Between 2002 and 2009 I was an active member of livejournal but recently I removed all my content, deleted my account, and then ceased using the service.
Unfortunately I had to un-delete the account after a week, as I suddenly realised that I'd used the URL skx.livejournal.com as an OpenID identifier.
(A few sites, such as stackoverflow, only use OpenID for handling accounts. So I just entered the livejournaly one without really thinking it through.)
Anyway the upshot of this is that I've hacked up a "quick" Perl OpenID handler, and I can now authenticate as www.steve.org.uk.
"Quick"? Well the thing about OpenID is that it is complex. More complex than you'd expect. (Common theme, here?)
I've now coded an OpenID consumer (again pretty livejournal-specific!) and an OpenID server - and both times the lack of documentation made the process annoying, fiddling, and more difficult than it should be:
So why something for me? Well if you're going to use OpenID then the whole premise is that you'll centralise. That means that you'll use the same identity everywhere - suddenly if your provider goes bust, gets bought out, or is compromised all the sites you use with it are vulnerable.
If I'm going to use OpenID I want the confidence that it is under my control. Completely. (Delegation does solve that to a certain extent; but not exclusively.)
Finally: No, Debian-Administration.org will not accept OpenID logins. Ask me to justify that sometime. In a pub. You're buying ...
ObFilm: Lolita
Tags: livejournal, openid 5 comments
