Someone checking out my last entry about CDN performance helpfully let me know that my site was displaying horribly in Opera. I’d checked in Firefox, Safari, and Chrome and hadn’t noticed a problem and just assumed I was pretty safe in other browsers since I wasn’t doing anything especially amazing.
A few minutes with Browser Shots later and I found out that it was broken in at least Opera 10 (8 and 9 seemed to work fine, oddly) and IE – all versions, including 8. I confirmed the IE8 problems in a virtual machine (compatibility view didn’t help) and dug in with the Opera developer tool Dragonfly.
Curiously, Dragonfly’s DOM view only showed the first div in the HTML that contained the header. It took a few minutes of head scratching before I started toggling options and idly hit the button to expand the HTML tree… Suddenly everything was there. I ran the W3C Validator again, just to be sure it wasn’t my imagination, and it checked out fine.
It didn’t take long backtracking up the tree to find out where I’d gone “wrong”:
<button />
Apparently Opera 10 and IE wanted the full closing tag:
<button></button>
I’m not surprised IE choked on the short tag, but I’m really disappointed in Opera… I’ve never been happy using it, but I always respected their rendering engine.
Year-End Browser Stats
Ed Bott published fresh browser stats this morning, and I thought I would comment on some aspects…
The point that Ed makes about the lack of increase in Firefox’s market share is disappointing, but are we really surprised? I’ve said for quite some time that Firefox is geared towards the tech enthusiasts among us and that it really offers no hard benefits for your average every-day user.
Back when Mozilla was competing only with IE6 (we’ll continue ignoring Opera and Safari), it offered great benefits like tabbed browsing and native popup blocking, etc. Unfortunately, by the time it caught any ground, Microsoft had already usurped a great deal of its momentum by releasing the most-needed features in IE7. Sure, Firefox’s amazing extension support offers a lot of flexibility to those of us who consider ourselves power users on the web, but does the average person who only has one computer need 10 different bookmark syncing extensions or the inspection capabilities of Firebug? No…
The most interesting thing I see in these stats is the market penetration of IE7.
I run some basic stats on all our sites at work (mainly to let me know what cool stuff I can and cannot use), and IE6 still has an 80% lead over IE7 in our user base. Since we’re getting traffic from totally technically inept users, that’s a somewhat unfortunate statistic. Even more distressing is that IE has a 95% lead over all other browsers combined.
Clearly both Mozilla and Microsoft have done a fair job marketing their newer products to technical users who keep up with such things, but they’ve failed miserably at extolling the virtues to the average user — and that’s something that needs to change.
How do we do that? I haven’t a clue… I develop the stuff, I don’t market it.
The really interesting question, given the recent announcement that IE8 has passed the ACID2 test, is whether this will matter in a year. If all browsers follow standards properly, do we care which browser anyone uses? Of course if IE8 doesn’t drastically improve upon the market adoption of IE7 thus far, it may take another 10 years before everyone is seeing the web as it was intended to be seen…
Tags:
Post Meta
Atom