Gajim - 2018-09-25


  1. Maranda <!-- Incoming 25/09/2018 02:05:48 --> <iq xmlns="jabber:client" id="juliet1" type="result" to="imadmin@lightwitch.org/gajim.QMD3WOH0" from="grimoire@muc.metronome.im"> <fin complete="true" xmlns="urn:xmpp:mam:2"> <set xmlns="http://jabber.org/protocol/rsm"> <first index="110">308c96c8-d8bc-48ff-a091-2380b1fd4008</first> <last>dc8424f2-69ce-44f4-ae5f-4b0a20e1e4b3</last> <count>161</count> </set> </fin> </iq>
  2. Maranda lovetox, look at index on first element
  3. stp > I'd be in favor of keep the same icon, but better design. The current icon is the identity of Gajim. A nicre version of the icon could be nice indeed +1
  4. Kirk Hello, I'm running gajim 1.0.1 on ubuntu 18.04. I also have the "Message Box Size" plugin enabled. When I make a large paste into a MUC, it crashes my login session and I need to log in again. Is this a known issue with gajim? If not, where should I report such an issue?
  5. lovetox Kirk, does it also happen if you disable the plugin?
  6. lovetox Maranda, there is only a index if you actually get a message
  7. lovetox but anyway, doesnt matter, its ok what metronom does
  8. bot Matthew Bell created an issue in _gajim_ < https://dev.gajim.org/gajim/gajim/issues/9357 >: #9357: < DBus org.freedesktop.login1 sleep handling >
  9. Maranda lovetox: obviously on an empty page you get nothing but the count (that's supposed to be that way in the spec)
  10. Maranda I previously thrown an item-not-found if you overflown archive with RSM but consensus was you should throw an empty page instead.
  11. Maranda (item-not-found is thrown only if the entry specified in either before or after doesn't exist)
  12. lovetox yes i wondered about that, i saw servers throw item-not-found
  13. lovetox i think its also much cleaner to send a finished iq
  14. Maranda (or specified <index /> doesnt exist forgot about that)
  15. Maranda lovetox: afair Metronome was the only one doing that btw 😆
  16. mrDoctorWho I installed PGP plugin. What's next? I don't see any option to enable or setup the PGP encryption
  17. vanitasvitae mrDoctorWho: iirc the (new) pgp plugin is still very experimental and lacks a UI
  18. mrDoctorWho I think you're talking about OpenPGP plugin that implements XEP-0373
  19. mrDoctorWho There are two PGP plugins out there
  20. vanitasvitae The 373 one is the new experimental one
  21. savant Hi guys! I'm using V1.0.3 of Gajim on Arch Linux and I would like to disable the chat history. However, I haven't found an option yet. Is it any possible?
  22. nico savant, there is a no_log_for config option where you can add your account name to disable logging for the entire account . at least the description states to function that way.
  23. lovetox you can set it in the Account window for the whole account
  24. savant https://upload.mailbox.org:5281/upload/ZqCxuxMbIh3L_rhf/Screenshot_20180925_160513.png
  25. lovetox or you can set it in the history window of each contact
  26. lovetox no savant, dont modify this value
  27. savant I think I need some help. What do i need to change?
  28. lovetox for what do you want to disable the chat history?
  29. lovetox for one contact or everyone?
  30. savant lovetox: I want to disable the entrie history function
  31. savant Like when u turn off the history in Firefox.
  32. lovetox Then Gajim -> Accounts
  33. lovetox then i think in the general options
  34. savant I think they remove the option. The checkbutton is grey.
  35. lovetox what?
  36. savant https://upload.mailbox.org:5281/upload/fAeTid89repL_ifA/Screenshot_20180925_160813.png
  37. savant Version 1.0.3
  38. lovetox please read what i write
  39. lovetox Then Gajim -> Accounts then i think in the general options
  40. savant I thin I found the option.
  41. savant https://upload.mailbox.org:5281/upload/U6p5W-QzrLMzsUdg/Screenshot_20180925_160940.png
  42. savant 1 sec
  43. savant Works! Thank you. :)
  44. bot Daniel Brötzmann modified an issue in _gajim_ < https://dev.gajim.org/gajim/gajim/issues/9357 >: #9357: < DBus org.freedesktop.login1 sleep handling >
  45. bot Marcin Mielniczuk proposed a new merge request for _gajim/master_ < https://dev.gajim.org/gajim/gajim/merge_requests/330 >: WIP: Refactor generating the groupchat nick hits into a separate function.
  46. marmistrz lovetox, do we have any tests for groupchat_control.py?
  47. Maranda lovetox is Gajim trying to send messages before joining?
  48. Maranda [17:22:02] ‎error: not-acceptable ‎[17:22:03] ‎error: not-acceptable
  49. Maranda is displayed in MUCs
  50. Maranda on latest master
  51. lovetox yes typing notifications
  52. lovetox no marmistrz
  53. bot Marcin Mielniczuk updated a merge request for _gajim/master_ < https://dev.gajim.org/gajim/gajim/merge_requests/330 >: WIP: Refactor generating the groupchat nick hits into a separate function.
  54. bot Marcin Mielniczuk updated a merge request for _gajim/master_ < https://dev.gajim.org/gajim/gajim/merge_requests/330 >: WIP: Refactor generating the groupchat nick hits into a separate function.
  55. marmistrz lovetox, so I'll add no mine either, GroupchatControl is too complicated to instantiate
  56. bot Marcin Mielniczuk updated a merge request for _gajim/master_ < https://dev.gajim.org/gajim/gajim/merge_requests/330 >: Refactor generating the groupchat nick hits into a separate function.
