The future of blogging, or how it should be, is what Matthew tries to describe. Many good points in there. I particularly like his focus on the ability to have all sorts of ways of sorting your posts and organising them, including reasonable URLs. Currently I do feel a bit locked into Movable Type's weekly scheme, which I decided for in the beginning. I like the way Blosxom seems to do things, but that would break all my old links. If I were ambitious enough, there'd probably be ways to redirect things (or aren't there for bla.html#entry links?) appropriately, but I'm not into scripting).
Matthew writes about Open Source, databases and future proofing things on the time scale of decades. While I appreciate Matthew's writing, I asked myself "What does he know?" (or I?, for that matter). How long has software development been around? in the way it's done today? How long for open source? open source databases? blogging systems? what's the track record of people coming up with software standing the test of time? For how long have you used computers? the same application? the same data? the same file format for that data?
My answers to these questions were, in that order: little, even less, a few decades, even fewer decades (perhaps only one), a decade or two (if you're generous), a couple of years (?), even fewer years, very bad, 14 years iirc, 11 years and a bit (TeX, though you could argue that TeX 3 with LaTeX2e is different that TeX 2.97 with LaTeX 2.09), 11 years again, 11 years again (although other of my old TeX documents may need a little tweaking to run these days: some style files don't exist anymore, others have Atari graphics included, which needed conversion and different inclusion techniques).
All of that makes it seem pretty unlikely that a blogging system will make it for that long. But Matthew got a few essential points there that'll make transition painless. And that's the most important bit, in my opinion. Programs, scripting languages, databases or operating systems may come and go – but that's not a big deal as long as there's an easy passage for your data.
And that's the reason why my old TeX files still reside on my hard drive today and my Signum files don't. The TeX files either just worked or needed only trivial conversions. And while I can see Matthew's point about databases, the sheer triviality of Blosxom's text-file based approach makes it very attractive. Your data will be easy to move – and we can let Moore's Law take care of the speed problems we may run into when managing decades' worth of content.
And for today, a KISS approach is appealing as well. I don't want to spend too much money on running this whole internet thing. And little money doesn't buy you all the flashy databases you may want to have. The lower the requirements of a system, the more people will be able to run it.
I am aware that Matthew only mentioned support for a database rather than a database-only system, but it was a good place to appeal to simplicity. Being the pessimistic person I am, I don't think there'll be a 'for-life' blogging system.
Received data seems to be invalid. The wanted file does probably not exist or the guys at last.fm changed something.