. We compromised on 96 hours, or four days. That was long enough to survive most weekends, but not so long that it would cause problems for our capacity planning or quota management.

But old habits die hard.

Why do I keep hitting F9 after deleting mails?

I did it just this morning, when removing a bunch of systems reports I'm cc'd on but have no interest in.

I think it's because it was always more efficient to select loads of emails and then hit F9 than to do it individually with the old Trash folder mechanism. I'm fine for single deletes - I just tap delete and it's gone. But the moment I want to delete multiple emails, I fall back into old habits.

What's really weird is that it's taken me well over a year and a half to notice this.

Still, hopefully now I can stop doing it.

 

Comments (0)
philipstorry November 6th, 2006 11:41:00

I had absolutely no idea about this particular tip until earlier this week, when it showed up as an updated Technote in my IBM support RSS feeds.

If you stamp an email with a "Please Reply By" date (Delivery Options, Advanced tab) it shows up in the recipient's to-do list. Which is cute, but you try getting your some of your recipients to use their to-do lists!

 

But according to Technote 1097972, if you stamp a Please Reply By date onto an email AND ask for a Read Receipt, then the sent email will go into YOUR to-do list. And the read receipts you receive will show up as responses!

 

Now that's neat.

 

It's not perfect, of course - you have to preview or open each Read Receipt, as it doesn't show the recipient's name associated with the receipt in the view. I'm sure that many with customised mail templates will probably read about this and want to improve upon it, though. (Perhaps add a specific read receipts view, and a button on the Memo form to provide a list of users who have read the email - in much the same way that Meeting invitations can show you who's responded and how?)

But it certainly saves you from playing hunt-the-read-receipt in your inbox.

 

What amazes me is that this functionality has been in there since R5, according to that Technote! That's seven years old functionality, which I didn't have the faintest idea existed...

 

(This short Show 'n' Tell Thursday item was shoe-horned in at the last minute because my longer item is STILL being written.)

 

Comments (4)
philipstorry November 2nd, 2006 23:19:00

Over at Genii Software, Ben Langhinrichs notes that some of their internal apps that were developed eight years ago are still running without modification.

I was stunned.

I mean, for starters, everyone knows that Notes R4.6 - which Ben was using then - was an awful product with an awful interface, in which it was impossible to be productive. So how Ben actually wrote this stuff is beyond me anyway. It probably took him months. Maybe he started writing it in R3?

And it's still running.

Which is ridiculous. I mean, just think of the lost opportunities!

By running eight-year-old code, Ben's lost so many opportunities to re-write and use new APIs and libraries (which replaced old working APIs and libraries for no good reason). Sure, he's had no reason to look at it because of Domino/Notes' excellent backwards compatibility. But he's also therefore had no reason to improve it, to add those little features that he didn't have a use for anyway, but that made it so much more fun to maintain and document.

Ah, yes. Documentation. In eight years, nobody's looked at this code. What horrors lie within it? Without comments and documentation, this code is probably lost to living memory.

If Ben did need to change the code, his complete lack of familiarity would be a major problem. Reliant upon comments that made sense at the time, and vague documentation that relies upon assumptions common before the liberating experience of developing with Buzzword 2.0, Ben will be lost if he ever does need to change this code.

He'll have his own little Y2K moment, except he'll also have to face the horror that HE is the COBOL programmer he needs to hire!

 

And I know what Ben's arguments will be. It's reliable. It's proven. It does the job is was written for, and doesn't need to do anything else.

Doesn't need to do anything else?

Good god, man! Of course it needs to do something else! This is 2006! If it doesn't generate an RSS feed, it's defunct! It's just that simple!

 

Ben should probably re-write this in Ruby. Or at the very least, Python. Then he'll gain some valuable experience in these hot languages, and be rich! Because we all know that in today's market, the business plan for the cool cats is:

1. Use the latest technologies before anyone else as self-important as you is doing so
2. ...
3. Profit!

(Business plan Copyright (C) Slashdot Comments System, 2001. Used with permission.)

 

Of course, what's saddest of all is that for the last eight years, through his wanton neglect of this application, Ben has actually had time to further develop successful commercial products like CoexEdit, CoexLinks, the Midas Rich Text LSX, the Midas Rich Text C++ API, and @Midas Formulas.

Where was the love for this application during all that time, Ben?

 

In the old economy, this would have been a triumphant tale of excellent return on investment. Luckily, in these enlightened times, we can see that he's actually just been denied excellent learning opportunities and intimate code familiarity, which will surely come back to bite him in the long run. As I believe an old Radicati report said:

"They that give up essential learning opportunities to purchase a little return on investment deserve neither learning opportunities nor investment. (Especially not from venture capitalists.)"

 

Comments (6)
philipstorry October 25th, 2006 08:27:00

The Web 2.0 crowd are always waffling on about how brilliant web applications are. And in some cases, they're right.

However, in some cases, they're absolutely wrong. Case in point - the idea that the support is easier and cheaper because there's no fat client binary to deploy, just an application on the server side.

Yeah, there's no fat client to deploy. Just testing against large numbers of browsers.

Now, in a corporate environment, you can limit the number of browsers people use and the versions of browser that they use. But for something like my organisation's webmail system, it's trickier. Very much trickier.

People use it from anywhere in the world. On any machine. We get to say which browsers we support, but believe me - and my webserver's logs - that doesn't stop them from trying all kinds of browsers with it. I mean, we tell people that we'll support access from a Windows PC with Internet Explorer 5.5 or higher. What part of that does Opera 9.02 on Mac OS X fulfill, exactly?

But here's what made me think about it seriously - Internet Explorer 7 is out. Yes, that's hardly up-to-the-minute news - but this IS the Not-So-Rapid blog, y'know...

The impact of IE7 for me is simple - not much. We're not deploying it internally on any desktops/laptops, so we don't need to test anything against it. Oh, except anything that's externally accessible.

Like our webmail system. Whoops.

And is DWA compatible with IE7? Um... Maybe.

Now, to be fair, IE7 has been out of beta for less than a week. It's grossly unfair to ask IBM to be able to say for certain that it'll work.

So, to be safe, I've just had to send our support team leaders an email saying that we can't and won't support it until testing has been done.

 

It was in drawing up the timescales for that testing that the folly of Web 2.0's "cheap 'n' easy!" claims became evident.

 

IBM currently say that R6.5.6, or R7.0.3 will support IE7. R6.5.6 will not ship until March 2007, and R7.0.3 currently has no advertised shipping date (they're working on R7.0.2 FP1, which will ship in January - that should give you an idea of the timescales.)

So, if Microsoft push IE7 out as a recommended security update in November or December - as they're expected to - then my application (DWA) will likely be running on an unsupported browser for most of our staff using Webmail. After all, they're not going to turn down a Microsoft recommendation, are they?

Therefore, we have to test it ourselves and get any necessary FAQs/Known Issues out to our support staff, preferably before everyone gets IE pushed to them by Microsoft. "We don't support IE7" is not likely to be an viable option within a couple of months, it seems.

In fact, we already have four people using it, as far as I can see from our webserver logs.

 

So am I just griping, or is this a serious issue?

 

I have no idea. And I won't until I complete testing of IE7 for myself, and have IBM's own testing results via their knowledgebase. At that point, I'll be able to correctly assess the problem. But not until that point.

And that is the point for web applications - any update to the browsers yur application is designed for must be tested, yet ironically you don't actually get to test it until AFTER the update arrives.

 

And this is an improvement over fat clients?

I seem to recall that my fat clients were usually tested on a test network, piloted on a group of test PCs, and then finally deployed widescale once we were s atisfied that there were no problems and the benefits of the new version were cost-effective. Yes, that was expensive. But if something broke, it most likely broke during testing or piloting, and didn't affect the organisation.

I also seem to recall that I had control over it. My testing and piloting was done alongside an already deployed fat client, which was already working.

 

This new Web 2.0 method can't do that. New browsers appear and disappear with each update from the browser manufacturers. They break your apps, unbreak your apps, and it's just chaos.

 

But of course, we can control that in an organisation by using the same QA procedures for the installed browser that we use for any other fat client binary software package. So in most organisations, this won't be a problem.

So let's just treat my webmail application as an abberation, right?

 

STOP RIGHT THERE!

(Dunh-dunh-dunh, I wanna know right know... *ahem*. Sorry. Meatloaf earworm...)

 

We can't just stop right there, unfortunately. Because one of the advantages to those nifty Web 2.0 applications is supposed to be their rapid deployment and open nature. The idea that they're sold on is that if you start working with another company, you can just throw an extranet up for that company and it'll all be fine.

Except we've just been talking about how that won't be fine, because now you're dealing with browsers that are outside of your control.

And what if it's not a problem in your code? What if it's a problem in your platform? What if your current version of whatever framework you have is causing your problems?

Well, just like me, you're now dependent on your own testing, advisories and workarounds until you manage to get an update for that framework/platform from your vendor.

 

Wow. It turns out that the lean, mean, agile Web 2.0 still has the sorts of dependencies that we're used to. With the same sorts of impacts. They just don't like to mention it, because it gets in the way of the AJAX-filled client-end demos.

So your choices, when this happens in Web 2.0, are simple - spend time and money working around it by hacking the framework/platform (if possible) to be compatible, or leave it and do nothing until the promised update. Oh, how familiar.

 

It seems to me that the Web 2.0 people don't have the answers to the same questions that we were asking of the Web 1.0 people over five years ago.

They have shiny interfaces and rapid deployment, certainly. But I don't want to rapidly deploy a fix for something outside of my control that broke. I want it to not break in the first place, by being under my control. That's the point of an available, reliable service. Which, by design, Web 2.0 can't definitely deliver due to its platform dependencies.

 

Of course, none of this means that Web 2.0 isn't viable. It's got some excellent aspects. It's a nice set of goals and ideas that we can re-use even outisde of the traditional Web 2.0 technologies. We should just be aware that it's no panacea.

 

And we should have seen the warning signs anyway. Like all high-noise, low-signal fads, Web 2.0 is promulgated primarily by developers. And I mean no offense to developers, but I can't help notice that you guys almost never have to MAINTAIN what you leave behind. Someone else is always left with the shovel, trying to grow the promised roses.

 

So, the web browser thin-client turns out to need a whole set of shovels that traditional fat clients didn't. Personally, I think they're at least even - browser clients have their own unique set of problems, as do fat clients. The work's still there, just in a different place. There is no panacea of very-low-maintenance clients that never require updates, never break and never need support.

 

And if that surprises you, I own a bridge over the Thames I'd like to sell you.

Assuming, of course, that I have time to sell it to you, what with all the browser testing I now need to do with my webmail application...

 

Comments (6)
philipstorry October 24th, 2006 12:51:00

The holiday blues are supposed to start after you get back form holiday, right?

I have them before I've even been.

The public sector in the UK is pretty generous with annual leave. I get 30 days standard allowance, and carried some over form last year. I currently have... 26 days remaining. No, wait - make that 28, because I have two days TOIL from a hardware upgrade I did over a weekend earlier in the year...

The holiday year runs from April 1st to March 30th, so I've got about six months left. In which to take over a month's holiday.

In that time, all I have to do is:

• Replace a mail cluster (more complex than it sounds - see below).
• Set up a dedicated Administration server (using some of the replaced hardware from the mail cluster).
• Migrate all applications from our old Intranet server to a new server.
• Set up a new mail cluster at a growing remote site, and migrate 200 users onto it.
• Upgrade two Domino.Doc Document Management (DDM) servers to DDM 7.

The first mail cluster upgrade is the scariest one, really - we have three servers in a cluster, but two of them are on quite old hardware. For various historical reasons, the third cluster member is newer hardware but is also our Administration server for the domain.

(It looks like they merged organisations at some time by adding the third server, registering all new accounts on it for the "mergees", making it the Administration server to speed up bulk recertification processing, and then just forgot to clean up afterwards. The third server must have been temporary originally, because it had less than half of the disk space that the first two had - we had to eventually upgrade its disk array just to prevent it from falling over with a full belly. If they didn't have a cleanup plan for this third server, then they badly specced it in the first place. But as documentation from that merger is spotty and difficult to find, I can only piece this story together using my "fearsome forensics skills", so I can't say which it was for sure.)

I'll be adding a fourth server to that cluster, and migrating about three hundred users (and a hundred mail-in databases) from the older two servers to that new server, then killing the old servers. The hardware for the newer third server will be upgraded to match the fourth server at some point too.

The old hardware for that third cluster member then becomes the Administration server for the domain, which will also allow us to dedicate it to statistics monitoring and other worthy tasks that only Administrators would care about.

Migrating applications to the intranet server is already underway - a customised variant of the excellent Advanced Domino Web Server Log template is making it very easy to see what's used and where on the Intranet, and log.nsf is identifying used applications on the server. Once we have a suitable data set, we'll go checking for those nasty evil absolute URI's in the Intranet applications and get them re-written, then simlpy replicate the databases to the new server and flip the DNS addresses. Should be a snap.

The new mail server in our Manchester office should be a snap too. We'll have learnt all we need to know about how to do "accelerated user migration" from server to server, so we just set up the new cluster using the two older bits of hardware we removed from the cluster here, and then shunt the users over.

But with less than five months, owing to Christmas (nothing gets done over Christmas here!) and my annual leave surplus, it's a tight timescale.

Even tigher now we can't do anything until mid-October as we're entering a change freeze whilst the organisation completes some very important work.

So make that four months.

Four months. One cluster rejig. One cluster setup. Two large user migrations. One set of application migrations. One DDM upgrade on two servers.

Did I me ntion I'm the only Domino administrator here? I shoud probably have mentioned that, I suppose...

