I would a few extra rules:
- MIME type is either application/xhtml+xml or application/vnd.wap.xhtml+xml
- page size (markup only) is under,say, 20kb

Luca

All Transcoding Proxies should Not Transcode if any of the following apply:

1) The page has a Mobile Specific DocType.
2) The Content Type Header being sent out is Mobile Specific.
3) The Cache-Control Header contains ‘no-transform’
4) is a .mobi domain name.

I really think folks are slamming Vodafone too much. I actually think the User-Agent behavior is correct since the proxy changes the header to say “you are talking to a proxy now, not the phone itself!” I think what people are really after is a way to say “don’t transcode me! I want to work with the phone directly.” The technique accomplishes this and works on the Google transcoder (which incidentally should be taking as much flak, since it does the same thing!) and maybe the VF proxy too already.

Now we can argue about whether this is the right standard, possibly in a productive context like the W3C. We can argue about whether any burden should rest on the content author in this regard. And, clearly this change was relatively sudden and disruptive for practitioners, regardless of whether it was right, and so maybe that could have been handled better. But those are the real issues here, not User-Agent this or evil-ness or sucks that or any of this other moral outrage. Calm!

This proves that you are not that great expert you would like the world to believe:
- first GAP and BPWG are very different (and incompatible to a large extent. If you follow GAP, you are not BP)
- GAP totally discard one-web, which is the foundation of BP
- GAP is based on the capabilities of real existining devices. BP is based on the wishful thinking of the companies that happened to sit in BPWG (your included).
- Among the plethora of document that BP is based on, there are some that were authored by (suprise surprise) me before W3C even realised that the mobile web existed.

This article (the author of which I am in not affiliated with) provides a good summary of the difference between the two:

http://www.fiercedeveloper.com/story/editor-s-corner/2007-03-13

don’t forget to read the comments too.

Luca

Duly noted.

…and thank you for confirming that this affects “the W3C MWI mobileOK which means it’ll affect .mobi – .mobi guidelines.”

David Harper

@David, “This implementation is affecting delivery of W3C mobileOK and .mobi standards-compliant XHTML-MP Basic as well.”

I think it does affect the W3C MWI mobileOK which means it’ll affect .mobi – .mobi guidelines are MWI guidelines with a little tweaking to make them more accessible. Jo Rabin does help ensure they both end up doing the same thing – which is good.

Please note that my post wasn’t covering the debate. All my commentary is directed at the termonlogy of WAP and Web – they’re two entirely different technologies, user requirements and marketing issues.