A Distributed Microblogging Proposal
I’ve heard enough about commercial Web services that offer "free" access to users eventually restricting their service in ways that upset the users. The villain of the moment is Twitter (judging by several posts on Hacker News).
I’m not here to lecture — this has been done already, and anything I added would be redundant. Instead, I’ll cut to the chase — the general solution to unsatisfactory centralized control is a distributed system. Here I offer a simple proposal for distributed microblogging.
A URL looks like this: <https://www.metamage.com/u/2012/06/30/160000.html>.
The path begins with the word 'u'
(resembling an asciified mu (meaning 'micro'),
or short for 'update', playing a similar role to 'blog'),
followed by a UTC timestamp consisting of the year, month, day, and time.
I haven’t thought much about the content format yet,
but it should be both displayable as a Web page and easily scraped by a robot.
In the above example,
it would be sufficient to locate the first <p> element
and grab it along with any following elements in the same container.
Merely using the first paragraph would also yield adequate results for some purposes.
There are many issues left to address, such as how to enumerate updates and how to link a reply to its referent, but I’m starting with what I have now. I’ll post updates on my progress, of course. :-)
In the meantime, you can follow me on Twitter (@joshuajuran).