There is some python code available to interface with the geonames web services, so that fuels my hopes we can also manage to link up in a more "web 2.0" way with the data services over there. It's about time we have some more mashing up than just our Google maps. If anyone is already working on a Django interface for this, I'd love to hear from you!

In other news we're working on a second instance of the platform, this one to support work Greenpeace will be doing in the middle east in the wake of the Lebanon crisis. You can read more about that project at it's holding site at www.speakup-middleeast.org . Hopefully that site will be up and running by the end of this week.

We've also been working hard to import our geo-data into the system. The GNS data we've been working with varies widely in quality. Some countries (eg Australia) can be pulled seamlessly out of the database, others (eg. Israel) suffer from duplicate or missing administration districts which has necessitated some manual fixes. The GNS data provides about ten times more data than the initial Geo-Rosetta data we deployed, so it's worth the hassle to get it in there.

Still, nearly there...

In the meantime, our colleagues at hosting have provided us a virtual machine to configure as the host for our platform. We are thinking about making such a virtual machine available to developers as well. It would make live so much easier: download the image, run it, and you're ready to help in developing our platform. No other downloads or setups needed, eveything including the automated testing ready for you to use.

Also, the data sets for our geo-location data are being merged. We have some 108,000 places in there, and about to add another 150,000 or so for the US, with a third collection still waiting. Getting everything together (longitude, lattitude, duplicate names, missing links to country or region, different spellings or transcriptions, there is some work involved before we get things on a map). Another interesting week ahead.

In the meantime, it has been a bit quiet here on the blog, while work was progressing. The naming and design are moving forward, and we should have final versions and be implementing the design in our templates soon. And... we now have the Melt project space at Sourceforge (thanks for your help, Misha!), so our software will become available through the new Sourceforge Subversion system.

In the meantime, we're coming close to the "show and tell" phase of our platform. The last "blocking issues" for a first live release are being solved this week, and we're making progress with a design overhaul. OF course, that leaves us with only a million or so new ideas to develop on the platform, so we're also moving the open source side of the platform to a new platform to open it up for more people to contribute.

Keep an eye on this blog, or get in touch with me if you can help us! (rolf.kleef (at) int.greenpeace.org)

Tuesday 30th May, 11AM PST (which is 8PM Western Europe, 7PM UK, 6PM UTC) I will have an hour session on our Custard Melt project, and especially on what our ambitions and expectations are. I hope that many people will join to explore how we can make the online platform help in galvanising offline action.

At 4PM PST, 1AM Western Europe, midnight UK, 11PM UTC, Micki Krimmel of Participant Productions will be online for a session on "Media that Mobilizes: An Inconvenient Truth, ClimateCrisis and more tales from Participate.net". Of course, as the producers behind Al Gore's movie about climate change, this is of particular interest to us.

Secondly I split out the non-functional stories onto a separate tab, since I've never felt quite comfortable having things like 'Must use standards compliant xHTML' sit next to 'Should be able to order search results alphabetically', it just didn't seem like we were measuring the same things.

Finally I've added the bugs we've found as stories in cases where they can't be listed directly against an existing story. I'm not sure how that's going to pan out but we'll see as we go...

You can see the new spreadsheet at http://svn.greenpeace.org/repositories/custard/doc/Custard%20Master%20Story%20List.xls?rev=592

As a more general update we're planning to get back to work on Melt this month, with a target of getting the first public site up next month. During our recent downtime we've been gratified to see Google adding more data to their Google maps, so that the system makes much more sense now when you're looking at results in Europe.

If anyone can recommend a good moveable type compatible service for captcha tests that would be great - seems preferable to asking people to get a username / password.

While I don't want to sound flippant about seriousness of this he may have a point. After all, the Great Bear Rain Forest used to be called the North East Timber Supply. Changing it's name was a key part of saving it.