Some impressions, in no particular order, from the day:

1. There was friction between some of the vendors represented; you could detect this particularly between the content-adaptation guys and the browser guys. The former offer a server-side solution, the latter client-side, and obviously each believe their approach to be superior.

2. The underlying idea of all this is to grow usage of the mobile Internet - that's a very worthy aim, and one worth remembering when we get mired in the inevitable politics of an effort like this. Despite differences in emphasis and opinion, every speaker and panelist was obviously getting behind this.

3. Chris Yanda gave some fantastic stats. 28% of WAP users only use the BBC WAP site - they don't visit the web site. That's 250,000 exclusive mobile users - most impressive. And the graph he showed of BBC WAP usage growing was interesting too... That said, the cynic in me is wondering how much effort is required to grow usage of the mobile Internet if we're at such a steep gradient in the curve right now. I'm looking forward to seeing the consumer research which .mobi are conducting and putting into the public domain too.

4. It is clearly compulsory to have one slide in any presentation referencing the "long tail".

5. After an initial gut reaction against .mobi which involve no small amount of retching (DNS does that to me - commercially, politically and technically), I'm actually not so against the idea now. It seems to me to be more of a marketing effort than anything else; as Jo Rabin said, it's about trust challenges not technical challenges. And I can't see what the downside of providing standards and support for mobile and web developers is: if it doesn't work out, doesn't get used, or isn't worthwhile then it won't have pissed on anyone elses chips.

6. I still don't understand what "one web" means.

7. I'd like to understand a little more more how the working group which may end up building an open database of handset capabilities will work alongside the existing, successful and well-deployed WURFL project. There wasn't much mention of the WURFL (though admittedly they only touched on the notion briefly) from speakers or panelists, which surprised me - because "in the wild" it's an valued tool for all sorts of folks... many of whom were in attendance at MWI.