For coordinating mapping efforts, you're welcome to create a page on the OpenStreetMap wiki. You'll just need to name the wiki pages to avoid clashes somehow. For now I've added you to this list at least.

Note to self: GIT for SVN users

If you're creating that page, add a fat link to this page: osm.wiki/Video_tutorials

Welcome! Contact the community if you get stuck.

Ah yeah mini... micro. Confusing. (I've swapped it) Garmin's take microSD, the seriously diddy size. It's also confusing that you can chose between several manufacturer brands of SD card, including one called ..."SD"

Ordnance Survey "StreetView" could be useful source. Here's how OS StreetView looks for Soho. So yeah. It includes coloured areas for buildings, but nothing very accurate. As far as I'm aware the vector dataset they released is about the same. Could be wrong though. They have not released the accurate building footprints of their "MasterMap" product

I think if we sketch in building outlines from Yahoo, we can probably arrive at something slightly more detailed. Also it's vaguely more interesting than tracing orange areas of StreetView. How boring would that be??

I noticed you've been circling London with landuse. Great stuff. Is that all just based on yahoo?

The funny thing about landuse is, it really needs to be done to a certain level of granularity. If you try to go too detailed with it, then you end up describing the use of individual buildings. It's a bit tricky in the city centre. Maybe the whole of Soho is commercial (?)

The difficulty is choosing targets to go for based only on yahoo imagery plus the data we have so far. London OpenStreetBugs could give us some targets. Maybe I'll pick a pub near a cluster of those, but there aren't actually that many bugs.

I had a few thoughts about POI gathering while I was on the bus yesterday. TimSC's somehow filled in a bunch of landuse data include commercial areas e.g.. I don't know what he's basing that on, and I can see he's missing a pink area by Belsize Park, so that's weird, but anyway it's potentially a helpful way of finding areas which should be rich in shop POIs. I wonder if there's some other ways of improving landuse coverage in broad brushstrokes. Mobile app to use while on the bus, where you just sit and categorize the building types as you pass.

I want to see more expansion of building outline coverage, but people have mixed opinions about that. Everyone has their own priorities. For example I really want to avoid mapping housenumbers because I find that utterly ridiculous! :-)

"torquing the data" What does that mean?

I take it there's some tension between you two. Think I'll try and stay out of that one. Hope you can work it out though. There's enough work to do fixing up U.S. data without people squabbling!

Oops. Turns out we are making some progress with ref fixup. See my update diary entry

Aha. Steve's blogged me. That's why this has got some attention. I thought it was just the green and red graph getting everyone excited. Incidentally I've just discovered something which might mean the graph is entirely wrong! checking now.

@Paul Johnson - I've added a reply explaining my thinking about relations on the talk page there.

@RussNelson - I've added some more details about refs with ';'s in them to the the wiki page. So I 88; NY 7 for example, reveals a curious gap caused by a short length which is tagged only "I 88" not "I 88; NY 7". To find problems, you tend to look at where the ends of the blue lines are pointing, and ask yourself why the waychain is ending there. Local knowledge can be useful. For example is "NY 7" just another name for the whole length of "I 88"? in which case there should be no ways tagged with just "I 88".

@Jean-Marc Liotier - Yes although it might not be very interesting. In any country where motorways have been mapped by the OpenStreetMap community (gradually, "manually", one by one) The data is of a much better quality than the U.S. TIGER data, and this crude tool would just give false negatives.

Aha. Steve's blogged me. That's why this has got some attention. I thought it was just the green and ref graph getting everyone excited.

@Paul Johnson - I've added a reply explaining my thinking about relations.

@RussNelson - I've added some more details about refs with ';'s in them to the the wiki page. So I 88; NY 7 for example, reveals a curious gap caused by short length which is tagged only "I 88" not "I 88; NY 7". To find problems, you tend to look at where the ends of the blue lines are pointing, and ask yourself why the waychain is ending there. Local knowledge can be useful. For example is "NY 7" just another name for the whole length of "I 88"? in which case there should be no ways tagged with just "I 88".

@Jean-Marc Liotier - Yes although it might not be very interesting. In any country where motorways have been mapped by the OpenStreetMap community (gradually, "manually", one by one) The data is of a much better quality than the U.S. TIGER data

Well the intention was to discover other more important problems (oneways pointing the wrong way, bits of interstate without a way in each direction, etc) But unfortunately it does seem to have mostly uncovered botched up 'ref' tags, which is a less interesting kind of problem. I don't think we need to add ref tags in many places. It's mostly fixing weird problems with the exiting ref (and int_ref) tags. Some of it requires a bit of local knowledge of how these roads are named.

Haha. Looks like a lot of fun!

Don't really know the pros and cons myself. I know that git is super-fashionable these days.

Well according to the README "you might as well just sell your soul to Adobe" :-P

Yes. At the moment there's only one renderer which makes makes maps which look nice, and Mapnik has a reputation as a mysterious C++ black box with hideous stylesheet language built on a massive "enterprise" grade database. Developers are making progress at improving the ease of installation though (just recently actually) You can actually run it as a quite a lightweight thing rendering a localised tileset reading direct from osm.xml and using split out css style or spreadsheet formatted stylesheets. As much as anything else we need some good friendly tutorials on doing these things.

Other technology stacks are fun of course. I can't really take tiles@home seriously, but I do agree that other rendering systems written in other languages / using other data access methods could at least foster a bit of competition. So far nobody's really given Mapnik any good competition though. Maybe Kosmos comes close to Mapnik's antialiased non-text-overlapping goodness. Nobody pays much attention to poor old Kosmos because it's M$.NET

Mapnik is built on top of PostGIS, which is a "proper" spatially indexed database. The tile data servers idea is kind of a half-assed naive attempt at doing something similar. ...but as I say, different technology stacks are fun. I did play with OJW's tile data server and phprender (before it broke) I haven't looked at "Roma" much but I imagine if somebody bridged the gap from roma to Mapnik, that could be an interesting combo.

Wow! Just spotted your work on the Salvador map. That's great! I did a little mapping there bit back in 2007 when I went on holiday there. It's looking a lot better now!

There's a few rifts in the OpenStreetMap community concerning the tags proposal process and in particular it's seen as quite troublesome the way people get excited about making new tag proposals, so I'm afraid you may be stepping on a minefield with this (see my talk http://www.harrywood.co.uk/blog/2009/10/04/community-smoothness/ )

There was a discussion a few months about the idea of having big group discussions at the 'State Of The Map' conferences, where we could agree to some sweeping changes to tags. Your first proposal might come to pass under that kind of framework.

I set up a wiki page for the event: São Paulo Dec 2009 Mapping Party including your cake diagram!