I found a lot wrong with this article. But the two that bothered me the most were these.<p>"Reduced scalability and increased maintenance"<p>The author claims code "can never be future proofed" because of different browsers. I don't see why that's so.
I personally use a renderhtmlforbrowser() function that uses a switch statement. So all I have to do when a new browser comes out is add another string to the switch statement and another function for the renderhtmlforbrowser function to call.<p>"It's Rarely Required"<p>The author says "Web standards have made browser sniffing far less necessary on the server" but I disagree with that. Standards have made things easier in that you can create HTML that won't be completely mangled by one browser while working fine in the other. But anyone whose tried knows that a site's appearance can be completely different in two separate browsers even though they are both standards compliant (even with CSS). So anyone who cares how their site looks still needs to pay attention to all browsers
Most the non-IE browsers work pretty well and get updated quickly to fix problems. IE needs the workarounds and for that you can use conditional comments (<a href="http://www.quirksmode.org/css/condcom.html" rel="nofollow">http://www.quirksmode.org/css/condcom.html</a>).
x100.<p>Mixi, the Japanese social networking service, has a mobile site. I was trying to surf to it via my DSi; but they refuse mobile connections from anything but a mobile browser that is recognized in their DB. They might have complicated that further by some awful IP address restrictions (thank you so much, jphone ruby package /sarcasm).<p>Result? I need to load the full website on a game system that simply cannot handle it. It cannot compare in the slightest to experience to the experience loading mobile Facebook. (Although to be fair, Mixi cannot compare favorably to anything done after Friendster anyway).<p>Please people, unless you really have to - don't browser-sniff.