Please stop listing every single change you make, it gets tedious scrolling through all the almost identical posts in my RSS reader - imagine what it would be like if all OSM contributors did the same thing!

However, longer posts with more detail about your mapping over a week, for example may be more digestable to most readers of the diaries.

(I'll just note that there's also been large arguments over calling 'namespaces' namespaces, since apparently although they appear like them, they aren't namespaces, again, refer to the mailing lists)

Further to our discussion on IRC, the software used to translate osm to the mp format had been announced several months previously on the mailing lists.

How could I have missed this? I realised the stadium was new and massive as I passed it on a train on the East Cost Main line, afterwards I looked at OSM to see where the tracks went, as the complexities of the routing through the area fascinated me. Never did realise it was missing though...

I hope we have other large stadia such as Wembley!?

It's been in a source branch for the past 5 months, we've all been waiting on it to be merged to trunk, and it finally has.
It is great functionality, and will hopefully allow OSM data to be indexed by search engines (how useful this will be, we've yet to determine)

So, well done to crschmidt to the databrowser.

Have fun! I find tracking down cycle routes a fun, and infuriating part of OSM!

This follows the way I feel bus relation roles should be handled. The ordering and numbering of roles is complete bs.

Go ahead and split the ways, correct data is always more important than (most) other issues.

Take a look at this wiki article for sample code:
osm.wiki/index.php/OpenLayers_Simple_Example

It's good to see someone who knows the area filling in areas of Central London in high detail.
As an aside, I saw you deleted a section of the Sutton & Mole Valley railway line, then redrew it. Assuming this was an accident, it is possible to undelete ways using U in potlatch (the online editor).

Why are people so obsessed about nodes and areas?

Please be aware that if you've used the TomTom's basemaps as a source for the positioning of the POIs, then they could be classed as being derived from that map data (and thus incompatible with the OSM licensing).
A bulk-import would therefore be unlikely to be allowable. However, plotting the points in OSM (referring to Yahoo, and existing OSM data), and copying across your own data would be allowable.

I've looked, and given up trying.

The 2002 map that I have is closer to what is signed on the ground than the more recent pre or post 2007 maps.

Regarding the blank blocks, there is a z12 'database' of seatiles that should always be rendered as sea automatically.
Take a look here for details: http://svn.openstreetmap.org/applications/rendering/tilesAtHome/tools/png2tileinfo/

I find the 'Draw large GPS points' option in the JOSM preferences (first tab) essential, since I can't usually see them otherwise!

Right clicking on a locally loaded gpx file will link up the tracepoints, but it's not recommended for the downloaded gps points - no timestamps are provided so the line drawing can't determine which points belong together.

Take a look at the Multipolygon relation wiki page for details on this, and a step-by-step guide how to make them.

Hmm, a reasonably good response, I'll push an update to remember the last layer selected ASAP.

The map features wiki page is the comprehensive document on how to tag stuff. If you can't find something there, just make it up!
Potlatch has some help also, you probably want to read this section of it.

If you like bleeding edge software, checkout OpenLayers trunk (its fairly stable), compile it, and then you can just set:
projection: new OpenLayers.Projection("EPSG:900913"),
displayProjection: new OpenLayers.Projection("EPSG:4326")
on the map options.

"I guess my wish would be for Tiger not to be reintroduced in this area anytime soon. ... One way to keep editors happy is to not nuke "their" data."

I believe that the Tiger data was tried to be imported without too much conflict with existing data. AFAIK, counties were only imported when someone requested them, or if that county had no data in OSM at all. Obviously this process was not perfect in this case.

Tiger is intended to be a one-time import into OSM, as a 'useful' base layer to get started on in places with no coverage. Tiger's data accuracy and correctness is known to be very poor in places, but the bulk is near-enough correct.