Over the last few weeks, a few high-profile accounts on Identi.ca, like @methoddan and @jezra, were inexplicably locked out of the system. I'm currently analyzing the possible causes, but I wanted to give a general idea of what's going on.
Here are my current suspected causes:
Incorrect lockout by a community moderator. We have about 40 community members who have the ability to "silence" another user -- preventing them from logging in or posting. These users have shown themselves to be responsible by doing effective spam reporting in the past, and granting them these limited privileges has been a very effective anti-spam measure. It is possible, but not likely, that one or more has either accidentally or mischievously silenced a "good" identi.ca member.
To identify if this is a problem, I'm working on audit tools to figure out by whom, and for what reason, an account was silenced. These should go live soon.
Mass-silencing by script. In late December 2011, we had a concerted and wide-spread attack on Identi.ca that brought down the system for hours at a time. The attack consisted of posts by somewhere between 5000 and 7000 registered accounts, posting randomly-generated text several times a second. The accounts were posting from thousands of different IP addresses.
The only commonality was a similarity in registration period and posted content -- the posts were based on templates like "I just ed with at ." I was able to isolate some unique patterns, and any user who registered in the particular date range and who had posted a notice with a particular text pattern was silenced automatically. About 5000 users were silenced using this method. It's possible that some "innocent" users would be blocked here, but more "senior" users would not fit in the date range.
Automated silencing. During this same period, to catch newly-registered users and users that had not yet posted, I added a posting filter that would look for posting patterns that matched the same text production. The new notices were blocked and the posting account was silenced. These automated silencing tools did not check for user registration date range, however.
The text patterns I used were pretty fine-tuned and, in particular, were clumsy and grammatically incorrect English. But it's possible that they were what triggered the silencing of some users. I think that this is the most likely reason that experienced Identi.ca users were silenced.
I've disabled these filters for now, although I may in the future re-enable them if we have a similar attack.
I personally apologize to the users who've had this problem; we've erred on the side of spam prevention a little too harshly this time around, and it caught people who didn't deserve it. It's not pleasant to be told you're not welcome as part of a community, and since there's no indication of what to do about being banned in our ban messages, it can feel brusque and unfair.
Compounding this was a change in mail server during the same period, which meant that some requests for help weren't delivered. Not helpful, either!
Again, my apologies to Dan and Jezra and anyone else who saw this problem.