If I was using any other product but Domino, that might be scary. As it stands, I'm confident we can get all of this done by the end of February, providing no higher priorities crop up.

Without cutting corners, without compromising any of our service levels or risking our data. And with very little noticable impact for anyone outside of the users being moved. (Except for the DDM upgrade, which should have an obvious change to the interface for starters!)

Which will leave March for a Blackberry Enterprise Server upgrade to 4.1, I suppose. Unless I think of anything else we should do in the meantime, that is...

Anyway, that's the near future. Busy, busy, busy - but only possible because it's Domino!

 

Comments (2)
philipstorry September 14th, 2006 11:45:00

There's a really good discussion going on over at Ed's blog, all revolving around why people don't see the value of Lotus Notes.

To my mind, it's shoved an elephant into the room, and we can't ignore it. (Even if it does juggle, stand one one leg and keep rings spinning - just like an old IBM advert!)

The elephant is simply this: Where is the Community Edition of Domino?

We have an Express edition of Domino, which many think is too restrictive. It fails to give people some of the things that the community praises most about Domino, which is not a good strategy to make sales.

What we need is something like the DB2 Express-C edition. It's very impressive - it limits you on the hardware you can use it on, and takes some of the Really Big Enterprise features away from the mix. But otherwise, it's free to deploy/develop/distribute. You'd be nuts to run anything critical on it, unless you want to rely upon the support of a kindly community.

So what can Domino do to match that?

First, I think we have to admit that Domino is not in the same market, so the strategy must be a little different. Domino isn't just a data store that you lob your bits into - Domino is a platform that speaks many protocols and is capable of automated workflows and data processing that DB2 can't even dream of. You don't redistribute Domino with an app you write, so some of the nice features in the DB2 Express-C licence just aren't applicable.

I suspect that we could argue this one with IBM until the cows come home. But I do think that this is an opportunity IBM are missing, and that this harms Notes/Domino more than we can imagine. So here's what I suggest:

IBM Lotus Domino Express-C

Free as in beer, but with the following restrictions:

1. Only available on the Win32/Linux platforms.
2. No clustering
3. No partitioning
4. No Extended ACL
5. No NSFDB2
6. No SAP features
7. Not to be used on a box with more than 4Gb of RAM or two processors
8. No more than 100 licenced users active in the domain(s) used by the organisation.
9. No more than 10 servers per domain.
10. One free Domino Administrator license, two free Domino Developer licenses, 100 free Notes Collaboration licences.

This works for very small shops and for those developing applications. It's not that the old Express edition wasn't good, it's just that most of what it provided could be cobbled together (and I do mean cobbled!) with other technologies. By dropping the price to gratis, you can now have it all in one package.

100 users and ten servers? Figures off the top of my head. Drop it to two servers if you like, but more than one is a must. 100 users seems about right - it grabs a good chunk of small companies that were never going to pay for collaboration anyway, and were otherwise just going to go with Exchange Server in the Small Business Server deal.

The free client licences? Yeah, that's a tough call. The free licences are usable ONLY with an Express-C licence, of course. Free client licenses make it even harder to play the "we have more seats than you do" game, but you can turn that to your advantage if you're smart. But we have to face facts and admit that a server with no users is useless, and people wanting a free server probably won't pay for that many clients either. Drop it fo fifty or even twenty, but you could really make headlines with a figure like 100.  ;-)

And finally, one more restriction:

1. Support from a community only.

Want support? Go buy Express.

But IBM, I promise you this: On the opening day of the Community Support Forums, you'll have at least two century's experience on hand from volunteers. Just abou t everyone with a blog will be dropping by, plus many more without them. Especially if you just rebrand the Notes.Net forums - which were, let's face it, always about Community Support anyway.

The only thing stopping an Express-C version of Domino is IBM. If that doesn't stop, then Domino will never have a look-in at the lower end of the market.

IBM Lotus Domino Express

Ah, Express. How maligned have you been until now? But with a free Community edition (Express-C) beneath you, you may now SING! Reach for the skies, and deliver (for a modest fee) unto us:

1. Clustering.
2. Partitioning.
3. NSFDB2.
4. Extended ACLs.
5. Still no platforms other than Win32/Linux.
6. Still no SAP.
7. No more than 4 processors, but all the RAM you can eat.
8. No more than 1000 users in the domain.
9. No more than 100 servers in the domain.
10. 2 free Domino Administrator licences, 5 free Domino Developer licences. You pay for all other licences.

Again, I'm very much pulling the latter two limits out of the air. But you get the idea - all the goodness of Domino down at the low-end of the market. Want more processors or Big Iron? Step up to a full Domino Server Licence. Let's drop the idea of the Application Server/Enterprise Server licences. We just want three simple tiers, where the benefits of each licence are clear and simple, and progress seems natural and normal.

And we want to give companies an incentive to develop applications. Hence the silly number of free Domino Developer licences. It's a gamble, but it might pay off. Even just the one free licence would allow IBM to make a song and dance and drive home the point that you can make Domino/Notes work for you...

If you also drop the frankly useless Lotus Notes for Email licence, you can give Lotus Notes/Domino a real shot in the arm. It becomes everybody's playground, thanks to the Express-C version. Businesses have a real upgrade path with the Express version, once they find that they need support or outgrow the limits of Express-C. And the Domino Server licence caters for those who want heavy hardware or a very stable OS.

The best time to do this would obviously be when releasing Hannover.

But imagine this AND a nifty fifty. Real collaboration solutions that you can run on a community-supported edition of an enterprise-class server. A clear upgrade path for when you start to hit your stride as a business. And building the idea of an in-house developer into the licencing structure.

That's got to get the value message across.

Meanwhile, consider me at the front of the queue for the Express-C forums. It'll be just like the old days in LOTUSC on CompuServe...  ;-)

 

Comments (0)
philipstorry August 29th, 2006 14:59:00

We have this occasional problem with the Notes client here... If anyone could throw any light on it, I'd be grateful.

Basically, sometimes people will open their Mail database and get the following error message:

Class or type name not found: CSEVENTNOTES

They then can't delete mail properly, use their calendar, or much else.

If I point them at Domino Web Access, all is well. If I - or anyone else - opens their mail database, then there's no problem either.

For most people, removing cache.ndk and desktop6.ndk will solve this.

However, for a few people the problem is in their bookmarks.nsf database.

And there lies my problem - we use Notes roaming. Removing the bookmarks database safely requires me to downgrade the user from roaming, clear down their notes client (using my NotesClearDown tool), and run first setup again to get a new bookmarks.nsf. Then, because they've lost their old address book, I have to copy any addresses back from the old roaming databases on the server befiore I can approve the deletion of their roaming data via AdminP. Once AdminP has done its work, I'm finally able to re-upgrade the user to roaming.

This is, as you can guess, a little time consuming. Especially with remote users. And the nature of roaming means that until this is done, the proble occurs on all machines for that user. My only recourse is to point them at our internal Domino Web Access implementation.

Thankfully, this is rare. Very rare. With 1000 active users, we see this about once every four months or so. But it's still a pain.

IBM have a technotes and an SPR for this:  1145239 and 8525701900420C98 (although I note that we're using R6.5.3, not R6.5.5). But all they say is "remove these databases and all will be well". I know that already - removing just takes longer than I'd like!

So, has anyone else experienced this? If so, did you ever find out what was wrong in bookmarks.nsf? I've tried removing all links to the mail database (including those in the history) and then using Restore Defaults to re-write the bookmarks, and had no joy. (Even if I closed Notes before using Restore Defaults.)

I have a tool that we use to do some client setup/fixes automatically. If this is a corrupt document that I can identify programmatically, I could use that tool to fix this - saving a few hours work each time. It's a vain hope, probably more of a dream, but being able to incorporate a fix for this into my Automatic Client Fixes tool would be really useful. Hence me asking if anyone has seen this and cured it by other means.

Anyone? Any ideas?

 

Comments (4)
philipstorry August 16th, 2006 10:43:00

Ed Brill asked if anyone was using POP3/IMAP, and has 72 responses so far.

I’m not surprised that people are using POP3/IMAP, but I am surprised at what they’re using it for.

Here, we just use it for integration with 3rd party products. Nothing more. So, for instance, our CRM system pulls mail from mailboxes via POP3.

To us, it’s just like using LDAP & SMTP so that our photocopiers can also act as scanners that then email scanned documents to you.

But we wouldn’t run POP3 or IMAP on our production mail servers. The idea that mail should possibly leave the safe, backed up, searchable Domino server is an anathema to us - you use Notes or Domino Web Access, or you have no mail. It’s not expensive to support POP3 clients as such, it’s just that the risk of losing email is too great for us to consider using it.

I was intrigued to see that the majority of people running POP3 for real users were doing so for short-term staff or for "malingerers" (as I call them), who would not or could not use Notes.

Now, here we have Blackberry - which helps alleviate some of the situations that some respondents face. For desktop users, we’d allow Domino Web Access at a push. (Although for some users on esoteric platforms, I can see that benig problematic.) But what if we had no Blackberry server? A lot of Ed’s respondents were using POP3 for mobile access to their email.

I wonder if some of these users might not be just as happy with a small, lightweight "Domino Web Access for Mobile Devices" mail template? Something that just does the bare basics of email, calendaring and to-dos via http?

It’s just a thought. I’m not sure if I’d have much use for it, but it might help out some of those who are using POP3 in Ed’s responses. Granted, it lacks offline access - but it at least provides access without removing the mail from the server, and possibly without the performance hit of IMAP or POP3.

And to re-iterate what I said in that thread, I’d like to see POP3/IMAP access control tightened up in future releases of Domino. This "every account on the server has access" lark might be simple, and is probably a good default. But I’d need a way to restrict access to POP3/IMAP per user, if only to prevent a proliferation of mail clients pulling mail off servers onto fragile laptops that get lost or broken with no backups available...

Comments (6)
philipstorry August 11th, 2006 13:41:00

This one's been covered before from some angles, by Greyhawk and the incomparable Christopher Linfoot.

But my last entry covered the conflict between the need for RFC-required email addresses and business needs. So I'm going to cover this again briefly, as my very first Show ‘n' Tell Thursday item.

RFC2821 requires you to have a postmaster address, to which undeliverable mail and queries for your messaging administrator can go. (Domino doesn't route undeliverable mail - it simply deletes it by default after a while.)

RFC2142 suggests abuse@domain.tld as a further email address, to which reports of "inappropriate behaviour" can be sent. This is a good idea, for obvious reasons.

Domino does actually handle both of these by default - see technote 1106677. You don't need to have a postmaster or abuse account/mail-in database - if the mail address doesn't exist, Domino will send the mail to the administrators instead.

Which is great, but has the following minor drawbacks:

• Lots more mail to your admin group, from bounced emails and spam amongst other things
• It's not clear that the mail addresses already exist
• If you create a mail-in database or account with that name for some other purpose, important mail could get delayed or even ignored instead of being handled

IBM do have a technote covering whether or not you need to create a postmaster address - technote 1132678. It is, in my opinion, woefully inadequate. Not least because it recommends creating a new account or adding the address as an alias to an existing account...

Best practice in this situation means creating a mail-in database record to point to a shared mail-in database.

A new account is an identity which must be maintained - it's not worth the bother unless you really need a separate account for some reason. Best practice is to limit the number of live accounts that can be logged in - partly for licencing reasons, partly because certificates have to be renewed, passwords changed and disseminated, and so forth.

So just create a new mail-in database for your postmaster address, and another for your abuse address. The benefits are numerous:

• It becomes clear that these addresses are "reserved", and the business should not use them for other purposes
• Administrators will know that they exist and that they should monitor them
• Mail sent to the addresses will not be removed along with all your administrator's mail when your administrator leaves

Just because Domino mollycoddles you by handling postmaster@ and abuse@ automatically, it doesn't mean it's right. Domino does this so that it's compliant "Out Of The Box" - it doesn't do it so that you can rely upon it for years upon end!

So if you don't have them, please create them. In the long run, it's the best thing to do. And it'll be one less thing to worry about as your Domino infrastructure grows with your business...

Comments (2)
philipstorry August 10th, 2006 23:19:00

Sometimes, the needs of business and the needs of IT can clash.

Most recently, this has surfaced in what - to most messaging administrators - seems like a very silly no brainer.

Our Corporate Services department wants us to turn over the Postmaster mailbox to them. Or delete it. One of the two, anyway.

They say that it confuses people, because they’re used to sending mail to a Postroom mail-in database to get postage/photocopying/binding and other services.

Our stance here in IT is simple. Postmaster is a required email address, as laid down in RFC 2812. We will not delete the address, but we will turn it over to the postroom - provided they check it daily and route issues accordingly. They will need to employ a messaging administrator to do that, of course.

The big problem here is that it quickly becomes something of a battle of perceptions. In IT, we know that the joy of Postmaster is that it’s one of the few addresses you know is going to work no matter where. So we do get some technical queries to it.

In the postroom, they know that some people have accidentally sent mail to the wrong address. It happens about once every three months these days, but it still happens.

The big problem here is that for all our quoting of RFCs, and for all our common sense "it’s just sensible for IT to own and operate this address", the political pressure can be difficult for IT management to deal with - especially when the perception is that the mailbox causes problems which may mean physical post (which is almost always, coincidentally, very important) might be delayed because someone sent a request to the wrong mail address.

Yesterday, however, we got a nice surprise in the form of handy evidence for our position.

We’re currently pursuing a DNS Complaint Resolution, against a typo-squatter. My organisation inspects, audits and investigates healthcare in England and Wales, and the typo-squatter has a one-character difference address that they are using to advertise laser eye surgery, drugs of dubious provenance and off-the-shelf heart surgery. (Yes, my mind boggled at that too.)

