- It’s read-only. It solves the easier technical problem.
- I’m not really a programmer. I was a programmer almost 15 years ago. But even then it was MUMPS, Python, XSLT, ColdFusion, other things.
- Asynchrony confuses the heck out of me. If you look at it and believe I’m misunderstanding how to code asynchronous JSON let’s have a hangout. I could learn.
- I had a version of this I wrote about a month or two back that I demo’d to some people, but there was an important difference — that version read the federated wiki format but couldn’t resolve links in the federation. This one resolves links by looking through the fork history of the page and querying those sites about whether they might have a copy of the linked page.
The weirdest thing about this code is that it’s so simple in a way. Once you carry the fork history in the page, the page can travel around wherever it wants, and links can be resolved without hard-coding brittle and rigid URLs in the wiki markup. Once you use JSON as the basis of pages instead of HTML, multiple sites can work as one giant site. Taken together these two simple ideas (along with the legal technology of Creative Commons) create a radically different vision of the web. I guess what I’m saying (and what I’m hoping you see in the code) is it’s a lot simpler than you might think to get to the vision of a World Wide Wiki. If a hack like me can get this done in a couple weekends, what could you do? Video on how the code works forthcoming. But download it now and start playing with it.
Update #1: Coding the item handler loop
Update #2: Resolving links from fork history