It appears that repeated (unauthenticated) calls to the API lock out the account.
The workaround is to use authenticated credentials when querying the API. It would be helpful if the API docs could be revised to reflect this. Sal On Dec 13, 8:01 pm, Sal Conigliaro <sco...@gmail.com> wrote: > Thanks Mark. I appreciate it. > > On Dec 13, 1:28 am, Mark McBride <mmcbr...@twitter.com> wrote: > > > I'll check with our abuse team, but this looks odd. > > > On Sat, Dec 12, 2009 at 10:23 PM, Sal Conigliaro <sco...@gmail.com> wrote: > > > Hi there- > > > > I have an app that compares who you're following to your friends > > > followers. To do this, I query > > > ttp://twitter.com/friends/ids.json?user_id=X > > > and compare that to my (saved) list of IDs. > > > > I noticed that if I make repeated (unauthenticated) queries to > > >http://twitter.com/friends/ids.json?user_id=X(ie, I'm comparing my > > > friends to friend A's friends, then to friend A's friend (B), then to > > > friend B's friend (C)) that user_id X gets locked out (I get the > > > "We've temporarily locked your account after too many failed attempts > > > to sign in. Please chillax for a few, then try again." when trying to > > > login to the website (or from a Twitter client). > > > > I'm guessing that the rapid, multiple queries look like abuse. > > > > I did notice, however, then if I make authenticated queries to the > > > same API method, the account locking does *not* happen. > > > > Is this an anti-abuse method? Is my only option to use authenticated > > > calls? > > > > Sal > > > -- > > ---Mark > > >http://twitter.com/mccv