Somehow, the contact details for the person we have dealing with that complaint got mangled. But the bright chaps at Nominet cc’d the mail to Postmaster@, so we saw it and forwarded it on to the right person.

Nominet could rely upon an RFC recommendation being carried out. If they couldn’t, we could have lost our fifteen day window in which to resolve our complaint, resulting in some unsavoury characters being free to advertise dubious services via our name (albeit mis-spelled).

If nothing else, the Postmaster address can now not be deleted - its value has been proved. I’d hope that this example also shows why IT is better placed to handle the mailbox - we understood immediately what a DNS Conflict Resolution was, and could route it accordingly. The postroom staff might not be able to distinguish between the spam and the real messages as effectively as IT staff can.

On a personal note, this kind of typo-squatting sickens and annoys me. People coming to our website could be vulnerable, having suffered from problems with the national healthcare system, and looking for help in getting those issues recognised and fixed. To view that as a business opportunity is at best disgusting, and shows a contempt for others which is quite unfathomable to me.

Comments (2)
philipstorry August 9th, 2006 10:06:00

There's a sudden urge in the community to get a new Nifty Fifty set of applications, to demonstrate what Notes/Domino can do.

This is excellent. We've always said that Notes was more than just mail, but had little to show for that "in the box" except for some pretty basic Document Library/Teamroom templates.

I, like everyone else, have plenty of ideas for what can go into the series. The Team Mailbox, for instance, is a must. And some kind of Calendar Aggregation/Team Calendar database would be handy, too.

We need real-world solutions that can show people "Hey, this product solves problems. It does more. And now you can run with it, take it, and do more yourself".

Lots of people asked how IBM can help. I'm late to the party as usual, but I do think IBM can and should help, for one simple reason.

Lotus Mail Licences - now known as Notes For Messaging licences.

I don't know if we still have these "in the wild". Does IBM even still sell the Mail licence? Their licencing details - especially their FAQ - seems to show that they do.

For me, IBM's involvement in the new Nifty Fifty effort boils down to this one licence. If they still sell them, then IBM needs to have a template developer briefly test and sign versions of the Nifty Fifty templates with the appropriate ID file so that anyone can use these templates. Otherwise, Notes isn't more than a mail client - not to the average user, not even to the average guy or gal in IT.

Alternatively, we can ditch the Notes For Messaging licences and just have a Community Nifty Fifty, with no involvement from IBM except changing their licencing.

The former solution - IBM signing community works - probably has legal implications, and could be interpreted as IBM approving or supporting the work. I doubt that IBM's legal eagles will like it.

The latter requires IBM to get rid of a licence that most of its community doesn't use and doesn't understand. Even if the upgrade to a Collaboration licence is free, anyone using them will no doubt spot that their next Collaboration renewal will be much higher. That won't be popular amongst all twenty-seven users... (I hope it's less than that, actually.) The solution there is to drop the Messaging licence and drop the price of the Collaboration licence to match the old Messaging licence. Which will make IBM accountants unhappy, I suspect.

As is sadly so often the case, it seems like true progress will be determined by the result of a cage fight between the lawyers and the accountants. (Two falls, one submission, no eye gauging.)

I don't much care which solution we use, to be honest. But IBM has to get involved somehow, or a Nifty Fifty won't be usable by all Notes users and Notes will still be seen by many as Just A Mail Client(tm).

Comments (4)
philipstorry August 8th, 2006 17:52:00

Recently, when not being exceptionally busy, I have mostly been engaging Slashdot users and the Ars Technica crowd in conversation.

The big announcement about Notes on Linux had caused some discussion, so I waded in.

And it was tough.

Very tough.

All the usual arguments were presented. And we all know what our replies would be. But for some reason, I just didn't have the fire in my belly.

Don't get me wrong. I still love working with Domino and Notes. I still want to get other people to see what I see in it. I just got sick and tired of having to beat people over the head with it.

So I was taking a much more measured, restrained approach. I picked and chose my fights - ignoring the people who were obviously working from out-of-date, third-hand, badly worded observations.

It felt good.

And then Ed Brill asked whether or not we sounded like zealots.

Mea culpa. That was me. I've been the loud, aggressive, incensed and (apparently) wronged fanatical supporter, who felt cornered and wasn't going down quietly.

It's easy to see how. It's a hostile world out there.

But I'd like to think I've changed a little. I'm not calmer. (Far from it.) I'm not going to be any more polite. (An ass is an ass. I can't change that.) I do hope I'll be more focused, though.

This conversation on Ars Technica is a great example of what I want to do. I want to engage without being too confrontational. I think that the Notes community has come back just a little TOO strongly in recent months - there's aggression, and then there's arrogance. We're treading the line.

I'm also changing tack on how I open and continue debate. Gone are technological superiority arguments. Not because Notes can't win them - it can, and it does frequently. But technological superiority arguments all too often leave the ball in your opponent's court. And because they just lost a round and are bitter, they bring up stupid arguments (IBM is dropping Notes) or old arguments (The User Interface Hall of Shame). Those arguments are tired, old, and often tinged with personal experience or opinion - not good arguments to get into.

Now, I like to start the debate with one word, and keep the focus on that one word: Value.

You just can't argue with value. And all the technological arguments keep coming back to value, as do the UI arguments (training costs for existing users) and investment arguments (IBM dropping Notes? Pshaw!). Value is the one thing that Notes runs consistently through all the pro-Notes points, and that’s why I’ve chosen it as my cornerstone for debate.

And there's one more change in how I debate – brutal honesty. Be honest about where Notes won't work, or needs more work. For instance, in my opinion, the SMB market is not for Notes. They want out-of-the-box solutions that give a return for absolute minimum investment. Notes can do that, but let's be honest and accept that Notes requires some investment to get maximum return - and some investment is exactly what the SMB market won't usually give. They just want email, and want it cheap. Very few SMBs will actually put the time into getting anything more than email working properly. So the SMB market is best served by a Linux box with an IMAP server and Thunderbird on the desktops. Sorry. It just is.

So with a focus on value, an aggressive but non-confrontational manner, and honesty, I think that I'll avoid being a slashdot-like Notes Troll.

I'll let you all know how that goes. But even better, perhaps you lot could let me know how it goes...

(And to jump back to the idea of Notes doing best when it gets investment, that will be the subject of my next blog entry. Back soon!)

Comments (2)
philipstorry August 2nd, 2006 17:00:00

Why did I not post a single thing during the month of July?

Because, as the title alludes, we had two power failures, one air conditioning failure, and plenty of work to do.

I wasn't completely offline during that time, but I did have less time to post here - and frankly, not much time to think of what to post here either!

Losing no data during all those environmental failures was a great bonus - the problem with shared storage is that if your disks overheat you're kind of stuffed. Overheating was a concern for both the power failures and the air conditioning failure - the UPS keeps the servers running, but not the air conditioning!

The downside to non-shared storage is twice the heat being generated. It's a small downside, though - we just powered down all the cluster members except one, and let failover kick in.

It would be easy to take the advantages I had for granted, but I have to say that I wouldn't want to do July 2006 again with any other product. Not for all the money in the world...

(All the money in the world and a strawberry ice-cream, however, may see me change my mind.)

Comments (0)
philipstorry August 2nd, 2006 13:29:00

A few years ago, IBM had a serious marketing problem with Lotus Notes and Domino. Their competitors were saying it was a dead product - but worse, other brands within IBM were also saying that it was a dead-end product.

Since then, IBM has really improved things dramatically. Their marketing and sales force for the Lotus brand seem re-energised, and their other brands realise that they're not going to replace Lotus Notes any time soon. It's almost as if IBM woke up one morning and realised that it had a $3.3 billion acquisition that it was not just wasting, but was actually harming.

However, because the competitors keep saying that Notes is dead, you sometimes get to wondering what other brands at IBM are saying. After all, I don't really watch every brand at IBM - some of them are just not of any interest to me, because of the kinds of work I do.

Recently, I've been looking at DB2. And I was made very happy by a simple diagram on page four of the book "Understanding DB2: Learning visually with examples".

It stakes out the brands clearly and simply. Rational is for Building, DB2 and WebSphere are for Running, Tivoli is for Managing, and Lotus is for Collaborating. Interestingly, Rational/DB2/WebSphere/Tivoli all feed into Lotus Notes for collaboration - so in brand terms, WebSphere is not for serious collaboration but is for running the big stuff on.

(I'll try to get a scan of the diagram if anyone's interested - but it'll take a short while, as my scanner is currently boxed away.)

But the diagram and the surrounding text are clear and concise, and don't harm the Lotus brand in any way. In fact, to quote:

Rational is the "build" brand; it is used to develop software.
DB2 and WebSphere are the "run" brands; they store and manipulate your data and your applications.
Tivoli is the "manage" brand; it integrates, provides security, and manages your overall systems.
Lotus is the "collaborate" brand used for integration, messaging, and collaboration across all the other brands.

I'm happy to see this kind of thing in print in an IBM Press book. If IBM could print this in December 2005, then hopefully they were circulating it internally somewhat before that. (See? I am an optimist! *grins*)

Now, can someone please post this diagram to Microsoft, so that they can understand this as well? ;-)

Comments (0)
philipstorry June 26th, 2006 10:17:00

Gregg Eldred points us at Google Trends, which is related to a topic I've been meaning to finish up and post about for a while now.

The statistics he's got show a similar trend to the ones I've seen in Google News alerts. Exchange Server gets a lot more mentions that Domino Server in my alerts, providing loads of crud I have to skim.

But what I find interesting is WHY. There are some general trends that I've noticed which distinguish the news stories about Exchange and Domino.

Exchange stories seemed to be far more likely to be third-party add-ons for high availability or disaster recovery services. Recently, that's changed and the most common reason to see Exchange Server is that a new mobile phone model has begun shipping, and can fetch mail from it.

The most common reason to see Domino/Notes? Well, for a while it was the R7 launch, reviews and LotusSphere. Since then, it's been Ray Ozzie. And for the past week, it's been nothing but the SAP integration.

What I find interesting is that Exchange seems to get into my news alerts through third-party products, whereas Domino tends to only get press releases from IBM.

These are, of course, very general trends. Observations made by one person, which are potentially skewed by my own bias.

Speaking of which - since when was high availability something you had to get from a third party vendor? *grins*

Comments (0)
Philip Storry May 15th, 2006 21:16:19

Sean Burgess responded to my previous entry, pointing out that people who seek to balance their work and home life may well not be aware of the risks that they are taking.

I agree - and I think it's down to the the human psychology of ownership.

This is not an easy subject for many people to grasp. It's the problem that content owners, represented by RIAA/MPAA, are grappling with every day in the media world. But put basically, people tend to associate physical possession with intellectual possession.

You bought that CD of music, so why shouldn't you be able to tape it? Or rip it to MP3s? Well, some jurisdictions will allow that under fair use - and some won't. But the fact is that the actual music is copyrighted, and therefore all you really bought was a licence to use the music. Usually, only to use it for personal use as well - if you take a whole load of CDs and set yourself up as a mobile disco, then each public playing of the music is probably illegal.

(Sorry to break that one to you. Yeah, it sucks. But it's true. Check the liner notes on your CDs if you don't believe me!)

People who do things like forward work mail to a home mail account are basically just not thinking about who owns what they're forwarding. They're intermingling company property with their own, and doing it just because they're - if you'll pardon me - ignorant about the issue.

And don't think this doesn't apply to you. Most companies these days include in their contracts a section on intellectual property, where they'll do a massive intellectual property landgrab for everything you ever create whilst employed with them.

I've been assured that the fiction literature I write isn't covered by this, because it's not work related - but that's not how it read on some contracts I've signed. And in future, I'll have to be very careful about this, as I've written software under the GPL licence as well as doing writing, photography and so forth. Future employers may find themselves - unwittingly - asking me to either stop my hobbies, or not work for them. I'm not looking forward to having to broach this with the HR department of employers future, I can tell you...

Anyway, companies will have a hard time educating their staff in who owns what exactly. Which did leave me with one other thought...

We all know that there are mail clients that some people prefer to Notes. I'll use Outlook as an example today, but I think that the same exercise could be done with many bits of software.

My question is simply this - is Notes really working with the psychology of the user, or is it working against it?

The most obvious example between Notes and Outlook is where you move mail to from your mail store (NSF database or Exchange mail store). Notes calls a spade a spade, and the feature is called an Archive. Outlook calls the feature a Personal Folder.

See? With Outlook, it's personal.

Yeah, this is petty stuff, I suppose. But do users react better to Outlook because of these little phrase differences?

I don't have a copy of Outlook to verify such things, but as a quick test I went into the Mail Rules section of an R6.5 mail template. And it says "When mail messages arrive that meet these conditions..." - which is lovely, corporate neutral language. I get the feeling that Outlook probably says something like "When messages arrive for me that meet these conditions..."

I'm probably wrong. But basically, I'm asking IBM to take a long hard look at its interface, and answer one simple question...

Does the Notes interface speak to the people that create an organisation's intellectual property, or to the organisation owning that intellectual property?

Yeah, I could ask whether Notes is speaking to the person or the organisation. It's a simpler way of putting it. But the problem is, I'm not sure IBM would take it seriously if I said it that way - judging by the Notes interface, I think the longer form I've used there is more likely to get noticed... *grins*

Comments (0)
Philip Storry April 17th, 2006 21:19:42

I edited the Wikipedia page on Lotus Notes today, amending a few little snippets that didn't quite match history.

What really amazed me was the way it handled concurrent editing - it appears to do a diff on your submitted text compared with the currently stored text. (I base this on the observation that Richard Schwartz made edits in between me starting and submitting changes - so if it wasn't doing a diff, my change would have written the old text over his edit. See? Very clever.)

This means that two people can edit different sections of the page at the same time, and you get no conflicts. None at all.

