Gajim - 2016-01-04


  1. kalkin How do i register with trac-plugins? I can not find any registration link
  2. kalkin Asterix plz ping me when you are there and i will send you my ssh public key
  3. kalkin Asterix: Or if I'm absent when you are here, just add my jid to you roster (I assume you see the jids in this muc) and i will send you the ssh key later.
  4. Link Mauve kalkin, why don’t you just put it here?
  5. Link Mauve He’ll see it in the logs.
  6. kalkin Link Mauve: Because this log is public. I do not want that everyone knows my ssh fingerprint.
  7. Link Mauve What kind of attack would that mitigate?
  8. kalkin Link Mauve: no attack at all, but it's an identifying information. I'm not sure if you could spin this some how. I must admit the only attack would be do MiM to figure out who is trying to connect to specific host over tor. I would notice the MiM but this should be enough for THE GOVERMENT[TM] to identify me
  9. kalkin Not that I'm using it over tor, but you have to be very careful what data about you is available.
  10. kalkin If Asterix can not reach me today/tomorrow i will just send him an E-Mail :)
  11. Link Mauve Better just do that, or by XMPP, no need to wait for him to be available then.
  12. Flow kalkin: pm to asterix a link to your pub key?
  13. kalkin I just send an email
  14. Flow oh, he's not online
  15. Flow kalkin: how's work on the omemo plugin going?
  16. kalkin Flow: Well I'm back in the office. I'm going through my github issues currently
  17. kalkin Will start this week fixing bugs and adding enhancements
  18. Flow so you are able to work on open source projects in your paid work time? that's great :)
  19. kalkin Flow: I wish. My office is in my apartment. I just do not like to use a mobile device for development. You should never underestimated productivity habits. :D
  20. Flow hehe ;)
  21. Flow kalkin: do you still develop palaver? Or is it put on hold/neglected now?
  22. kalkin Flow: I'm freelance software developer and I develop gajim-omemo in my spare time
  23. kalkin Flow: Put on hold/neglected, because it's a full time job for around 12-16 months and I have to earn a living
  24. kalkin I remember there was a way to disable E2E sessions in gajim for good. Any one remembers how it worked?
  25. arune kalkin: a setting in ACE...
  26. kalkin could it be autonegotiate_essesions?
  27. kalkin has currently no one online who uses gajim, so he can not verify it
  28. arune kalkin: I think so
  29. arune It was mentioned in this chat just in the past days
  30. kalkin arune: lol. Just found out that i told 0xAFFE on 14.12.15 how to disable it :D
  31. kalkin Just for the history enable_esession can be set to disable to disable to completly disable essions. autonegotiate_esessions can be set to disable to disable autnegotiation.
  32. Darlan Hello, tmolitor.
  33. tmolitor hello darlan :)
  34. Darlan I was asleep two days ago. I am sorry about it.
  35. Darlan Would you like to make more tests?
  36. tmolitor darlan: no problem :)
  37. tmolitor and yes...
  38. Darlan When?
  39. tmolitor now ;)
  40. Darlan So I will restart Gajim and start logging. I will be back, in a few minutes.
  41. kalkin Darlan: we almost got worried you got killed by some crazy developer :D
  42. tmolitor darlan: you don't need to log it....I have my own gajim logging here with more debug messages...I only need you to trigger the bug at my side...
  43. tmolitor darlan: why is this bug triggered much more often for you than for me? :o
  44. tmolitor darlan: is the bug also triggered when I write you something using my conversations client at my phone?
  45. Darlan I do not know =/
  46. Darlan Yes
  47. Darlan I think it is a malformed file on my end.
  48. Darlan That is causing this.
  49. Darlan How did you know it was triggered?
  50. tmolitor darlan: because you went offline
  51. tmolitor pn me again please...
  52. Darlan > we almost got worried you got killed by some crazy developer :D kalkin, HAHA xD
  53. tmolitor darlan: it triggered when you came back online...
  54. tmolitor but I have to tune my debug output further...
  55. Darlan I think I know why you send messages in MUC instead of PM :P
  56. Darlan Getting this bug triggered often is annoying!
  57. tmolitor darlan: yes :D
  58. tmolitor I'm restarting my gajim...just wait a few seconds...
  59. Darlan Yes, sir!
  60. Darlan Yes, Sir!
  61. tmolitor darlan: please pm me again...
  62. Darlan Do you want me to send to you a PM?
  63. tmolitor yes
  64. tmolitor I thinkI got a good trace! lets analyze it :))
  65. stp Does anyone know, how to trust the OMEMO fingerprint from your own Conversations? A friend got everything working, but his own messages sent from Conversations don't appear in his Gajim.
  66. kalkin You can click on your avatar. This will show up your Account Details. There you can trust your Device Keys.
  67. kalkin stp: iNPUTmice mentioned today that he had this issue fixed, it is not released yet.
