[rwolf]

The tone is overly-bombastic, but there's the seed of a good idea in here. Getting the a consistent look in FF3 and IE6 is Sisyphean; making a IE6-only layout much easier.

The author recommends dropping the whole "one app, one look" conceit and shuffling the dinosaurs to a dinosaur pen, with a simpler layout and the same content (you have been separating content and presentation, right?).

[tobyjoe]

There is definitely some bombast there. I've been fed up with folks displacing their frustrations onto users instead of getting creative and practical with the issue.

I've been in this game a long, long time and I saw (and, admittedly, took part in) the "insult and block" approach with Netscape 4. Looking back, it was immature, cocky and lazy. I hate seeing people repeat the same childish mistakes. Instead, we can build an economic model to decide how to classify each user agent for each property we control and act accordingly. It's not emotional or dogmatic - just practical.

To paraphrase a reply to a comment on my site:

I think my approach helps bring innovation while continuing to support the people stuck on less modern user agents.

Reclassifying IE6 into the category of feed readers, screen readers, and printers is a perfectly viable solution.

Oppose that to the childish antics behind blocking content based on user agent.

In each case, you get to provide interesting experience design for modern user agents, but only by reclassifying IE6 for content support do you also retain (some portion of) the IE6 audience.

[pbhj]

Is anyone really suggesting going back to the days of blocking on sniffed UA? I don't think so.

When people say "not supporting IE6" I think they mean no longer doing browser specific bug fixing - that's what I mean. Provided the majority of textual content is viewable I consider IE6 done (ditto links2 my only currently tested text browser). If text is not viewable (as in an MS site I visited recently with IE6!) then the client has to pay a little extra if they want to up the support.

Graceful degradation is surely the lowest level of support any designer will accept, no?

So I'm targeting XHTML1-trans and CSS2.1 with enhancements for JavaScript and CSS3 capabilities.