Yeah, I know I'm probably very late to notice this - this is the first time I've edited a Wiki whilst someone else was obviously editing it. But I was very impressed with this solution.

It reminded me of the first time I saw replication in Notes R4 - which brought field-level replication to the playing field. On large databases, it was so much faster and more convenient that it just amazed you, even though it was a pretty obvious solution when you thought about it.

(And, ironically, as I wrote this Nathan T. Freeman just edited the Wikipedia page to mention field-level replication! What a freaky coincidence... *grins*)

Anyway, there's quite a bit of work to be done before that Wikipedia page truly reflects Notes/Domino properly, but I think that the community is up to it. If everyone adds their small part, we'll be fine.

Comments (3)
Philip Storry April 10th, 2006 21:31:25

I'm guessing that I don't need to tell most of my visitors how a Notes client is configured - which files do what, and so forth.

I've had debates with people in the past, usually extolling the virtues of storing configuration information in clearly defined stores like names.nsf, bookmarks.nsf and the notes.ini. It makes support pretty easy, and reduces the problems caused by file corruption sometimes.

However, one of the things I've never quite understood is the complete lack of support options from Iris/Lotus/IBM for handling problems with the configuration. It's assumed you'll just know how it works, and roll up your sleeves and get to it.

For occasional problems, that's fine. However, when we did our Notes R6.5 rollout last year, we used contractors to set up each client. (We had to rip-and-replace an old roaming system with the new R6.5 native Windows multi-user one, so re-setting up clients was somewhat inevitable.)

Unfortunately, not all contractors are equal. Some particularly dim ones were there on the first weekend, and an entire floor got set up incorrectly. Rest assured that they didn't come back for a second weekend - we quickly learnt who was capable and who wasn't!

However, every now and again we found ourselves faced with someone returning from annual leave etc. to find that their Notes client didn't work as expected. Specifically, they were trying to send mail directly via SMTP at the client, and their bookmarks weren't exactly what we wanted. The easiest, fastest solution was to clear down the Notes client - delete all the config files, edit the notes.ini - and go through setup again for the user.

With ten years of Notes experience, that takes me about thirty seconds. I'm sure some of my readers could do that faster, as well. (Maybe this should be an event at LotusSphere? *grins*)

But our first line support staff found it cumbersome. So I wrote a little app to do the trick for us instead. And here it is:

NotesCleardown.exe

You run it, pick an option (I recommend Smart Cleardown), and it removes what you want. It always leaves the help files intact, it removes full-text indexes when it finds them, and it always removes the Notes roaming cleanup files for you. Oh, and all deletions go to the recycle bin - if it can't delete to the recycle bin, it won't delete the files at all. The rest of the application is self-explanatory from the display you'll see, really. It works fine in my organisation's environment, but please note that this is by nature a destructive utility - so please do back up before you test it!

I GPL'd the utility, but never got round to releasing it to the general public.

Now, I should warn you all that this doesn't work with R7 clients. Only R6.x clients. But if you have R6 clients, could you please download this, back up your notes.ini and your data directory, and then run it to see what you think? Then let me know if I should continue developing this.

Obvious future plans include supporting the R7 client, and possibly putting in a task-killing capability a la the infamous NotesKill application. But at the moment, I just want to know if this is useful to you.

Oh, and IBM - if you're listening, please note that this solution only runs on Windows. A notes.exe -cleanup option that does the same as my app's Smart Cleardown would be quite nice, and would go some way to lowering the TCO of Domino/Notes even further...

Anyway, download and let me know folks. Feedback welcomed, just leave a comment!

Comments (2)
Philip Storry April 5th, 2006 22:36:27

The Domino/Notes community seems to be pinning its hopes on Hannover to win over dissatisfied users of Notes.

I'm sure it will be pretty slick. But the one thing that worries me is the legacy issue.

The recent Workspace vs Bookmarks issue is a good example. The Workspace represents the R4.x interface - everything that's bad about Lotus Notes in interface terms. Bookmarks represent the newer, shiny interface.

Yet the workspace seems to survive in Hannover, because IBM can't break such a legacy item.

Damned if they do, damned if they don't.

That's just a recent example, of course. The fact is that in some areas, Notes just works differently. A sent view instead of a sent folder, for instance. Folders which contain references, rather than the actual email. (So you can have one message in multiple folders, and it's still just one copy of the message. Deleting it deletes it from all folders, though.)

There are some fundamental differences in Notes, which can't - or shouldn't - be removed. Shouldn't gets my vote, because they're often differences that are very useful and powerful.

And we're still going to have to educate people on those issues, even after Hannover has shipped.

What I'm hoping is that Hannover makes the differences more obvious, and easier to handle. Difference isn't, in itself, bad or good. It's Hannover's job to sell the differences in the most positive light.

Comments (0)
Philip Storry February 14th, 2006 07:00:08

Over at Ed Brill's weblog, there's a spirited discussion on the Guardian article of last week.

I wrote what I thought was a fairly reasonable response, but referred to the Slashdot crowd as "Web 2.0 Wonders" in it. And got called on that as a prejudice.

Continue Reading "Web 2.0 Wonders..." »

Comments (2)
Philip Storry February 12th, 2006 14:17:50

There is much fuss in the Notes world - especially in the UK - about the Charles Arthur article in today's Guardian.

I bought a copy, and was most unimpressed. It seems inherently biased and very poorly researched. But then, the Grauniad seems to have put the article in the Opinion section anyway. At least, I think they have - the page opposite (page four) clearly states Opinion. Sadly, the new Grauniad style seems devoid of any kind of section tag on the page that holds the article, but as it reads like opinion and that's the last section title I saw I have to make the assumption...

It's very badly researched opinion, too. All credibility was lost about a quarter of the way through, when I read the paragraph on the early interface design:

..."Around this time Apple Computer released the Macintosh with a new easy-to-use graphical user interface. This influenced the developers of Notes, and they gave their new product a character-oriented graphical user interface."
So was it character-oriented, or graphical? You can't be both.

Actually, Mr. Arthur, you can be both. The last character oriented graphical program I probably used in anger would have been either PC Tools or Norton Utilities. Both drew dialogue boxes - often animated ones that exploded into view - on a text screen.  And in some programs in those suites, they even allowed them to be moved around, minimised, and moved over/under each other in a kind of z-order stack. All that moving could be done via either the mouse or the keyboard, naturally.

Later versions even re-write the console font in video memory, to provide a slicker look-and-feel with custom widgets for re-sizing, minimising and so forth.

Plenty of other DOS programs used this trick - and I even did it myself with some basic programs I wrote. In BASIC, as it happens - so this was hardly advanced stuff.

Of course, I'm being incredibly petty in picking up on such a trivial display of the author's lack of knowledge. But that's intentional - if the author of the article doesn't get his trivial facts right, then what does that say about the important facts that they're trying to present?

Comments (0)
Philip Storry February 9th, 2006 22:02:06

I'm afraid I'm on the side of Richard Schwartz in the Bookmarks vs Workspace debate. Partly because bookmarks are so much more usable, and partly because they're much easier to support.

Continue Reading "Bookmarks are the future" »

Comments (2)
Philip Storry February 9th, 2006 21:57:02

Sean Burgess has picked up on something said by Graham Chastney:

Influential people hate the Notes client, and they are the people who count.

I think Sean's analysis of this stands by itself, and I won't attempt to add anything more to it. However, I would like to make some observations about Influential People.

In my experience, Influential People do sometimes hate Notes. Not all the time, but they are more likely to dislike it than to like it. This is mostly because they don't actually use it.

I'm being serious here. Influential People are often the least trained, least informed people when it comes to making good decisions about technologies. They're too busy to attend training, so they use their influence to skip it. They're too busy to understand technical issues properly, so they commission high-level papers and reports to condense it down for them. (And, in my experience, it's also not uncommon for the people who write those reports to simply reflect the Influential Person's own feelings back at them with some nice-sounding reasons that will never be investigated. Because report writers want to keep their jobs and be paid a living wage.)

But the most important thing about Influential People is that they often have PAs. Which means that don't actually tend to handle mundane things like email themselves.

I first noticed this when iNotes came out. Influential People who used Notes to work from home or whilst travelling were impressed initially, but soon frustrated. It's not that iNotes/Domino Web Access is poor - it's just that the Web Browser medium it has to run in places serious limitations on what it can do and how it can do it. Domino Web Access - and just about every other webmail client - is quite poor at selection, searching and mass operations. And that's just the start of a long list...

Google's gmail product is the best of the lot, and that relies heavily on fragile scripting, the user actively tagging their emails and on having a very good search engine behind it...

But I digress a little. My point is that those Influential People that used Notes for remote working were initially impressed by iNotes/Domino Web Access, because it looked nice and slick. But after a few days, they stopped using it. When I politely asked, the most common reason was that they just didn't find it as useful as the Notes client.

Conversely, those Influential People that have preferred the iNotes/Domino Web Access system have done so on mostly aesthetic reasons, and do not usually work remotely for longer periods. In fact, they often delegate all such things to their PAs whilst they are away from the office. They use email so little that they prefer not just Outlook to Notes, but would probably also prefer Outlook Express - which nobody could possibly defend as a business email client!

Interestingly, I've also noticed a correlation between inbox management and Notes appreciation. Those Influential People who dislike Notes are also far more likely to have horrendously large mailboxes than those who prefer Notes. This is, of course, because they usually delegate inbox management to their PA without any kind of clear indication as to what should and should not be kept - so the PA simply keeps everything, so as not to get into trouble.

I don't know how this can help the discussion, other than pointing out that you're not limited to DAMO - you could always try them on Domino Web Access, which is probably easier to set up and maintain in some ways. But I was wondering if anyone else had noticed these trends, and if I'd missed any other indicators?

I am, of course, going to decline to draw conclusions at this stage. Influential People might be reading, even if they only found this site via Google - so I'd rather you all drew your own conclusions! ;-)

Comments (4)
Philip Storry February 3rd, 2006 08:45:00

Ed Brill seems to have Ed-Dotted himself to death, by posting his Boss Loves Microsoft slides.

I've just read them, and they're the usual excellent stuff - I recommend that you all read it - but not right now, as the man's website seems to be dying!

Instead, give the man's bandwidth a rest, and go and read these two things first:

• Lotus Notes/Domino 7 application performance: Part 1: Database properties and document collections
• The DigestSearch method for Lotus Domino databases

Both are new articles from DeveloperWorks, freshly posted yesterday. (As you can tell by the odd markup mistake that's yet to be fixed.)

I've seen some of the information elsewhere, but it's always nice to see this kind of thing written about in a lucid and engaging style.

The parts about document collections in the first article are exactly the sort of thing I love - practically useful, yet also a nice challenge to the mind. Domino, like most other development environments these days, has so many ways of letting you do something that you're never quite sure which one is the best. I love the challenge of picking apart performance problems like this in my own coding, trying to understand where my assumptions were wrong.

The second article on the DigestSearch method seems like blue-sky stuff at first, but is really just a re-hashing of an old idea. (There's a clue there, folks!) But again, this touches the whole performance/design part of software that I find fascinating - the fact that it really does so often come down to trading off functionality against performance.

To close, I'd like to distract you all with the wackiest support document IBM have shown us in a long time: Can the messageID in log.nsf be related to a specific e-mail?

The technote says no, it can't. But then strongly intimates that yes, it can.

I think what IBM meant to say is that you can relate the MessageID in log.nsf to a specific e-mail, but you shouldn't rely upon being able to do so.

But it does make for a very strange technote!

Comments (0)
Philip Storry February 1st, 2006 16:37:00

David Schaffer talks briefly about documentation for Domino. And I have to admit that he's sort of got a point.

Continue Reading "Domino Documentation" »

Comments (0)
Philip Storry January 10th, 2006 22:27:26

Chris Miller asks us for our comments on an article on Domino Domain Monitoring (DDM).

Well, personally, I always expected this response to DDM to come along eventually.

Continue Reading "On DDM and how ground-breaking it is..." »

Comments (2)
Philip Storry January 9th, 2006 23:59:37

Yesterday, I spoke about the background and the aims of my Administrator's Manifesto for 2006.

Today, I'd like to talk about the tricky aspects of policies and implementation.

Policies

Policies are the part where, I admit, the Manifesto struggles most.

It's tempting to have just one policy:

• Address the shortcomings I find.

Is that it? Well, in order to accomplish my goals, that would be about all I'd need.

But it still feels like a distinct lack of policy, you have to admit.

There is another area of policy - licensing. And that's REALLY tricky.

I'm drawn, by personal preference, to the GPL as a license. It protects my work from exploitation, and means that the community is also protected from myself - I can't decide to take back my work, or charge for it, or anything else. That should engender the trust that I'd like.

But have I got that right? Will people really take to software that's licensed in that way?

Implementation is the easy part, though. For the most part, everything will be done via NSFs, usually with Formula Language and LotusScript.

I've drawn up a list of areas I want to address, and assigned tasks to "products" - individual databases or sets of databases which do the job. Then I racked my brains for a naming scheme, to tie them all together. Upon finding one, I googled like mad, and I think I'm legally safe on that front.

It's been a busy day of planning, but I think it's been worthwhile.

However, I'm afraid I'm keeping all of the details under wraps until I have a deliverable first product. Hopefully this will happen sometime within the next couple of months.

Comments (0)
Philip Storry January 2nd, 2006 21:51:07

Rocky Oliver strikes again. His Surely Template project is a great example of the kind of thing I was talking about yesterday - the kind of admin tool that's perfectly possible, but not in the box.

I haven't had serious template administration complications since I was managing a development server in a consultancy, where I suspect that this would have been a really useful tool. However, I'm pretty sure that this is the kind of tool which will probably surprise me and be useful in just about every environment.

