Infrequently Noted

Alex Russell on browsers, standards, and the process of progress.

+1.5 Years: Where Are We Now?

I wrote this post well ahead of the release of IE 7, and I've been holding back on writing a follow up until things seemed reasonably settled. A year and a half since that post, and almost a year since IE 7 hit the streets, I fear things are more settled than anyone would like.

The "worst case scenario" that I've described to folks in private for a long time is that IE 7 is the end of the line. The last hurrah. The final gasp of life in Trident's creaking limbs. IE 7 could either signal the beginning of a renewed commitment to the web by Microsoft, or it could be the minimum MSFT can get away with to prevent customer mutiny. To assuage the latter scenario, I've directly and firmly asked every member of the IE team I've talked to since then to outline in public Microsoft's commitment to new versions of IE. We need to see timelines, feature targets, and distribution plans for those releases as well. This might seem like putting the cart before the horse but I think it's not too much to ask. In fact, it might even be the minimum the web development community should expect.

To contrast the IE commitment to a better future with Mozilla's, it's instructive to look at the roadmaps for each. Brendan Eich has an entire blog dedicated to communicating outward about the features that we can expect from the web as delivered by Firefox (and the platform behind it). The IE Team's blog is eerily silent on the future of what is still the most important browser on the internet. We're reduced to getting information from third parties and conference talks. The features planned for Firefox 3 are impressive and the work is being done in the open, meaning it's easy to have confidence that not only will Mozilla ship what they say they will, it'll be here when they say it will. Same goes for the excellent work the Safari team has been doing. Even Opera keeps its community on fire by shipping regular updates, showing tech previews at conferences, and blogging about the progress being made on many fronts. If the IE team is holed up working on something stonkingly good, they certainly aren't doing themselves any favors by not telling us about it. The result of their radio silence isn't mystery, it's distrust. Deep, divisive, troubling distrust of the kind you can only get when folks who break up stop talking altogether. The problem is that IE still has half of my DVDs and my friends keep telling me it's just using me. Not a great way to rebuild a relationship.

Part of the distrust springs for what wasn't in IE 7.

We did get some good things, of course. The tireless vendor-focused work of Molly Holzschlag and other (relative) pragmatists in the standards community helped get a lot of the most egregious issues fixed. But for as important as those fixes where, they were important in a historical context. They helped enable better versions of what we started to develop with in 1998.

When I caught up with Ted Leung at OSCON this year, we had a long discussion about "RIA's" in general, but in talking with him it became clear to me how little of what we (rich app developers) needed was delivered by IE 7. Of the 10 items on my off-the-cuff list, only gzip issues and memory allocation have been substantively addressed. And quite frankly, it was a reserved list; it only included things that might have been hard but shouldn't require fundamental architectural changes. Progress has been made on memory leakage, but we're still finding new leak patterns that need guarding against.

And then there are the regressions. On my list from last year was VML, and far from fixing its performance IE 7 broke critical VML functionality. Yes, yes, VML is old, bad, silly tech, but then so is the rest of what we've got to work with out here on the open web. Microsoft has subsequently released Silverlight which we're using to hack around some of VML's liabilities for limited situations, but we might as well be using Flash or Java at that point. Silverlight has nearly zero market penetration and it's yet another "draw in a little box" solution which isn't Open Web tech. Yawn. The IE team also removed the only WYSIWYG infrastructure that was worth a damn in their browser. The only reason web developers didn't raise holy hell over this is probably that every other browser is worse than even the IE fallback position.

Judging IE's progress feels like making excuses for a rotten kid with absentee parents: at some point explanations cease to matter and the behavior itself needs to be addressed. I'm not giving up on IE's potential, but it's harder and harder to trust that the baby steps taken with IE 7 are going to solidify into a pattern of real progress when we've gone a year without any real communication from Microsoft as to what's next.

I'm pretty sure the IE team isn't sitting still. Chris Wilson is heading up the HTML 5 working group and there's reports of some real progress there. HTML 5 is the most important web spec under consideration anywhere so this is truly good news. But it hasn't yet been accompanied by the kinds of communication that allow us to trust MSFT as a custodian of the web's future.

Getting IE 7 and watching it ramp up among IE's installed base has been good, but it's only half the answer. The web needs to know, unequivocally, when we can expect more information about IE.Next, what OSes it will target, and what standards, improvements, and major fixes are on the roadmap even if they slip. Without that much honesty, this relationship probably won't get off the ground again.