The current issue of A List Apart has an article by Aaron Gustafson, one of those brave articles that is sure to cause a storm of un- and mis-informed comment. It’s a real attempt to solve a problem, and one that shows more than a little bravery.

Eric Meyer’s reply (published side-by-side in a point/not-quite-counterpoint approach) covers a lot of the thought process I went through when encountering it. I’ll freely grant that Meyer, having had more time than I did to come to grips with the idea, may have thought this through deeper than I, but I’m still at the “Ooooh, ick” stage with this idea.

I can see the point, and as I said I congratulate Gustafson for taking the bull by the horns in such a bold manner, but I cannot agree that he escaped goring.

I’m afraid I see this as an attempt to offload the responsibility for good design, to pass the buck, as it were, from designers to browser makers.

The principle of version targeting, as I understand it, is to allow a designer to specify in the source code what browser was the intended recipient of the markup. It becomes, then, the responsibility of the browser manufacturer to render that page according to all the known warts and bugs of that particular browser version.

The chill that ran down my spine as the full implications began to dawn on me hasn’t stopped, yet.

All My Sins Remembered

Since this approach is being pointed towards all browser makers, not just Microsoft, it now makes it necessary for Mozilla, Firefox, Konqueror, Safari, Opera, iCab, and the rest to build their own versions of IE5, IE5.5, IE6, and IE7, bugs and all. So now every maker except Microsoft, if they take this initiative seriously, is required to recapitulate the evolution of IE. Good browser makers now would be expected to deliberately write code that doesn’t conform to the spec, just because the Big Boys in Redmond once decided not to. How can this be considered a Good Thing? And how long will further progress be delayed by the time it takes to build in all these errors?

And if you think this is simply an anti-MS rant, stop and consider the other point of view. Once again, if Microsoft is going to take this as seriously as we appear to be expecting others to, then MS will now have to take resources away from other projects in order to duplicate the foibles of earlier versions of Netscape, Opera, FireFox, Safari and others. How much time will that effort take away from progress toward worthier goals?

Andy Clarke is already upset over the glacial progress toward useful design features. Can you imagine how much longer we’ll have to wait while the browser makers catch up with the arduous task of replicating each other’s bugs?

The Law Of Unintended Consequences

And what will be the effect on design? The production of new browsers with better rendering has, in the past, been one factor driving design refreshes. The effect of this proposal will be to take away that particular factor. While I can’t and won’t claim this, by itself, will be sufficient to completely eliminate the design refresh, I have to believe it will have a noticeable effect on the practice. And I don’t view that prospect as good.

I can see this further retarding forward movement. Are we expected to believe that while writing code to correctly render the standards was so difficult it took nigh unto a decade to achieve, it will be a trivial matter to recreate the bugs in past versions of everyone’s failures? That somehow it’s easier to write code that fails in a specific manner than it is to write code that doesn’t fail at all? I think, rather, we’ll suddenly have a quadrupling of the bugs. Will the solution then be to demand a feature that will allow IE9, for example, to emulate Safari 4’s buggy emulation of IE6? Where do we draw the line?

Clients are already notorious for accepting whatever Explorer does as being the standard of behavior for the web. By bringing this into play, it will become even harder to explain to a client why the way IE5 rendered a web site is not correct.

The Immortal Electron

While pondering the implications, it struck me that one of the things being lost in this was the nature of the web itself. For all we like to use terms implying durability (information architecture, document structure, robust design, etc.) the web is, well, a web. Something lovely, something useful, but something that is, at its root, ephemeral.

And isn’t there something fundamentally wrong with demanding that the ephemeral never change?

The web is delicate. Some days I feel like the guy on the old Ed Sullivan Show who had to keep running all over the stage trying to keep all those spinning plates balanced on top of the sticks. But even at the most hectic, when I dread the phone ringing, it still seems right, somehow.

I dunno. Maybe I’m just an old frontiersman grumbling about the dandies moving in from out east, but it seems to me the web will lose a great deal of its dynamism if something like Version Targeting becomes the norm. There will cease to be a difference between the web and print. For some, it’s possible this would qualify as “a consummation devoutly to be wished.”

To me it just seems like death.