  68. kalkin assumes it will be released with next Conversations version
  69. Darlan daniel, kalkin, do you know of somebody that is working on bringing OMEMO plugin/support for Psi?
  70. kalkin Darlan: There is an open issue (in russian) on Github for Psi+. Assuming that Psi+ is just a patchset, which not even has carbons, mam & http-upload (there are issues open for it in russian), I have no hope it getting OMEMO any time soon
  71. tmolitor darlan: I think I am very close to this bug now...
  72. tmolitor from the source:
  73. tmolitor class LegacyContactsAPI: """ This is a GOD class for accessing contact and groupchat information. The API has several flaws: * it mixes concerns because it deals with contacts, groupchats, groupchat contacts and metacontacts * some methods like get_contact() may return None. This leads to a lot of duplication all over Gajim because it is not sure if we receive a proper contact or just None. It is a long way to cleanup this API. Therefore just stick with it and use it as before. We will try to figure out a migration path. """
  74. Darlan > I have no hope it getting OMEMO any time soon For what reason, do you think so?
  75. kalkin Darlan: general impression on the development process.
  76. tmolitor and exactly that is happening (get_contact() returns none)...
  77. kalkin Also same is true for original Psi. It's almost dead even if there are minor bug fixes every month or two
  78. kalkin had evaluated different clients for XMPP some time ago
  79. Darlan You mean, years.
  80. Darlan Psi is not dead, you may look at the official git repository. Tough, I do not know what is the process of decision making there for releasing.
  81. Darlan tmolitor, what exactly _is_ happening?
  82. kalkin Darlan: I'm https://github.com/psi-im/psi/commits/master and https://github.com/psi-im/psi/commits/master and https://github.com/psi-im/psi/commits/master
  83. kalkin s/I'm/I am looking at/
  84. tmolitor darlan: File "/usr/share/gajim/src/session.py", line 165, in _nec_decrypted_message_received self.control.contact = contact and contact is None in this case because of the comment in the code I posted a few minutes ago...
  85. Darlan You have pasted 3 identical links.
  86. kalkin Darlan sorry
  87. Darlan tmolitor, do you know how to fix this?
  88. Darlan No problems, I wanted to know if there were two missing links that maybe you wanted to paste.
  89. kalkin Darlan https://github.com/psi-plus/psi-plus-snapshots/commits/master https://github.com/psi-plus/main/commits/master
  90. Darlan I think, we have a lot to expect from Psi; I hope, soon.
  91. kalkin But i do not want to stop anyone from implementing OMEMO and sending patches to any Psi client. I'm even pretty sure the Psi+ guys would accept them if they are reasonable.
  92. tmolitor darlan: I need some more tests....stand by :)
  93. Darlan tmolitor, fixing this bug would make Gajim usable for me, once again.
  94. Darlan Imagine how it feels to have these pop-up loops bugs with Gajim on a 32-bit single core laptop (1.6GHz CPU and 1GB RAM) while having a web browser running.
  95. Darlan That laptop still exists. I think of turning it into a wireless repeater.
  96. tmolitor darlan: please pm me again...
  97. tmolitor test...
  98. tmolitor darlan: pm me again ;)
  99. tmolitor darlan: sorry...reconnect was not working...
  100. tmolitor next pm please :)
  101. tmolitor darlan: ping
  102. Darlan tmolitor, pong
  103. Darlan BOOM!
  104. tmolitor darlan: next pm please...
  105. tmolitor darlan: complete disk crash?
  106. Darlan I can't. POPUPS!
  107. tmolitor darlan: oh :D
  108. Darlan I hate it.
  109. tmolitor darlan: if you start gajim in the console with logging activated you won't get popups as the messages would be visible in the console :)
  110. Darlan I know, but, apparently, this is not the case with these loop errors.
  111. Darlan I am using this: $ gajim -l DEBUG 2>$(date +%Y-%m-%d_%H:%M:%S)_gajim_debug.log Logger gajim level set to 10
  112. tmolitor darlan: triggered now...
  113. tmolitor darlan: investigating...
  114. Darlan Stable, here.
  115. tmolitor I think I have a patch ready in about 5 minutes :)
  116. tmolitor darlan: do you know in which directory the gajim source is at your laptop?
  117. tmolitor I'll give you some line of code you have to insert and then we'll test again if the bug is still there (I hope not :) )
  118. Darlan Yes, at /usr/share/gajim/src/
  119. Darlan I am using a station computer.
  120. Darlan Beware that my HDD is faulting...
  121. tmolitor darlan: I know...