Rocky's looking for someone to take over the bulk of the development. I'm tempted, but I doubt I'll have time to handle Surely right now, along with everything else I want to do.

But I'll be downloading a copy tomorrow, and having a think about it. I need to know that I can understand and maintain the existing code before I can even think about taking an active interest...

Comments (0)
Philip Storry January 2nd, 2006 17:12:02

Over at the charmingly titled You Had Me At EHLO, a post appeared yesterday about the Exchange 12 64-bit only decision. This is something I stayed relatively quiet about a month or two back when the rest of the Domino world was pointing and sniggering, except to comment on other people's blogs.

The reason for that is that I occasionally dabble in development, despite being a systems administrator. I have a reasonable idea of what sorts of decisions developers are sometimes faced with, and I've collected a few wise quotes from computing luminaries which guide me in my own development efforts. One of my favourites is this:

"When in doubt, use brute force." - Ken Thompson

Yes, THAT Ken Thompson. I think it's safe to say he knows a thing or two...

Continue Reading "The penultimate 2005 edit - Exchange 12 and 64-bit" »

Comments (2)
Philip Storry December 30th, 2005 15:16:42

Ed Brill asks if we're working over the holidays.

Not likely, matey!

That having been said, I am still checking emails and so forth. In order to save a little battery power on my Crackberry, I'm monitoring mail via our secure Domino Web Access-based webmail service.

Curiously, this service wasn't available earlier today. I tried to VPN onto the office network to check the server, but found that the VPN connection was also down. So I gave up, figuring that if they needed to contact me they know how to - and that contacting work might just be an exercise in interrupting them mid-disaster, which nobody likes!

But the webmail server is now back up. Checking my mail shows no automated alerts from the systems, which is a good sign. I fired up the good old Web Administrator application, fully expecting to find that the server was up all the time - indicating that there was a problem with our internet connection. But the server log document suddenly ends, as though the server had crashed or had some kind of other failure... The next log document shows me the server doing a normal recovery restart. Not good!

So I dug a little further. Thankfully, the web administrator allows me to see the Windows Server logs. And what's this about the Event Log Service starting? This certainly wasn't a problem with Domino - I have a hole in the logs, followed by signs of a restart!

So I'm currently logging on to the network via VPN. Maybe they had power problems or something similar... Whilst I'm not technically working, I'd like to know if there was a large scale problem which affected a number of my servers. Give me ten minutes, and I'll get back to you all with the results of my investigations...

Comments (0)
Philip Storry December 29th, 2005 23:38:36

A few days ago, someone put in an odd request. They want a new Send method in the mail template. Not content with normal Send and the slightly saucy Send And File, this person had looked at their mail usage and decided that they'd really like a Send and Remove Attachments option.

This requirement then went to someone, who passed it to someone else, who passed it to someone else, and somewhere along the line this became Send And Transfer To Domino.Doc. Which is probably technically possible, but the idea of people being able to transfer documents immediately into Domino.Doc in such a manner would probably terrify the Records team, who already have to deal with more than enough people determined to get below mail quota by transferring last year's Christmas goodwill messages into our permanent Records storage. (Hint: They're not Records, folks. Whilst it's nice you wanted us to have a merry annual celebratory event, in seven years time we won't need to know about it.)

I was all set to nix the idea immediately, and add it to my ever-growing list entitled "fun things to scare the Records team with". Until I read down the email and found that actually, the person wasn't asking for Transfer to Domino.Doc. They just wanted to remove attachments from emails, in a bulk manner.

Continue Reading "Rocky brings shininess..." »

Comments (0)
Philip Storry December 20th, 2005 00:24:49

So, having foretold doom and gloom and made you all check your servers, maybe it's time to mention just why the Server Controller is so great...

The Lotus Domino Console (as the client for the Server Controller is called) is fantastic because I can connect to multiple servers at the same time, and simply switch between each console as and when I want to. If a server crashes or becomes unresponsive, the controller is usually still responsive. Connecting to it shows me the last few screens of output, which is useful.

Then there's the ability to kill a dead server, so that I can restart it. This is immensely useful on a couple of my more temperamental servers, which like to occasionally monopolise an Agent Manager thread so badly that although everything else will shut down, it just won't. Kill Server solves that one... And the server's still back up five minutes later, which is a good five minutes sooner than it would be if I had to go to the server room, restart the whole box, watch it go through the hardware POST, watch Windows boot and so forth...

And my last, most favourite feature... Being able to send a command to multiple servers, simultaneously. Just connect to the servers you need, and click on the little arrow by Send. Pick Multiple Servers, and select the ones you want individually, or pick a group. Then send your command to the servers.

I find that most useful for getting lists of agents running on my mail cluster... Just to ensure nothing's slipped through!

And lastly, but not least, the logfiles... The Server Controller puts everything out to text files by default, named dcntrlrMMDDhhmm.log (where MMDDhhmm is the date/timestamp - by default, it rolls over every day, but restarts of the Server Controller create new logs as well). These can be searched with tools like grep, making log analysis fast and easy.

If you're already using the Server Controller, I hope this helps you use it better. If you're not using it.. DO SO!

Comments (0)
Philip Storry September 26th, 2005 14:00:00

(Note: This post applies to Windows servers only.)

One of the things that many Domino admins I've talked to over the years have agreed on is really very simple. They all like the fact that Domino has a console, so you can see what's happening on your server.

It's not even that it always shows you the right information - as anyone who uses the http server task knows. It's more that it gives you somewhere that you get instant feedback. Even the most recalcitrant task at least shows that it's starting up or shutting down, and when you're restarting something that's useful to see.

However, it's not all great - that nice interactive window only shows up on the physical console (screen 0), so our Domino servers can't use Terminal Services like all the other Windows boxes can, which can cause friction sometimes when others don't understand why we MUST have VNC on the server... *grins*

Of course, the solution was also quick in arriving - Domino R6 gave us the Server Controller, which is great. Now many admins can watch the server console. You can start and stop the entire server within the controller, and even kill it if you're having problems.

But I still run across people who aren't using it. Perhaps they don't know it's there? Most likely they don't care - I had to edit the registry entry for the service to get the controller options I wanted, and that's a little more hassle than many want.

This is more than a shame. This is actually going to be a problem... Larry Osterman mentioned recently that the Interact With Desktop option for services will not be available in future. Now, I'm sure that this is a planned change which Microsoft are advertising to the development community and others... But it was sure news to me!

So I'd recommend that everyone get used to using the Server Controller in their environment as soon as possible. I'd also recommend that when using it, you use "nserver.exe -jc -c" - this starts the Server Controller, and starts the Domino Server in it automatically. But it doesn't start a Console session. The Console session would be useless, of course, because in the future it won't be visible - just like the Server Controller itself!

And if anyone in IBM is watching, I'd like to suggest that in R7.5 (if there is such a release), you make the default for the service on Windows platforms "-jc -c", to ease the migration.

And in the community, we should also be mentioning this frequently. Preparation is key here, so whenever you meet someone NOT using the Server Controller, get them to look into it. Point out that time is limited for the status quo - it WILL change, and it's better that they learn the new ways now and at their leisure than be forced to learn them quickly in the future.

Comments (2)
Philip Storry September 25th, 2005 22:21:46

For level-headed security advice, Bruce Schnieir is your man.

He points us to the Devil's InfoSec Dictionary - mostly because he's mentioned in it.

The dictionary is short but fun. But the comments on Bruce's weblog are much funnier, providing you skip the Microsoft bashing ones... Some quick examples:

• EBCDIC: An American data encryption standard.
• Password Policy: A written document which makes users write their passwords down and put it on the monitor.
• CEO: An untouchable who should have access to all networkable computers. This person should be regarded as trusted and secure by default.
• Internet Worm: Free test to check if you applied all your security updates
• Security consultant: The person outside your organization that you pay to secure everything you don't want anybody outside your organization to know about.

(Yeah, I edited at least one of those to remove bias...)

This is all fairly relevant to me at the moment, as I'm implementing our webmail system at work. Logging, ACLs, mail journalling, patching - all of these things are quite current to me. But over and above features in the product, we're putting in some serious procedural security - which is to say that when a user requests webmail, we have to do quite a bit of work. Not just creating the replica on the webmail server and adding the user to the access group, but also doing things like creating an ACL monitor for their mail database on that server, and looking at ways of detecting compliance with or abuse of our policies via the web server logs.

We just can't satisfy all our security requirements via ticky-boxes in the product alone, so we're starting from the assumption that if the product isn't already broken, our users will find a way to break it. And it's our job to make sure that's as hard a thing to do as is humanly possible, without actually making the service unusable.

How secure are your systems? Do you assume that your product is handling it all for you?

If this entry makes just one reader look at their system's security again, then it will have been worth writing. :-)

Comments (3)
Philip Storry August 15th, 2005 15:30:00

There are some things that users just keep asking for, and a yearly view seems to be one of them. Yesterday, the request came in yet again from one of my users.

The thing is, she's sure that she's used a product alongside Notes to view and print her calendars in a yearly format at her last organisation. But I racked my brains, and could think of nothing. I know that we used to recommend Lotus Organizer GS for such things, but I didn't think they still shipped/supported that - and I only see the stand-alone Organizer on IBM's website. I'd rather implement something that's supported, and will consider third-party solutions quite happily. Especially if they require little or no deployment. *grins*

So, what do people use to handle the more advanced calendaring requests of their users?

Oh, and IBM - I know you can deliver improvements to the Calendar interface. You did it in R6.x with the Summarize feature, for instance. So could you consider a yearly calendar view, capable of just showing busy days as filled blocks and non-busy days as empty blocks, a bit like a gantt chart? And a group calendar that's actually usable - that's a common request as well. Both these things would please my users, and probably quite a few other people too. So, IBM... Pretty please? With icing and sugar on top?

Comments (2)
Philip Storry July 26th, 2005 10:00:00

Chris Miller has some great slides from his presentation at Advisor Vegas 2005.

One set of which covers optimising the performance of Domino Web Access.

And at work, I'm just securing and optimising our new shiny external Domino Web Access machine.

Great timing, Chris! Thanks for the slides - they're very useful. :-)

Comments (0)
Philip Storry July 25th, 2005 10:20:00

Going back to Ed's posting on the differences between mail clients, Norm Van Bergen asks why Exchange can send Out of Office replies immediately, whereas Notes sends them only every six hours.

Fair question, and it boils down to the way in which the two products handle out of office.

With its strong heritage as a workflow and application platform, it was natural to implement it in Notes using the tools to hand. That meant using scheduled server-side agents. Unfortunately, the agents are a little complex - they not only check the details of incoming emails against a list of addresses/subjects in order to decide whether or not to reply (or reply with a special message), but they also have to maintain a list of people they've replied to already, and update that list when necessary. That might not sound like much, but as someone's out-of-office period continues, the number of people in the list grows, meaning more comparisons every time new mail arrives. This increases the time taken to run the agent. Not by much, but by just enough.

The agent could be run after mail arrives - but if it were, that ever-slightly-increasing workload, combined with the overhead of starting/stopping the agents themselves each time they're run, would gradually kill a server. So instead, the agent is set to run every six hours.

On an Exchange Server, this is no heritage of workflow or application handling. So Microsoft simply looked for something else to use when implementing the Out of Office agent. They settled on their server-side mail rules system. The Out of Office agent is just a special system rule, which never shows up in the normal rules list in Outlook.

So the Out of Office agent is run as mail is delivered, as all other mail rules on Exchange are. Of course, the Out of Office mail rule is considerably more complex than any other mail rule, but that's about it. The behaviour of it is hard-coded into the Exchange server application, meaning it has less startup and running overheads than the compiled LotusScript agent that Domino/Notes uses. It also has no scope for modification, whereas you could change the behaviour of the Domino/Notes agent if you wished.

I'm unsure that we should drop the agent method, as I have no doubt that some large customers have customised the functionality a little. But I am thinking it might be nice to offer an alternative method. Since R5 of Domino/Notes, we've had the ability to run server-side mail rules in mail databases. In R6, the option to forward mail from within those rules was added. So it seems as though we could fairly easily extend these rules to do it the same way that Exchange does. It would have to be a special system rule, as it is in Exchange, but otherwise I think it would be pretty simple to extend the current technologies to implement this - the R6 mail router is obviously much closer to being able to do the job than any previous one has been.

But if we offer both methods, how do we decide which one to show the user?

In the R6.5 mail templates, we offer two methods of deleting mail - soft deletions and the old Trash folder method. The mail template switches between the two depending on whether or not soft deletions are enabled. The same trick could be used here - let the Administrator set either a database property, or a field in the profile document. (I'd prefer a field in the profile document - no need to go adding database properties for this feature, really!) Then let the mail template hide the alternative method's menus/agents and so forth based on that setting. It's probably easier to maintain than the other obvious solution - two versions of every mail template! Which is nice and clean, but would be a nightmare for IBM to maintain.

Am I overlooking something? Is there some mistake I've made? Or could we really have what users want - immediate responses from the Out of Office agent? If so, it would be lovely to see this in Release 7.5...

Comments (2)
Philip Storry July 22nd, 2005 11:47:00

Remember Ed's recent article on what you miss from the mail client you normally use when you're forced to use another?

Hosun Lee and David Price have chimed in that they like the fact that they can simply drag and drop mail from their mailbox to a local PST file.

OWCH! My head!

Guys, this really isn't a good thing. REALLY.

Yes, you feel organised after you do it. And it helps keep you below quota. But on the other hand, it's an Information and Corporate Governance nightmare.

Here's the scenario. Your organisation is being sued. You have a copy of the email that's needed to keep the shareholders happy and your bank account healthy. Except it's in a local PST file.

