A bit o'travel has left me with some time to run through the mobile web best practices document. Some thoughts typed up:

1. The document states that its "primary goal is to improve the user experience of the Web when accessed from mobile devices.". Does this imply that there are many situations where the document doesn't seek to apply? e.g. in any WAP/XHTML-MP services being produced (which are mobile-specific, not web sites)? Or am I reading the document too pedantically? To me, the web on mobile seems to be distinct from mobile apps which might share elements of infrastructure or technology.

2. Given the comparative rarity of devices which can browse "the mobile web" today (we're talking PDAs and a few high-end handsets aren't we?) then am I right in thinking that this document is focused more on a hypothetical future of 2-3 years away than around services being built today/tomorrow?

3. I'd agree with most of the recommendations. They seem in some parts to be head-thumpingly simple and obvious, but I've been very close to this industry for some time now so that's to be expected, I think (and IMHO one characteristic of a good idea is that it's obvious in retrospect). I don't believe these are necessarily obvious to a target audience of web developers.

4. I'd be interested to hear more about the advantages of mobile devices (section 3.7). This might provide a more positive counterweight to a document which is otherwise on a fairly negative slant (in that it's otherwise focused on limitations, things we cut out of the web to make mobile, etc.)

5. I'd agree with the assertion in section 4.2 that adaptation is required to deliver all non-trivial mobile services. We even saw this on the fixed internet (netscape vs ie) - sure, we got the users to make a decision as to which version they see (mmm that's a nice user experience) but it still happened.

Overall, this seems like a reasonable starting point to me. I don't see anything here that I violently disagree with; much of it seems vague but I suspect (not that I've ever been involved in an effort like this) that with many vendors pulling in different directions, any consensus is a success.

One weirdie tho: why is the semantic markup section there, when even the group admits it doesn't know what it is? Is TBL standing over them with a Big Stick?