The web, like most new technologies, was going to make life easier, speed delivery, cook you breakfast in bed.

And it did.

My life is easier thanks to the web - I can do research faster, I can order things without getting dressed, I can order from companies I would never have known existed and get that one geeky bit that never would have made it to Green Bay, Wisconsin, the Technological Mecca of the Midwest United States.

But have you pondered what it's done from a software perspective?

The web browser was designed to be a thin client, no intelligence, just a rendering engine with the ability to send bits of data back to the server.

But let us test today's reality. Do a clean install of your OS of choice, make it minimal for use as a GUI desktop. Then hit the web.

The first thing you find is that you need to install Java. Once you're done, you'll hit a site that requires Flash player. Need documentation? Download Acrobat. Then you hit that inimical site that requires the .NET engine. And don't even think about going to iTunes and downloading anything without installing something.

The web requires more software than we needed before it came along (unless you unistalled your word processor and are using Google's now. NOT), and in many organizations, the responsibility for installing that software falls increasingly on users. If it's not part of the corporate install then users have to hunt it.

Is it easier? Yes. You just go find it and it installs in most cases. But that doesn't change the fact that the thin client model is gone, tossed to the side of the information super-highway like the bag from your MacDonald's breakfast. You need disk space, you need processing power, and now, instead of just the OS sucking up resources you have the OS and the browser. The ever larger browser.

Is that necessarily a bad thing? No, if you as an IT staff member keep in mind that your users rely upon you to do their job and make certain that all of this software is part of the standard image. Disks have gotten huge, memory is cheap, and each rev of chips seems to add cores, so the only loss is in time invested. You can invest it once in the desktop/laptop image, or your users will invest it over and over on their own.

Performance is still an issue - many corporations run IE, which ships with limitations on the number of connections that can be open. Other browsers do too, but IE has to be changed in the registry, so if you lock the registry down, make certain the settings are set to something acceptable to your organization before the image is burned. In fact, most users shouldn't be mucking with browser config files either, so make certain the number is sane for whatever browser your corporation is settled on.

And we as an industry should be careful about how far we let this trend go. While I'd love to see our acceleration products fly off the shelves to solve your browser performance issues, we don't want to recreate the client/server environment with the client being the browser with 10,000 add-ons. If we're going to go there then why did we bother to upgrade our client/server systems? You haven't yet? No worries, you will.

The 90s were a great time to be in computers, lots to learn, lots of fun. But the progress was forward. Let's keep it that way, and not go revisit the 90s with a fat-client browser.

And don't even get me started about browser 'security'.


/imbibing: Mt. Dew

/reading: See yesterday's post.