And you're on holiday.

How is your legal department supposed to find that email?

Don't try to excuse it. Yes, you can put it on a network - where someone now has to get a search tool to search both your Exchange Information Store AND all the PSTs. (Not to mention keep nagging people to cull mail from their PSTs and compress them, because they're running out of space on the server.) It's a maintenance and information management nightmare.

Not to mention that there's always people who don't even put their PSTs on the server. So when their desktop PC (or laptop) goes boom, they lose all their email. Was it backed up? Puh-leeze. Be serious. Why backup a desktop? All the data should be on the server.

And server backups are no guarantee. In a disaster recovery event, not all servers might be brought back up at the DR site. Sometimes, "less important" servers are left off the DR plans. So you could find yourself at a DR site with functioning email, but no idea what that phone number you need is, because it's on a server that's not available right now.

Oh, and anyone using PST files also loses the right to talk about Outlook Web Access, for obvious reasons.

It's not that PSTs are bad themselves, or that Outlook is bad. You could do similar things in Notes (I've banned archiving before now because users do it locally, then lose their data when their laptop dies...), but Notes tends not to encourage it.

Information like how to archive locally (in Notes) or how to create PST files (in Outlook) is viral in nature, I'm sure. Someone finds out somehow, and uses it to stay below quota. Before you know it, everyone is doing it, as the secret gets out. And once the genie's out of the bottle, you're shafted.

I have some remote users who came to us from another organisation, and didn't get a policy to stop them from archiving. Trying to stop them now is almost impossible politically, because they've just become packrats. Worse, they do it on their local machine - their laptop! The only solution we've found is to use the same grapevine that the problem got round on in the first place - we're allowing them to archive for the moment, but when someone's machine dies we point out that the local archive wasn't supported by us, and wasn't backed up. People are slowly stopping doing it as more data is lost, and we'll soon be able to apply that policy and have them grateful for it!

So be warned. Keep it in your mailbox. And if you're an administrator, get them to keep it in their mailbox. By hook or by crook. Because otherwise, those emails are going to disappear sometime...

Comments (4)
Philip Storry July 19th, 2005 22:37:00

So, in my last blog entry, I said that the only thing I missed about Outlook/Exchange was the delayed mail sending feature. And I meant it - it's a nice idea. The implementation's a bit sloppy, though, and I've always thought it could do with some polish. So, here's a proposed feature request that I might submit to IBM for their implementation of delayed mail sending...

Delayed mail sending in Notes/Domino

The purpose of delayed mail sending is to arrange for emails to be sent at a suitable time - e.g. after an embargo has been lifted on information, or when a specific time comes around such as the close of a market or a deadline.

User interface changes required

When composing the email, the user can select an option to delay mail delivery until a specific date and time. When the mail is submitted to the mail server the mail router, seeing that it the message is marked for delay, checks that the intended delivery time has not already passed (just in case!), and then marks the message as "Held, pending delayed delivery".

If a user chooses to mark a message for delayed delivery whilst using a local replica of their mail database, they should be warned upon sending the mail to replicate as soon as possible to prevent misunderstandings causing mail to not be delivered before its intended time.

Changes to the mail.box design

Messages marked as "Held, pending delayed delivery" show clearly that they are pending and their intended delivery time when viewed in the mail.box, to assist the administrators of the messaging system when troubleshooting mail. Additionally, a new view with pending delayed delivery mails sorted by intended delivery time (sorted descending) should be added.

A new action should also be added, allowing an Administrator to remove the delay for a message, causing it to send immediately.

Changes to the router software

In order to prevent abuse or mistaken misuse of the delayed sending feature, the router should have some specific changes made to better support administrators' abilities to control mail flow and prevent capacity problems from occurring. It is by no means expected that all users will need to send mail with a delivery delay - quite the contrary. It is more likely that only specific departments (IT, Facilities, Communications/PR, Senior Management) will require such an ability. Therefore, it is suggested that the following options be made available for configuration in the Server Configuration Document:

• A radio button option, allowing the sending of delayed mail to be:
  • Disabled completely
  • Enabled
  • Enabled for users in the following group(s) in the Domino Directory:
• An option to send "confirmation messages" to reassure the user that their message has been accepted for delayed delivery
• Specifying a maximum time window during which mails can be delayed - expressed in days
• Specifying a maximum size of email that can be delayed
• Specifying a maximum number of recipients per email that delayed delivery allows
• Specifying a time period during the day in which delayed delivery is not allowed
• Specifying a maximum number of delayed mails to queue at any one timeNotification options for emails that do not meet any chosen above requirements:
  • Deliver the email immediately, and inform the user
  • Reject the mail, and inform the user why in the mail delivery failure notice
  • Reject the mail, inform the user and additionally send an email to the following address(es):
• An option for a custom message to attach to all delivery failure notifications to users, in addition to the standard delivery failure message
• An option to check that the sender still exists in the Domino Directory before sending, and mark the message as dead otherwise.

I think all those options are sensible, A maximum forward timeframe during which mail can be delayed is a must-have. The ability to restrict to just one or a few groups of people who can send delayed mail is very important, as it reduces the scope for abuse. (As does checking the user still exists - no sending nastygrams just after you've left the organisation!)

Most of the other options should be self-explanatory, but I'd like to explain one option I put in - the option to send an additional failure message to another email address other than the sending user. That's for the helpdesk staff, to be honest. In a modern IT environment, you don't want the helpdesk staff getting all your administrative alerts - but you would want them getting something like this, because they're going to be getting the call from the user about the strange delivery failure they just got. The option of alerting another email address helps out there, going that extra mile so that your helpdesk can possibly even call the user and explain what went wrong before they read the failure message themselves!

I don't recall the Exchange implementation being anywhere near as complete as this, by the way. I do recall it not delivering mail for users that had been deleted, but that was about it. As a feature, the Exchange implementation is pretty basic. But the Notes/Domino version should be better, and I think that the edge given by that feature list would do the trick. Feel free to tell me if I missed anything important, though...

Comments (0)
Philip Storry July 19th, 2005 21:48:00

Ed Brill asks what bugs us when we move away from our normal mail clients to other ones. Naturally, being Ed, the focus is on Notes, with a particular focus on comparing with Outlook.

Well, I really have to agree with some of the comments made - Chad's comment about the Copy Into New feature is spot on. That even bugs me when I'm using my Blackberry - the very idea that I can't just create a new to-do or calendar entry from an email (or any other combination/direction)  bugs me no end, and it's possibly the most under-rated feature in the Notes mail template in my view.

Subhan's comment about the way Notes handles folders is good - the fact that I can easily put a document into many folders is very nice. I think he misses the more important fact, though - even though the document's in a folder, it still shows up in views. Which means that I always have the All Documents or Sent views for finding those pesky lost emails.

Gerald Mengisen touches on those all-important views, and also mentions Send and File. Send and File is, of course, a spin-off of the view/folder system - but nicely done. He also mentions the fact that you can just hit the escape key and choose to only send a mail, rather than send and save - which is something I use every time I send a test message across my servers, I must admit!

Lyset mentions the Outlook feature to delay mail sending. Good call, that's about the only thing I've EVER missed about Outlook. Yesterday, my mail servers took a hit as someone in Communications sent a 2.5Mb PDF to about four hundred people out there on the internet. The poor router was quickly bogged down by the recipient's slow ISDN connections, leading me to a hairy ten minutes of hoping that SOMEONE on the mailing list had a half-decent pipe. Luckily, it was just a temporary thing, as all the slow pipes clumped together for a short while. As they got dealt with, the outbound queue gradually slipped downwards in numbers. But I can't help thinking that the ability to delay the mail sending might have been useful- if combined with training for users who'd need it, naturally... I'll have a quick think about that, as I'm in two minds about it. Part of me says "just build a mailing app to do it!", and part says it should be something the router can do anyway...

I've thought for a short while about what it is I miss most in other applications when I leave Notes behind, and for a short while I was going to say "Private views" - the ability to create my own quick and dirty view to show something of use to me. But that feels like cheating.

However, it does give me my answer. My favourite feature of Notes is the customisability. The fact that I can change the basic design of the mail template to do something that benefits my colleagues is great. We run with a customised mail template at my workplace, and I'm currently evaluating the OpenNTF Mail Experience design to see what we can (and can't) use from it to improve the experience we give. I'm betraying my administration roots here, but that's certainly what I've always likes about Notes.

Of course, if I think of anything else, I'll be sure to let you all know.

Comments (4)
Philip Storry July 18th, 2005 23:26:13

Chris Miller has started what looks to be a great (albeit probably short) series on the basics of replication topologies.

This is, he correctly points out, not often understood all that well, or tended to properly by administrators. Head over and check it out. I'm sure that this will be a great reference for both newcomers and the experienced alike...

Comments (0)
Philip Storry July 14th, 2005 22:32:51

Just musing, but something occurred to me about the future of collaboration. When you speak to people who've used collaboration systems like Notes, you find that the reason Notes works is that it allows you to take your data with you.

Of course, Notes extends that beyond just data and into the actual use of the data - allowing you to take your business processes (workflow) on the road as well.

In a nutshell, that's the success of Notes. Getting people working when they're not actually connected to your network by a rather short network cable.

Of course, this doesn't come free. You have to massage your data to get the benefits - Notes requires the use of the NSF data store, which is nowhere near as interchangeable as the average Word document. It's all swings and roundabouts.

There's heritage there, driving that particular downside - Notes began to evolve when networks themselves were rare, and when sneakernet was dominant. The novelty of being able to take your data with you so easily was definitely something that outweighed the additional steps needed to get your data into a suitable NSF (or set of NSFs). Notes has benefitted most from visionaries who saw this as an opportunity to look beyond the data itself, and see how people actually work - and then fit the solution around that.

This is important, because all of Microsoft's attempts to collaborate have fallen a bit flat. The reason for this is that they have an enormous advantage that they don't want to lose - by the time they started shipping collaborative software, they owned the de facto standard file formats. Microsoft's attempts in collaboration are restricted by an unwillingness to move away from those file formats.

That makes a lot of sense in one way - you can then try to sell collaboration as an add-on to your existing technologies, rather than taking collaboration as an opportunity to reflect upon your workflows and perhaps improve them. Microsoft's collaboration strategy is one of demanding as few changes as possible from the customer in how they handle their data.

The problem is that the file formats themselves are now the restriction for Microsoft - if nothing else, because they just won't move around as easily as data in a database can. This is why Microsoft's acquisition of Groove is so important.

But it's only important in the short term. I'm going to go out on a limb here, and make a quick prediction - that Microsoft's long term collaboration strategy relies upon WinFS.

WinFS is Microsoft's much-delayed searchable, fast, database-oriented storage system that will replace the use of a folder (or set of folders) on your hard disk for the storage of your files. The desktop version has been delayed, but it still coming. What's faded off the roadmaps is the server-side version, which basically had massive scaling problems. Searching tens or hundreds of thousands of documents, totalling gigabytes or more, is difficult. But there was to be a server-side version, and I have no doubt that it will arrive eventually.

So, there you are with WinFS on your laptop, and WinFS on your network. As both are databases, what's to stop you from automatically exchanging data between the two of them? From replicating a small subset of the WinFS server side down onto your laptop?

I know what you're thinking - you're thinking that they tried this with Briefcase in Windows 95, and quietly retired it some time later when it became obvious it didn't work.

But WinFS will be closer to the way NSF works than the way a normal file system works in many ways. I believe that replication could well be a capability it will have. Maybe not in the standard setup, maybe as an add-on install for corporates. But it's certainly possible.

With the ability to take your data with you, Microsoft are a hop and a skip away from a solution that truly competes with Notes. An application framework, strong security and document routing are all that's needed. With .NET as the application framework, WinFS/Active Directory providing the security and Exchange/Outlook providing the routing, you have a solution.

Of course, the big question then will be which approach works better - an approach which attempts to fit the tools around your workflow, no matter how bad it may be (or good the tools are) - or an approach which puts the tools foremost, and asks that you change your workflow/working practices where necessary.

I have a feeling that the former will probably be more popular, no matter how inappropriate it may be sometimes.

So, do you see the same future I see?

Or am I just bonkers?

Comments (1)
Philip Storry July 4th, 2005 09:00:00

Microsoft's next version of Office will use XML to store the data by default, and that XML format will be something we can get the spec for and use royalty free.

This is, of course, massive. Not very original - the actual format description (XML files stored inside a ZIP file) should sound suspiciously familiar to users of OpenOffice.org's products, for starters. But the Office document formats have been a pain for years, as have most other formats.

Take RTF, for instance. Rich Text Format. What could be more of a lingua franca than the venerable RTF that we all know and love? You know - the one that's akin to a Word 6.0 document from years ago, and hasn't changed since?

Well, even there we're ignoring the fact that there are loads of different RTF specifications from different vendors, and different products use different RTF standards. (Like Notes, for instance - there's a reason that the development documentation constantly calls it "Notes Rich Text", and that reason is to avoid confusion.)

But even with Microsoft's own RTF specification, they're up to version 1.8. Which was news to me, because I thought that they were still on 1.7 - this change must have come about within the last year or so. So much for an unchanging, constant format. In fact, RTF seems to change once per Office release (funnily enough), and what most people think of as RTF is somewhere around the version 1.1/1.2 mark. And many users - even technical experts - assume that the RTF that Word exports is this RTF. Well, I can introduce you to some developers who work with RTF, and have some unkind things to say about how compliant Word if with its RTF export on the current standards - let alone how well many RTF libraries handle the format. The bottom line is that you should never assume that RTF is just RTF and can be read by anything - RTF is a minefield of incompatibility in real life.

