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.
I'm afraid I'm not going to apologise for that phrase. I used it because the context was correct, and because there's no other term that I'm aware of for the class of technical people I'm referring to.
Some people think that LAMP (Linux, Apache, MySQL, PHP/Perl), or Ruby On Rails, or whatever emerged fifteen minutes ago will solve every problem in the world. And I'm not going to knock the technologies - they're often very good ones, which are very capable. But they're also often not actually suitable for replacing whatever's being discussed.
The "Web 2.0 Wonders" will, in my experience, usually not have even considered scalability, reliability and availability. They made their decision based on how easily they could get a working product up and running - not on how suitable it is for the requirements.
They will knock products that they don't understand, often without having thought about why the products are successful. This doesn't just apply to Domino/Notes, but to other products as well. But given my expertise, let's stick with Domino/Notes.
They don't understand the need for data to go offline, they don't understand the need for a worldwide replication network, and they don't understand the need for high levels of security.
Show them any Notes app, and they'll say that they can do it for you cheaper and faster in their preferred technology. Which may or may not be true, but that's not the point.
They probably didn't even consider that not having a replicating, distributed system means that the organisation will have to upgrade its WAN links worldwide, because they'll only have one central application rather than a replica at each site. They ignored the fact that you might need to have a cluster, to ensure that you have no downtime in the event of hardware failures. They didn't even think about the fact that a centralised solution is going to be used in many different timezones - so now there might not be any "out of hours" that you can do maintenance in, because it's always business hours somewhere in the world. They ignored the security implications of having to put these applications into a DMZ, so that people can access them from outside the offices securely - and the implication this has on data security. And the solution might not be available for offline use at all.
But things like scalability, reliability, availability and a distributed architecture are often required by business - precisely to avoid these kinds of issues. Not all of those things, and certainly not all the time - but they are important. And few of them are ever considered by the sort of people who push "Web 2.0" or even "Web 1.0" as a solution to all problems. Often, the closest they get is bragging about uptime...
This doesn't mean that there aren't solutions to this problem. You can cluster their solutions. You can use more lightweight designs to reduce bandwidth utilisation. You can try to make it work on a phone's web browser to help those out in the field. But all this is making the solution more complex, more expensive, and more difficult to support. And suddenly, the costs are going up and this "Web x.0" thing looks like less of a solution when compared to your current (but apparently archaic, nasty and expensive) Notes application.
Of course, Domino/Notes isn't p
erfect for everything. Some applications can't be distributed, and therefore
you have to have WAN links worldwide for them. (Airline ticket booking
systems might be a good example of this, especially if I recall my Notes
R4.x Application Development Level 1 exam correctly.) But the Domino/Notes
system allows you to at least dedicate more of your WAN bandwidth to such
applications - which are, let's face it, likely to be mission critical.
I will remain prejudiced against anyone who tells me they can set up a Linux box, Apache, MySQL, PHP and phpBB in a couple of hours, and that this means Notes is dead. Or crap. Or horrible to work with. Because these people aren't focused on what their employer or customer needs - they're focused on how they can do the least work for the job. In other words on how easy they can make their own lives. And I'm afraid that issues of scalability, reliability and availability just aren't easy.
Of course, I'd like to apologise to the Web 2.0 people who do get it, and are building appropriate solutions. (I know you guys are out there!) You guys aren't "Web 2.0 Wonders" - you're just technology consultants who are getting a job done. Which sounds boring and maybe even a little offensive - but it's meant as a compliment, I assure you. *grins*
In summary, yes I'm prejudiced against people who say that they're web developers, or system administrators, but then show now real knowledge or concern for the hard arts of scalability/reliability/availability/distribution. And I'm pretty proud of that prejudice, as I think many others are too.
Comments (2)Philip Storry February 12th, 2006 14:17:50
So, Philip, why don't you tell us how you really feel? Try not to hold anything back next time.
I completely understand where you are coming from and agree for the most part, especially when it comes to security. IMHO, web developers outside the Domino world don't know how to do security at a granular level and don't care to. They work in an all or nothing world or they expect their javascript or PHP to limit the access for them. True security done at the database level is foreign to them. I see that now in the helpdesk system that is used at a client of mine. It's built on Oracle and the security is all done by the application, so if you can get access via SQL, you can see anything you need to. Not exactly what I call real secure.
Sean---
*laughs*
Yeah, I guess that came across pretty strongly. But I am sick and tired of hearing the web 1.0/2.0 message for inappropriate means.
And of course, few things rival Domino for attention to detail in security. :-)
Discussion for this entry is now closed.