Phillip Pearson - web + electronics notes

Dave Bryson is working on a 'Java Community Server'. Interesting! The last attempt to do the same kind of petered out a while back. Good luck to Dave; it sounds like he's going well!

Of course, anyone implementing this sort of thing is welcome to steal code from my Python Community Server and phpStorageSystem apps.

Non-Internet Explorer-using folks like Scott Rosenberg will be pleased to hear that I fixed a silly bug with my comment monitor, and now it works in both Opera and Mozilla. Enjoy!

bzero v0.04: now with RSS 0.94 support.

Sam Ruby suggests a modification to RSS 0.94: use Dublin Core for the things it does well.

This brings the 0.9x branch a bit closer to the 1.0 (RDF) branch, but without sacficing compatibility with aggregators that only handle <rss/> documents.

Second p0st, leading authority on hacking blogger blogs.

My experiment with Thinlet. Turns out it's a Googlewhack.

(Found while searching for Mike Cannon-Brookes's ThinRSS.)

Has anyone else been doing stuff with Thinlet? There was a brief bit of buzz about it around the time ThinRSS appeared, but I haven't heard anything since.

Update: The word is that the next version will be LGPL. Cool. That means I'll be able to hack in the things I want, like a text editor that will wrap properly. Perhaps this will be the way to make a minimal (just bold/italic/links) cross-platform HTML editor.

Found out about reStructured Text from Zia this morning. Interesting. Another wiki-like text formatting system. I wonder if I should put something like that into bzero. I need to implement shortcuts; typing out the whole link every time I like to bzero or PyCS is a bit tedious.

Working on getting bzero to generate RSS 0.94. Is Second p0st's feed working?

RSS 0.94 has some nice new things. Posts can now have a 'guid' attribute, which (in a roundabout way) lets you tell people about the permalink for the post. For the moment I'll make this the same as the 'link' attribute. I realised I wasn't generating 'title' or 'link' attributes at all; that's fixed now. Oops!

There's also a place to note where to go to get the comments for the post. I've been putting that in as a little link marked 'Comments' at the bottom of each article, but now it's machine-readable as well.

In short: I like it. Not quite perfect, but a nice incremental improvement.

One suggestion: clarify 'link' and 'guid'. How about adding 'permalink' and making it ALWAYS contain the permalink, then making 'guid' optional? That way, when someone puts together some weird mirroring system that has a non-URL GUID, it will fit in nicely. Although maybe we don't need non-URL GUIDs anyway ... hmm. Something to think about.