Because XML is almost self-describing, I'd expect better compliance from all programs outputting these new XML formats, as otherwise they won't validate. So this is a good thing, I hope. There will still no doubt be the ever-present almost yearly feature creep from Microsoft, which will mean that there will be a mad scramble to implement the new features in import/export filters. But otherwise, this is a positive move towards interoperability.

Of course, I note that only the document formats are open. No opening of a format for Outlook, for instance - nor for Access. Richard Schwartz asks whether or not IBM should open up the NSF format, and I think that the absence of an open Outlook or Access storage format shows us that he's on the wrong track. This is about opening up formats for data interchange - and you're more likely to send an individual item than a whole collection that's stored in a database.

Stan Rogers pointed out in a comment that this has serious security implications, as a lot of the Notes security relies upon the API rather than any inherent security in the format (unless you use the encryption facilities, of course!).

Basically, I don't see the need to open up database formats like NSF. I'd rather see the access methods for those storage formats opened up - which we already have with Notes, through access to NSFs via the Notes client (LotusScript/Formula language/Java), DXL, COM, C and C++ APIs. They could perhaps be made a bit better, but APIs are certainly the way to go when you're looking at a database, and Notes offers the most choice of any database format I know of when it comes to allowing access via APIs.

And most importantly, an API means that we don't suffer massive problems with different sub-versions of NSF out there. If the NSF format was documented, we might end up with the same kind of mess that RTF ended up in. Which would be totally unacceptable for NSF, given how many users it has and the good standing it maintains with them...

Comments (2)
Philip Storry June 6th, 2005 20:05:22

Ed Brill reports that Notes 7.0 will have the ability to filter the calendar view by entry type.

So close, yet so far away... For people with very busy calendars, this might be useful. But I think that they've just missed the mark on this one, myself.

We can already see what a calendar entry is, because they have different icons and different colours. There's plenty of differentiation there. Granted, in the views where there's no space, this will make the calendar a little more readable - providing you know what kind of entry type you're looking for. But that's about it.

Whereas we've had a Category field in the calendar entry form for years now, and not once used it well in the views. If you gave me the option of filtering by entry type or category, I'd pick Category every time - it's my calendar, after all, and Category is one of the ways I can "personalise" it to fit my needs.

I also think that the Category field needs a little more work. I can see why it's not used that much in most environments - it's a free-for-all right now, which means that across a whole corporation it's probably possible to get more category entries than you could fit in the picklist dialogue box, causing an error. So I think that we should have the Category fields for Calendar Entries and To-Dos added to the policies and profiles, allowing both a pre-filled set of categories and a restriction on how many each user can then add on top of those. That would allow some freedom, but also reduce the risk of hitting a "maximum number of entries" error in the interface. The more you think about filtering by Category, the more difficult it seems to do.

Frankly, the other reason the Category field isn't used is that we don't have views for "Calendar Entries by Category" - the Category is rarely, if ever, mentioned elsewhere in the Notes interface. That needs to be changed, because the Category field could be very useful. It's interesting to note that in the comments for Ed's blog entry, although someone asks to filter by room/location/subject/timeslots and so forth, nobody mentions the Category field at all. And although I know I'm not the only one who uses it in my workplace, I also know that it's not exactly commonly used either.

I'm increasingly of the mind that the Category field in the Calendar Entry form needs to either be made useful, or removed...

Comments (2)
Philip Storry June 2nd, 2005 14:10:06

Cliff Reeves is asking some excellent questions at Ed's Blog. I can't possibly resist sticking my oar in here. Specifically, he seems to want to know how we get from Domino/Notes - using NSF as a storage model, Formula Language and LotusScript and a fat Notes client - to Workplace, which is DB2, J2EE/Java, and thin clients using web standards (or protocols layered on top of them).

And I can see where Cliff's coming from. But let me share my revelation with you - because I have seen the future.

And it is DOLS.

DOLS. Domino Off-Line Services. Most Domino administrators know about it, but don't necessarily use it. A few developers may know about it, but even fewer have used it. But if you'll bear with me, I'll show you how DOLS will take us to Workplace, without undue pain and misery.

About DOLS

DOLS is a pretty smart piece of work. It makes use of the cunning architecture of Notes. Because, deep down, Notes seems - to my untrained eye - pretty heavily influenced by old-fashioned design philosophies from UNIX. In UNIX, they like to create one small program that does its job, and does it well - and then they do bigger things by tying those small programs together with pipes.

Notes isn't small, I'll grant you. And even its components aren't that tiny. But then, the things it does are quite complex. And to be honest, size isn't the point here - the modularity is the key.

The modularity of Notes and Domino is most obvious at the Domino Server itself. Want a webserver? Load the http task. Want a pop3 server? Load the pop3 task. Want ldap? Load the ldap task. Want coffee? Leave the server room, please - no liquids allowed! *grins*
But the fact is that Notes and Domino are pretty modular, right down to the fact that the basic server calls itself a "Database Server". You can see the message when a Domino server starts -

Database Server started

- which tells you it's ready to accept connections. Kill that database server, and believe me, many of your tasks will die - because they often work by making loopback connections to the database server. That's good design - they have one good database server, and they re-use it when they need to.

The competition - especially in its infancy - often went with a more monolithic design, in which the servers for internet protocols were just thrown into other tasks because it was convenient. Thankfully, such days are behind most of us, but I thought I'd mention that because it nicely contrasts the two approaches.

And what does all this have to do with DOLS?

Well, DOLS filled a specific need. Having put a webserver onto the Domino server, many were chuffed with their suddenly lightweight, easily accessible applications that didn't need a heavy Notes client. But there was a rub - the heavy Notes client was what made mail usable when you didn't have a connection to the server. And a web-based application was, to put it bluntly, useless without a web server to serve it.

The solution? Well, luckily Notes/Domino is modular. So, let's just take the replication engine, the web server, and anything that they happen to depend upon. Build a quick interface so that you can easily control them. And hey presto! DOLS. DOLS-enabled applications - like the iNotes/Domino Web Access mail templates - can be replicated to a local machine, taken away from the server, and still accessed! The DOLS interface application even includes a replication scheduling interface! Brilliant!

DOLS and Workplace

So what does DOLS have to do with Workplace?

EVERYTHING.

DOLS shows that Domino/Notes isn't monolithic. It's modular. So you can easily start replacing or migrating it BIT BY BIT. (Literally, if you happen to lack a bit-bucket with which to carry those bits... *grins*)

Yes, folks. We're seeing IBM say to us: "Well, we're moving towards that. But don't worry - we can do that over several versions, whilst still supporting each individual technology (NSF, LotusScript, Formula Language) in both the old and the new products."

That may not seem pleasant, but the alternative in the industry appears to have been the VB model - "Yes folks, this is the future. Stop using that old thing, move all your code to this, and everything will be fine.". This model is OK, unless your code base happens to be a few hundred thousand lines or so...

The VB model might seem like a cheap shot, but it's actually a great link to the final part of the puzzle: Modularity makes good platforms, monolithicity makes bad ones. (Although interestingly, they can both make good products. This is partly a scalability thing...)

Workplace and Domino will be two parts - well, actually, two collections of parts - of the same platform. The modular underpinnings allow them to interchange and interact far better than if each one were just a huge monolithic product. The VB migration model was disastrous not because you were migrating from an old version of the product to a new one, but because VB.NET was part of a platform. You were migrating from a product to a platform - which is never going to be a pleasant experience, because there are likely to be differences - possibly huge differences - in how the product and the platform work.

Conclusion

A platform has to be modular. Domino and Notes were already a platform in that regard, so moving to a new platform is always going to be easier because all you need to do is get a set of compatible modules to move towards. Companies that migrated from the Domino platform to the Exchange Server product were often disappointed because of the great difficulties involved in migrating, and then upon completion of the migration finding out how much flexibility they'd lost. This is typical of the platform/product distinction.

DOLS is a perfect example of how a product often thought of as monolithic - the Notes client - is actually modular, and how that platform's technology can be used in an almost unrecognisable manner. It proves that a gradual migration by intermingling technology modules between Domino/Notes and Workplace is not just possible, but achievable without great efforts for anyone currently using Domino/Notes.

Comments (0)
Philip Storry April 11th, 2005 21:54:00

I'm not dead yet.

And neither is Notes/Domino, by the way.

I was most disturbed to see e-Pro Magazine close down, but I can see why.

Because here's something really interesting, folks...

For the last month, I've been too busy - both at work and home - to follow much on the Domino/Notes Blogosphere. At home, I subscribe to about 20 blogs of developers and administrators who work with Domino/Notes - and, of course, the Ubiquitous Brill. (You can't escape him in this game, I tell ya!). At work, I only managed to subscribe to a few - Maybe five. So I've been lurking, vaguely following those few core people. Meanwhile, my RSS reader at home continues to slurp feeds in, and I don't have time to read them.

It's now the 11th of April, and the last time I looked at any feeds appears to have been before Whisky Live - on 11th March. And I have 312 articles to read.

312.

Wow. I've got lots of reading to do. But you can see why e-Pro Mag would find it difficult to keep up with that. For a diverse group of admins and developers, many of whom have not actually met each other, that's a pretty hefty figure for a month.

And I'm sure that's just the bloggers, by the way. I've marked the couple of thousand or so articles from Slashdot, The Register, Boing Boing and the like as read - leaving just Domino/Notes material.

I have to say, it feels good not to be dead. But better to be in such illustrious undead company. *grins*

Which has led me to wonder about the future. But you'll have to wait for the next posting to read about that...

Comments (0)
Philip Storry April 11th, 2005 20:36:00

Some analysts (well, PR shills actually) continue to say that Notes is dead. And yet Microsoft goes out and buys Groove... Let me tell you what I think that means.

It means Microsoft knows that Notes isn't dead. It also means that they feel very threatened by IBM's Notes/Workplace strategy.

To find out why, we have to go back to the days of the DOJ trial against Microsoft. It produced possibly one of the most interesting documents I've ever read on Microsoft - the Findings of Fact, which declared Microsoft a monopoly and threw some interesting light on the way Microsoft thinks. I recommend everyone should read this document at least once, and possibly that they should dip back into it every time Microsoft makes some kind of strategy announcement or an acquisition.

That document tells you that Microsoft felt very threatened by Netscape Navigator, because it had the potential (however small) to deliver applications to a desktop without anyone needing any Microsoft technology at all. Microsoft is very protective about its platforms. They see themselves as a platform company, not a product company.

In trying to defeat Netscape, they threw away a lot of money by giving IE out for free. But they also knifed a pretty big potential cash cow - MSN. They had to open up their desktop to an AOL icon to get AOL to use IE instead of Navigator, and that made MSN "just another service provider". Not good for a service which was, when launched, supposed to kill CompuServe, AOL and The Internet within a few years. Remember - MSN was originally the only on-line service on the desktop of every shipping Windows PC. Not a small advantage...

So if Microsoft is willing to throw away an entire revenue stream to protect their platform, they're most certainly a platform company.

What does this have to do with Groove?

Windows is one of Microsoft's platforms. Office is another. And then they have their "Backoffice" platform.

Groove sits in both the Office and Backoffice platform camps, and does things that traditionally Microsoft has not been very good at. But Notes has been very good at some of the things Groove does - well, Notes and the rest of IBM's offerings like SameTime, Quickplace etc...

When Microsoft last invested in Groove, they put 38 million dollars into them. Stick your finger in the air and feel the strength of THAT breeze, and guess how much Groove must have cost Microsoft. OK, Microsoft has deep pockets - but they didn't get where they are today by wasting money buying junk they don't need.

The purchase of Groove seems to indicate that Microsoft are worried. I'm starting to see some nice parallels with their past behaviour when they were worried that something threatened their platforms. I'm not saying the Notes is a Netscape Navigator, of course. But it certainly can't be dead - and IBM's Notes/Workplace strategy may just have them a more than a little bit scared...

Comments (0)
Philip Storry March 11th, 2005 12:09:00

I went out for a quick drink, and nine hours later when I return I find that MS has bought Groove.

Luckily, me drinking and MS buying aren't (as far as I know) related - or MS will have bought the whole IT industry by eight o'clock on Saturday evening. (More about that tomorrow...)

But as usual, the moment Notes is mentioned the Slashdot Hordes start up with their usual "I used Notes and it was awful" drone, which has long been established to actually mean "I used Notes back in 1997, for a week, without understanding what I was using it for, and I didn't like it, and I think I could replace the world's number one groupware product with one webserver and a few PHP scripts...".

The great thing about having been drinking whisky since three in the afternoon is that it tends to make your mind function in, um, alternate modes. I was about to wade in and show them the light, but luckily my brain knew that I needed food and sleep. So it threw me a curveball to keep me less enraged and - hopefully - make me just write a single blog entry before getting some kip...

In my somewhat inebriated state, a question occurred to me that I'd never actually asked before:
"Why, as late as 2002, were these people using Notes R4.x?"

The only possible answer is - in my current state - blindingly obvious. The infrastructure that was put in place much earlier had happily scaled, so that in 200x Notes was still happily trundling along just as it had back in the nineties, often with a gap of five years or more in rollout and the upgrade to R5/R6.

When the R4.x interface was unveiled, I think its major competition was - in no particular order - cc:Mail, MS Mail, and perhaps MS Exchange Client (Remember that? It came before Outlook, shipping with the first release of Exchange Server). cc:Mail was also from Lotus, and died because Lotus wanted to concentrate on just having one messaging platform. MS Mail scaled like a paralysed chipmunk that forgot its climbing kit but still turned up to the mountain-face. The MS Exchange client remembered its climbing kit, but got a cramp fifty meters into its ascent, and decided that it would simply lie and say it had scaled the mountain, and hope nobody ever tried to make it do it again. (MS have been talking about scalability - and high availability - ever since, but have delivered very little, despite ripping & replacing Domains for Active Directory...)

