I managed to get the old behaviour back by modifying the oauth gem to not set a default oauth_callback (oob)
For some reason the twitter-auth gem is not passing over the configuration to override the default I'm too tired to investigate further at the minute but will keep looking in the morning (GMT) On Jun 10, 12:16 am, Matt Sanford <m...@twitter.com> wrote: > Hi there, > > I just checked the tokens generated on several of these services > and I see oauth_callback was set to "oob". Doug is working on the docs > right now to make it clear how all of this shakes out. The end result > is that if you want to use the pre-configured callback url don't send > an oauth_callback parameter at all. If you're seeing this error but > are not sending the oauth_callback parameter please email me off list > with a copy of the URL, headers and body where you make the > request_token call so I can try and debug the issue. It doesn't seem > to be all apps which is what I would expect in the case of a bug. > > Thanks; > – Matt Sanford / @mzsanford > Twitter Dev > > On Jun 9, 2009, at 3:53 PM, lebreeze wrote: > > > > > > > I'm seeing exactly the same behaviour and it just started happening a > > few hours ago > > > App ishttp://moodmapr.com > > > Users just cannot login but instead are provided with a PIN > > > On Jun 9, 11:37 pm, Keith Hanson <seraphimrhaps...@gmail.com> wrote: > >> I'm actually not using an oauth callback parameter and am getting > >> this > >> behavior. > > >> I'm running on Sinatra at the moment, but have implemented my login > >> routine by pretty much copy/pasting the Rails tutorial in the API > >> Wiki. > > >> I'm using the gem OAuth 0.3.5 for redirecting and what-not. > > >> I did take a look at the redirect url, though, and didn't see any > >> oauth_callback params set. It sounds as if they shouldn't be there > >> anyways, correct? > > >> On Jun 9, 5:28 pm, Elliott Kember <elliott.kem...@gmail.com> wrote: > > >>> Sorry - having said that, I've removed the oauth_callback parameter > >>> and the behaviour is still persisting - and it also doesn't save the > >>> authentication so I have to hit Allow every time. > > >>> On Jun 9, 11:21 pm, Elliott Kember <elliott.kem...@gmail.com> wrote: > > >>>> Hey Matt, > > >>>> Yep, I'm passing oauth_callback - and it does look like that's the > >>>> problem, because I have another app which doesn't send it, and it's > >>>> working fine. > > >>>> Is this by design, or will it be changed back? I don't need it to > >>>> use > >>>> the oauth_callback url that I pass, but it'd be good to return to > >>>> the > >>>> specified callback URL by default. > > >>>> Thanks, > >>>> Elliott > > >>>> On Jun 9, 11:17 pm, Matt Sanford <m...@twitter.com> wrote: > > >>>>> Hi there, > > >>>>> Are you by chance passing anything in to the request_token > >>>>> call > >>>>> for the value of oauth_callback? I checked out a few other > >>>>> services > >>>>> and they seemed fine. If you're sending oauth_callback=oob (a.k.a. > >>>>> "out of band") then the system is forced into the PIN flow. We're > >>>>> working on docs for all of this now but let me know if that's what > >>>>> you're sending. > > >>>>> — Matt > > >>>>> On Jun 9, 2009, at 2:55 PM, Keith Hanson wrote: > > >>>>>> Same here, actually, for tweetingtoohard.com (tried to post here > >>>>>> before but it looks like it got eaten :P) > > >>>>>> We've put up a snarky message in the meantime about the > >>>>>> blunders :P > >>>>>> But please do correct us if we have done something incorrect. > > >>>>>> -- Keith Hanson > >>>>>> @big_love > >>>>>> keith (at) tweetingtoohard.com > > >>>>>> On Jun 9, 4:47 pm, Elliott Kember <elliott.kem...@gmail.com> > >>>>>> wrote: > >>>>>>> Hey guys, > > >>>>>>> I'm having a few problems with the OAuth API and my browser- > >>>>>>> based app > >>>>>>> - it's giving me PIN numbers at the /oauth/authorize page, even > >>>>>>> though > >>>>>>> it's set to return to a callback in the OAuth settings - I've > >>>>>>> rechecked my settings, and the application is definitely set > >>>>>>> as a > >>>>>>> browser app. > > >>>>>>> I'm guessing this is something to do with the new PIN-based > >>>>>>> desktop > >>>>>>> app code, just wondering when it'll be fixed - or whether I'm > >>>>>>> doing > >>>>>>> something wrong! > > >>>>>>> Thanks, > >>>>>>> Elliott