  122. tmolitor darlan: okay...we're changing the file /usr/share/gajim/src/session.py
  123. tmolitor https://xmpp-upload.datenknoten.me/-br3Vey7Wu4NBfiBX37qzQ3WxpA/hCsC7gfEQOpkx/session.py
  124. tmolitor darlan: yust copy the file over the old one and restart gajim :)
  125. Darlan I figured you would send me a patch, just like Dicson used to do.
  126. Darlan Just for confirmation. $ sha1sum session.py 2c347429f4891e1b14fa319c4668e38904df7ef6 session.py
  127. tmolitor darlan: yes...that's the file :)
  128. Darlan # rm session.pyc session.pyo
  129. tmolitor darlan: say if you are ready for the next test round :)
  130. Darlan I have seen an empty chatview box for quite some time.
  131. Darlan Patch, for record: --- /usr/share/gajim/src/session.py.original +++ /usr/share/gajim/src/session.py @@ -38,6 +38,9 @@ import dialogs import negotiation +import logging +log = logging.getLogger('gajim.interface') + class ChatControlSession(stanza_session.EncryptedStanzaSession): def __init__(self, conn, jid, thread_id, type_='chat'): stanza_session.EncryptedStanzaSession.__init__(self, conn, jid, thread_id, @@ -67,6 +70,12 @@ return contact = gajim.contacts.get_contact(self.conn.name, obj.jid, obj.resource) + #check for None and use the already saved contact then (if available) + if contact is None: + if self.control: + contact = self.control.contact + log.error("_nec_decrypted_message_received(): gajim.contacts.get_contact(self.conn.name [%s], obj.jid [%s], obj.resource [%s])..." % (self.conn.name, obj.jid, obj.resource)) + log.error("contact is now: '%s'" % contact) if self.resource != obj.resource: self.resource = obj.resource if self.control: @@ -162,7 +171,9 @@ if ctrl: self.control = ctrl self.control.set_session(self) - self.control.contact = contact + log.error("*** WOULD ALTER CONTACT TO: '%s'" % contact) + if contact: + self.control.contact = contact if not pm: self.roster_message2(obj)
  132. Darlan Hold on, please.
  133. tmolitor darlan: okay...
  134. Darlan No delay, now. Maybe Gajim created session.pyc and session.pyo at that time.
  135. Darlan tmolitor, ready.
  136. tmolitor darlan: ready :)
  137. Darlan Yes.
  138. tmolitor darlan: very nice...its working!! :)
  139. Darlan I hope it is.
  140. tmolitor darlan: it is...I see the bug would have triggered at my side numerous times...I added this debug output: "*** WOULD ALTER CONTACT TO: 'None'" to catch this :)
  141. mokasin Hey. Is it possible to debug and profile Gajim plugins in vitro?
  142. Darlan mokasin, please post this question at the gajim-devel mailing-list.
  143. mokasin Darlan, alright
  144. Darlan https://lists.gajim.org/cgi-bin/listinfo/gajim-devel
  145. kalkin Where does Gajim expects the plugins to be stored at Windows?
  146. Darlan Maybe in %APPDATA% or Program Files.
  147. Darlan Please ask mmlosh to attend in this channel more frequently; I would like to test the new fix of tmolitor against him.
  148. Darlan tmolitor, can you detect what malformed settings (or, maybe, strings) cause to that loop error?
  149. 0xAFFE kalkin, %APPDATA%\Roaming\Gajim\Plugins
  150. 0xAFFE if you install the plugin via the plugin manager
  151. kalkin 0xAFFE, Darlan thanks
  152. 0xAFFE %APPDATA% is "%USERPROFILE%\Application Data" IIRC
  153. 0xAFFE oh no, its %USERPROFILE%\AppData and %UserProfile% is in most cases C:\Users\%USERNAME% but can be different
  154. 0xAFFE so for the user tim this would be C:\Users\tim\AppData\Roaming\Gajim\Plugins
  155. Darlan LAST TEST OF A RECENT FIX: @Everyone, please send to me a PM or two and initiate E2E encryption with me.
  156. Darlan has received zero (0) private messages.
  157. tmolitor darlan: what do you mean?
  158. Darlan Are you able to detect what file or setting caused to the error message?
  159. tmolitor yes...the line I fixed in your session.py file ;)
  160. Darlan So would it be correct to assume that this bug has occurred to many other users?
  161. bot RSS: Feeds for Gajim • Ticket #8215 (Programming error on attempt to open nickname dialog twice) created Bug description Traceback (most recent call last): File "/usr/share/gajim/src/groupchat_control.py", line 2219, in _on_change_nick_menuitem_activa[…] https://trac.gajim.org/ticket/8215 • session.py attached to Ticket #[…] https://trac.gajim.org/attachment/ticket/8214/session.py • session.py.patch attached[…] https://trac.gajim.org/attachment/ticket/8214/session.py.patch
  162. Darlan tmolitor, please confirm the validity of the attached session.py files in ticket #8214.
  163. Darlan #8214
  164. bot Darlan: http://trac.gajim.org/ticket/8214 (Endless error messages)
  165. bot RSS: Feeds for Gajim • Ticket #8214 (Endless error messages) updated The patch is not the right one (it still contains nummerous debug outputs) I'll create a clean one in the next couple of days…stay tuned :) https://trac.gajim.org/ticket/8214#comment:2