Whatever way you cut it, the companies that were still using old versions of Notes with horrific interfaces were doing so because it still worked. It still met all their needs, and probably could have done more for them if they wanted it to. Notes was a victim of its success - it may not have been pretty, but it was an honest and hard worker who could be relied upon and trusted.

Not bad for something that's supposed to be a piece of crap, really. So here's an open challenge to the Slashdot Hordes. Install the oldest copy of PHP you have. Preferably a copy from the same time as Notes 4.x, which would probably be PHP 3 or earlier... Now implement a solution in it that does, say: mail, discussion, workflow for some simple things like holiday booking, and a knowledgebase application. Then leave it for five or more years.

At the end of those five years, have your needs changed? If so, does your little PHP-based website still meet them?

Answers on a postcard to /dev/null, please...

Comments (0)
Philip Storry March 11th, 2005 00:30:32

Much of this week's time at home has been spent at work anyway. Virtually speaking, that is.

Our R6.5 client rollout approaches, and the fact that we're replacing an antiquated manual "roaming" system makes it a complicated operation.

But I want to get it right first time. And to that end, much of my development over this past weeks has been focused on a database I call "Automated Client Fixes". This database can:

• Reset client's network address cache and delete/rebuild all connection documents at the press of a button
• Automatically delete the standard location documents that we don't want, and rename the ones that we do to the names we want
• Add bookmarks in a bulk, automated fashion
• Rename the dreaded "Workspace Tab 1" bookmark folder to something more... Homely

Nothing much to tax the development muscles there, you'd think. And some of it was ridiculously easy.

The first two, for instance, were quick and easy. But bookmarks?

Ah, bookmarks. IBM want us to use them. We can create them automatically with Setup Profiles (R5) and Policies (R6). We can roam our bookmarks with R6. We get nice ways of looking at our bookmarks, and in R6 they even tried to make them look like the workspace... You get the feeling that bookmarks are definitely where the future is, these days.

But can you deal with bookmarks programmatically? No. Not very easily, anyway. We still have ways of adding workspace icons and selecting them. But the only way you can add a new bookmark is via @Command([AddBookmark]). Nothing via script - you can't Evaluate @Commands, and the various options which say they'll add a bookmark as they add a workspace icon were... Unreliable.

Or, in other words, I couldn't get them to behave consistently. Especially not if there already was a workspace icon. As the Automated Client Fixes database will be used to support users after the rollout - in a "Go here, click this, it's fixed" manner - I need something that works every time. If a user deletes their Bookmark but still has the workspace icon, I need it to create a bookmark anyway - not just select the workspace icon and forget to create the bookmark. Which was not exactly documented behaviour, I might add...

In the end, I rolled out my Formula skills from the dusty corner I keep them in, and forged ahead with the @Command. It was either that or spend a while reverse-engineering the bookmarks database and tinkering in it directly with LotusScript.

(OK, I admit it. I half-tinkered anyway, and now understand how the bookmarks database works - hence we gained the renaming tool for the ugly "Workspace Tab 1" default folder. But I decided that it was better to go with the supported method and use the @Command, as we're on a tight time constraint right now. We can't afford to call up IBM and say "Hey, we've stuffed up 1000 bookmark databases." We can afford to call them up and say "Hey, @Command([AddBookmark]) stuffed up 1000 bookmark databases." Sometimes, the decision is just that easy to make, no matter how much you may feel that there's a geekier solution available...)

So what's bugging me now is the programmatic support for bookmarks - the feature we should all be using, or so it seems. After some work, we now have a database which shows a list of databases of our choosing, and from which the user can pick databases and we'll then add bookmarks for those databases for them.

Why did it take me almost three hours to get that working, and yet it only took me five minutes to add some security to the database definition documents, so that not all users can see all databases?

I think this is why some developers have a love-hate relationship with Notes & Domino. In any other system, that apparently simple UI feature would be the easy part, and the security of the system would be hard. But in Notes, the amazingly easy and the infuriatingly hard are not always what you'd think they'd be...

(Or maybe I'm just very, very rusty at development these days...!)

Comments (0)
Philip Storry February 24th, 2005 21:14:03

Many things attracted me to Notes at first, as a product.

Although I'm an Administrator by trade, the rapid application development aspect of it was one of them. It was the one I took the longest to figure out, but it's also the one that suddenly made me appreciate what Notes was all about.

In this age of wonder, we're used to seeing apps that do all sorts of things. But to see a WYSIWYG form designer and a quick view creator that can build a whole application back in 1995... That was something special. That was Notes R3.33.

Notes R4.0 brought LotusScript to the fore, and as I'm a BASIC man I loved it.

But I always stuck to Administration when it came down to the choice. Mostly because there's more control over my own work there - development is very hard to manage, and there are precious few good managers in the software industry. So I always felt under more pressure to hit impossible deadlines on projects that had no research or planning behind them, and it was a fairly easy choice in the end.

But that doesn't stop me from developing. I'm one of those strange people who feels that a good administrator can look at a problem and know that one of three things will get them the information they need - a by-hand count, a full-text search, or building a custom view. A great administrator knows which one is the most time efficient for their current problem. And a really great administrator has templates full of the design elements they need!

I lack the templates, but I'm slowly building them up now. I have a few projects I'm working on, and I hope to be posting something that's approaching finished soon. None of it is stellar - it's all "scratch-an-itch-ware" - but I'm going to GPL it and donate it to the wider world, in the hope that it's useful to someone.

Of course, first comes the mandatory tidying up. And that's what this is really about - I want to use tools that I was happy enough with that I could say to the world, "Hey! You can use this too!".

My current schedule? Ship code by next week.

Ship documentation by 2010.

Ship bugfixes before Longhorn arrives.

*grins*

Comments (0)
Philip Storry February 22nd, 2005 21:27:48

A while back - before LotusSphere - a call went out for some examples of Notes being blamed for things it's not actually at fault for.

My entry went in, and I thought nothing more of it. Work continued to go more and more nuts, and it's only recently that I've gotten back into the land of the active, so to speak.

So I started to catch up on things. I'd been meaning to read some of the LotusSphere stuff, as I missed it except for headlines and summaries. A natural place to start was Ed Brill's sit, where he'd posted two PDFs of the presentations he did.

And here's my infamy - page five of the How To Sell Notes Internally PDF.

Wow. It's me alright - loquacious as ever, taking up a whole slide to say "Notes didn't print, but then neither did anything else." Nobody can fake that kind of blathering, in my experience!

Oh, how I wish I could have been there. To be at the 'Sphere as my words are pushed, by means of bright projection units, directly onto the retinas of a horde! It's just like a scene from all my world-ruling fantasies, you know!

With less whips and leather, obviously.

Anyway, this should show you all how far behind I am - I'm only just now catching up on LotusSphere.

So, to wrap up: Big thanks to Ed and Libby for picking me from a crop of fine, fine quotes.

Congratulations to Declan and Mick, who were also credited with quotes on this issue. They've also earned a slot on my private list of competitors for world domination, and will therefore be crushed underfoot by me when they least expect it...

I'm presuming that Libby or Ed are the source of the final quote, as there's no name there. They escape the list, because I never add someone to it without proof.

But they'll be scrutinised closely from here on, I assure you!

Comments (0)
Philip Storry February 21st, 2005 20:08:18

Many of my recent troubles have been caused by GroupShield, a product I'm steadily beginning to loathe. It even seems unable to shut down properly half the time - if I issue a quit command at the server, I end up with a hung server that never quits. So I restart the machine, and get a fault recovery email from my server when  comes back up, because the last shutdown wasn't clean. Which I then have to explain to my boss, who takes great interest in these sorts of emails...

So, does anyone have a recommendation for any antivirus products for Domino which aren't GroupShield?

Preferably one which works, does both on-access and on-demand scanning, and rarely - if ever - marks all my mail as dead for the fun of it, or decides to use up all the handles on the server because the name of the day has a vowel in it...

Comments (9)
Philip Storry February 15th, 2005 23:52:29

Do you have to organise meetings in your daily job?

It's a pain, isn't it? Getting all those people into one place at one time... Of course, your computer systems can help here. A good calendaring & scheduling system will be able to tell you when people are and aren't busy. So you won't need to call around and get everyone's schedule, instead you can just look it up yourself. Handy!

But if the system already knows everyone's schedule and can give each one to you, then why can't it check them to see if there's a common free time that everyone has, so that you can meet then?

You're in luck. Microsoft is going to introduce just that feature in the next version of Exchange. They're currently calling it a "Smart meeting picker".

So, in 2006 or 2007 - when Exchange 12 ships (Microsoft haven't created a "ship date picker" yet, it seems) - you'll be able to pick a suitable time for your meeting quickly and easily.

Or you could be using IBM's Domino/Notes combination - where this feature has been in there since Release 4.5, which shipped in 1996. Yes, by the time Exchange offers you this feature, Notes will have had it for a decade.

And here's something to think about: I confirmed which release of Notes this feature was added into by finding a technote from IBM on a problem with it.

A quick glance at that technote shows that not only is this feature mature and stable, but it's been reasonably well thought out. Which the Exchange 12 solution won't be able to match until at least 2016, according to my back-of-the-envelope calculations.

Comments (0)
Philip Storry January 20th, 2005 21:45:00

Going to the one-day training course today was a bit of a risk, I must admit.

Much as I wanted to find out how to Deal With Difficult People, I knew that we'd had some pretty serious mail problems in the past week.

They started last Friday... One of our mail servers died. This is, of course, bad. The server had run out of space for transaction logging (we use archival style logs for backup purposes), and so it died in a grand fashion. One quick restart and the server sorted itself out, which was quite handy.

Two hours later, a second mail server did the same thing. A second server in the same cluster, no less. Of course, I wasn't too worried - I have three servers in the cluster. But it took a little more coaxing to get this server back up, especially as it lost is cldbdir.nsf database and therefore dropped out of the cluster.

I'd come in early that day to do some unrelated maintenance, arriving at quarter to seven. I left the office at twenty past nine in the evening, having coaxed two servers back to life after losing their transaction logs. On Monday, I was going to look at what had caused this in greater detail...

Of course, the plans of men are never of concern to technology. They are a mere fantasies, to be destroyed whenever possible.

I jest. Technology had very little to do with what happened on Saturday morning. Incompetence at our ISP had everything to do with it. In some routine maintenance, involving moving us to a new netblock, they managed to completely forget to update any DNS records.

So when I got in on Monday morning, I was confronted with a "no mail" situation. Nothing was coming in. And because our ptr records were wrong, very little was going out either.

And the cherry on the top was the fact that our ISP's relay server wasn't configured to relay mail for our domain.

I think my director has the private parts of our ISP account manager on a small plaque in his office now, actually...

So, as you can imagine, there was something of a little trepidation in being out of the office today. Thankfully, everything seems to have gone without problems.

Comments (2)
Philip Storry January 19th, 2005 21:18:26

Ed Brill asks "what gets blamed on Lotus Notes?".

Despite not having finished tinkering with my blog design yet, I'll respond.

I can think of two examples of Notes being blamed for things that are outside of its control.

The first is "Notes won't print".

Everyone who works with Notes knows that its printing engine is a little old, and can be cantankerous at the best of times. So when a problem with printing from Notes came into the helpdesk, they escalated it to me.

I went out to the person's desk, and asked them to reproduce the problem. They opened up an email, and picked an attachment. It opened in Word, and sure enough, it wouldn't print.

I carried out one quick test, then politely excused myself and returned to my desk. I then passed the call back down to the second line support staff, with the note attached: "Nothing prints. Not notepad, nor Word, nor Excel, nor Notes. If you get Notepad, Word and Excel printing, I'll have a look at Notes. Thanks."

Interestingly, this is a case of both the user AND the support staff immediately blaming Notes. The user obviously required training (which I did give on the spot, but to have said that would have sadly interrupted the flow of the tale...), but the support staff should at least have asked if anyting else could print... Instead, they simply stuck it in the pile of Notes problems.

This example is a good indicator of a larger problem - users who just aren't aware that when they open an attachment, they're usually going into a seperate program. I've heard users curse Notes for features it doesn't even have...

My second offering is also partly a training issue - but mostly a cultural one.

For an entire group of users, Lotus Notes is slow. They're a group of users most organisations that have Notes willbe familiar with - remote users.

They dial up, onto our network, and then connect to the Domino servers and send/replicate their mail. This takes too long, apparently.

Yet if you look at the mail statistics and see who sends and receives the largest mails, it's the departments who have predominantly remote users. I'd like to put a maximum message transfer size in place, to help users stay below quota (amongst other things). Yet I know that because of the way our users work, it will predominantly be the remote users who will resist this.

Which is doubly crazy when you realise that we have Domino.Doc in place, and they can send a http or DocLink link to documents from within there, thus removing the need to send large documents everywhere by email. The only people who can't use Domino.Doc in our organisation are those who work in Finance with linked spreadsheets - virtually none of whom work remotely.

This particularly isn't helped by one department, who were merged into our organisation from another one. They used to use Outlook, and didn't all have laptops back then. So naturally, Outlook (usually on a network) was much faster than this horrible Notes (usually over a 56K dial-up) that we make them use... :-(

It's late, and I think I've responded far too rapidly to this entry anyway. Not a good start, considering how carefully I picked the name for this blog. :-)

[Edit: Before I turn in, I think I should just point out that both Dan and Andrew Pollack make an excellent point - users typically spend more time in Notes than in most other applications, and will therefore be more likely to find faults there anyway. Just think of how many man-hours of flawless performance Notes must deliver world-wide in one year...)

Comments (0)
philipstorry January 10th, 2005 00:58:00