This story is so gloriously beautiful that I want to find the old man and shake his hand. 
Talk about making demons out of daisies: Roy Blunt (R-Missouri), the House Majority Whip, has said that the Democrats, if elected, will "plot to establish a Department of Peace, raise your taxes and minimize penalties for crack dealers". Raising taxes sucks - but two out of three policies sound pretty good, especially when only one out of every three Americans think that the incumbent president is capable of running the country.
Brian Suda has some example microformat XSLT. He's written a O'Reilly Shortcuts on microformats.
Paul Kedrosky has some useful guerilla travel tips.
Projects: one improved and one new 
I'm writing code again! I went to the microformats event this evening in London, and it was really interesting - I didn't liveblog it as I expected, but that's because I was too busy thinking. Anyway, they've recorded a podcast, so it shouldn't be too much of a bother. As soon as I find the podcast, I'll post it. (I won a Yahoo-branded USB stick, which is rather nice - I've been looking to get a USB stick for a while.)
Anyway, I've improved my digg and Grazr mashup. It now lets you click through to see topic-specific user pages (eg. "what Apple stories has kevinrose dugg?") - unfortunately, digg doesn't provide this data in RSS, so it's just a plain HTML link.
The next thing I've been building that I'll hopefully launch soon is something which will hopefully bootstrap people in to putting OPML auto-discovery tags on their pages (since there is some cool stuff that people are doing with OPML autodiscovery) - a Weblogs.Com style service for OPML. I've got the most significant bit done - I've written the code which does all the detection and data gathering and then dumps it all out to the database. I've got a few tweaks I've got to make - but that bit is mostly done. I'm using PHP 5's DOM and SimpleXML functions to gather this information (it's served a very useful purpose in helping me learn how to use both of these) - and I'm using mysqli to write the data out to a SQL database.
It's not massively quick, but it doesn't really need to be - yet anyway. I know how to improve it, if I need to (basically the bit which takes the time is reading in the OPML files - something which I can spin out in to a separate thread if necessary).
Eventually, what you'll be able to do is visit the OPML changes page, submit your change and it'll then go in to the database. I'm also working to implement some form of XML-RPC interface (hopefully weblogsUpdate.ping) and also a POST-REST interface.
The output formats? I'm planning on HTML, a standard XML format, RSS/Atom and OPML. Hopefully, I'll also have a querying interface so you can ask for details of OPML files. The idea is that you'll be able to subscribe to new OPML files (this is also to replace the now dead OPML Pop service which I ran until I screwed up - sorry!).
What I've just written may seem very geeky and dense, so let me say it in a simpler way - this is going to be a simple way of finding new and interesting OPML and sharing your OPML with others.
Cui bono? Well, you, me, Web Standards, the OPML community at large, the OPML tool makers (Grazr, iJot, RSS Labs, Dave et al.).
How can you help make this a success? Simple. Put a link in your document's 'head' to it's OPML. So long as it's marked as being mime type either text or application followed by a slash, then either xml+opml, opml+xml or x-opml.
What else? Well, I'm hoping to figure out how to set the OPML Editor to notify my server when you post a new outline. I'm also figuring out a way so to handle the OPML Editors daily blog outlines. I'm not sure that getting twenty new blog outlines in the "new" feed every day is particularly useful to anybody. It's really a static 'announce' list.
Also, the plan differs slightly from the way that a weblog updates site like weblogs.com works - it is, as I've said, an 'announce' page - so much so that if you ping us with the same file, it gets ignored.
Hopefully, we'll have something that we can play with in the next few days. In the meantime, get those LINK elements fixed on the top of your pages pointing to the outlines that are used in rendering them.
Technorati: opml, autodiscovery, microformats, Grazr, digg
