matrix-hacks / matrix-puppet-skype

23 stars 12 forks source link

Error in handleThirdPartyRoomMessage { [M_UNKNOWN: No row found] #11

Closed MurzNN closed 6 years ago

MurzNN commented 7 years ago

I got the error on receive incoming message in Skype room:

Error in handleThirdPartyRoomMessage { [M_UNKNOWN: No row found]
  errcode: 'M_UNKNOWN',
  name: 'M_UNKNOWN',
  message: 'No row found',
  data: { errcode: 'M_UNKNOWN', error: 'No row found' },
  httpStatus: 404 } { roomId: '<redacted>l',
  senderId: '<redacted>',
  senderName: '<redacted>',
  avatarUrl: 'https://avatar.skype.com/v1/avatars/<redacted>?auth_key=<redacted>',
  text: 'Не <redacted> ли случайно? )' } 
MurzNN commented 7 years ago

This was be fist message from new user of room, next messages from this user received successfully. So maybe bridge receive user message earlier that new room member event?

MurzNN commented 7 years ago

Yes, when new user send me a message and no room exits in Matrix, first message is lost, because it received by Matrix puppet bridge before adding remote user to room event. Next messages received normally.

MurzNN commented 7 years ago

Same problem repeating when user is not added to Matrix room, but exits in Skype room. So now Skype room contains 4 users, but in Matrix room there are only 3 users and 1 missed, and all messages from missed user from Skype - generate No row found and missed in Matrix room. Will be good to check if Skype user exists in Matrix room, and force add it if absent.

MurzNN commented 7 years ago

Seems this is problem with specific contact: If this contact try to send me direct messages in new room, I see same error and no messages in Matrix. This contact have - in login, so nick is like asd-qwe, maybe the problem is in - symbol, I try to find other contacts with same symbol.

MurzNN commented 7 years ago

Seems the source of the problem is in #15 issue.

twouters commented 6 years ago

Should be fixed when matrix-puppet-bridge 1.15.4 is released.

MurzNN commented 6 years ago

Thanks, seems fixed - at now after upgrading and 4 days of testing - I don't see repeats of this error.