This reminds me of the software article I just wrote

 - 
08/14/2009
So, I just got done writing this quite lengthy special report on what "stability" means for physical security software, covering everything from testing procedures to metrics, network compatibility to the stability-feature dynamic. Most people agreed, referring to that last bit there, that in order to have robust features you've got to have a long track record, as each feature must be thoroughly vetted, and you can't thoroughly test a brand-new piece of software that has a million features. It's just not possible to test them all in a real-world environment. Rather, many companies (especially the established ones) said, you need to start with a relatively simple and stable platform, and then build on that proven technology by adding feature sets, testing each feature thoroughly before it's added. Makes sense to me. Funny that today I should run across this article, telling the browser makers to just knock if off. No new browsers, please. Basically, the jist is similar: sure, this new browser might be really cool, since it's based on a new engine and everyone else is working off an older engine. But, the thing is, those older engines sort of work, right? They might have some drawbacks, and we might all fight over which is faster, but many of the bugs have been kinked out. With this new browser, RockMelt (a crap name for a browser, if you ask me - sounds like a bad sandwich), the bugs will be all brand new, unless the vetting process is remarkable, which it probably won't be (I'm just cynical by nature).
But let's look at the flip side of that statement. When you build something from scratch, you still have to make it work with all the other stuff that exists. The Web is a vast place, with billions of sites and countless plug-in technologies, many of them considered a standard part of the Web. There's Flash, Acrobat, QuickTime, Java, JavaScript, CSS, and HTML. Now, you can build new interfaces for all of that and force the companies that own these plug-ins to work within these standards to support your browser. However, none of that may be very easy.
Wait a second - I thought IT had standards... Doesn't everything work with everything else? But I digress...
I imagine that these companies have a hard enough time supporting the growing list of upstart browsers. Google Chrome, for example, which has about 2 percent of the browser share, is never first on the list when someone's building a plug-in. My favorite password manager, LastPass, is still working on a Chrome toolbar.
Hmmm. Remind you of any security industry problem you know of? Exactly - every new VMS vendor that comes out is sort of a pain in the ass, isn't it? Not that I begrudge them their business models, but don't we have enough VMS vendors by now? Shouldn't we be working on whittling them down rather than adding to them? Anyway, if you get through the monster that is that software article I wrote, feel free to drop me a line of feedback. I'll be working on follow ups as the year goes along.