  57. lovetox yes its ok you dont need to write tests, but if then i would test only your methods with sample data
  58. lovetox its only about choosing the right next nick
  59. lovetox out of a list
  60. lovetox you dont need the chatcontrol
  61. marmistrz hmmm, I could just mock things
  62. bot Marcin Mielniczuk updated a merge request for _gajim/master_ < https://dev.gajim.org/gajim/gajim/merge_requests/330 >: WIP: Refactor generating the groupchat nick hits into a separate function.
  63. bot Marcin Mielniczuk updated a merge request for _gajim/master_ < https://dev.gajim.org/gajim/gajim/merge_requests/330 >: WIP: Refactor generating the groupchat nick hits into a separate function.
  64. lovetox what do you need to mock?
  65. lovetox your methods only operate on a list, self.attention_list
  66. lovetox thats it
  67. lovetox ah i get it you mean you need the chatcontrol object
  68. lovetox because the methods are in there
  69. lovetox yes could be tricky thats why i meant you make a class that takes care of choosing, not a method on the chatcontrol
  70. lovetox this class will surley be expanded later with more logic
  71. lovetox then just lower()
  72. lovetox and is not blocked jid
  73. Maranda lovetox, if eventually you wanna play with SCRAM...
  74. Maranda https://conference.gajim.org:5281/pastebin/cd179f64-2dff-4968-9b36-c45b874b48fa
  75. lovetox great Maranda !
  76. lovetox i dont know shit about SCRAM though
  77. bot Marcin Mielniczuk updated a merge request for _gajim/master_ < https://dev.gajim.org/gajim/gajim/merge_requests/330 >: WIP: Refactor generating the groupchat nick hits into a separate function.
  78. bot Marcin Mielniczuk updated a merge request for _gajim/master_ < https://dev.gajim.org/gajim/gajim/merge_requests/330 >: WIP: Refactor generating the groupchat nick hits into a separate function.
  79. marmistrz https://conference.gajim.org:5281/pastebin/5494ee77-44a0-48ed-8060-1f5e9ef77ce5
  80. marmistrz lovetox, this happens while `python setup.py test_nogui`
  81. marmistrz Have you ever come across something like this?
  82. lovetox its a import cycle
  83. lovetox you see in the above lines that you import gajim/config.py
  84. lovetox we dont test gui code, thats why the test its called test_nogui
  85. lovetox everything in this traceback is importing gui code
  86. lovetox try to first import gajim.gtk
  87. concerto Any progress with the smarter completion?
  88. concerto I just typed S<tab>, intending to respond to SamWhited, and it completed to SaltyBones :D
  89. Daniel concerto, look closely, there is even a MR now for this
  90. Maranda lovetox, well adding support for scram-sha256 shouldn't be hard me thinks anyways
  91. Maranda I dunno how it's implemented in gajim but you should just need to change the digesting algorithm
  92. lovetox so from SCRAM-SHA-1
  93. lovetox to 256 is just changing digest?
  94. lovetox we can certainly try that
  95. lovetox why does the PLUS exist?
  96. Zash channel binding
  97. lovetox is this better?
  98. Zash mixes in some stuff from TLS that should make the auth fail in case someone MITMs you
  99. Maranda lovetox: gajim does channel binding already
  100. Maranda Btw if you weren't aware
  101. lovetox i am aware
  102. lovetox and it is nbxmpp
  103. Maranda 🤸‍♂️🙂
  104. lovetox wait i adapt nbxmpp then we can try
  105. Maranda lovetox, oh make sure to change password on lw.org first btw
  106. Maranda lovetox, else you won't have the sha256 keys in the account store to use scram-sha256
  107. lovetox hm
  108. lovetox tied get a malformed request error
  109. lovetox Maranda, do you see whats wrong serverside?
  110. Maranda retry now
  111. Maranda hm
  112. Maranda no text
  113. Maranda I think I know
  114. lovetox malfromed seems to be the wrong error in any case
  115. Maranda https://github.com/maranda/metronome/blob/master/util/sasl/scram.lua#L143
  116. Maranda either there's no gs2_cbind_flag
  117. lovetox hm could be, do i have to use it?
  118. lovetox i thought its only for PLUS
  119. Maranda or you sent y
  120. Maranda and server supports channel binding so you should use PLUS (that's the stuff to protect from MITM)
  121. lovetox yes could be
  122. lovetox mom
  123. lovetox seems to work
  124. Maranda :)
  125. lovetox still weird that it doesnt work without channel binding
  126. lovetox i think i send indeed y
  127. lovetox im not sure how to find out if channel binding is supported
  128. lovetox should the ssl lib tell me that?
  129. lovetox ah i have it, weird that it doesnt work though
  130. lovetox Maranda, yes we send y,,
  131. lovetox but why is this wrong?
  132. Maranda because if you send y, and don't use plus, it means you don't think the server supports channel binding and there's a downgrade MITM attack going on.
  133. Maranda at least that should be how it works if I read the specs right
  134. Maranda lovetox, if you don't wanna use cb just set the flag to "n"
  135. Maranda lovetox, the server must refuse authentication otherwise.
  136. lovetox yeah i read it
  137. lovetox client has to find out via advertised mechs
  138. lovetox so if i find any PLUS in the mechs, i send automatically p
  139. lovetox and if i find no PLUS i send n
  140. lovetox but this seems weird, because it means the client does not support cb
  141. lovetox when in fact the server is not support it
  142. lovetox lol wait
  143. lovetox Clients that support mechanism negotiation and channel binding MUST use a "p" gs2-cbind-flag when the server offers the PLUS- variant of the desired GS2 mechanism.
  144. lovetox but then
  145. lovetox If the client supports channel binding and the server does not appear to (i.e., the client did not see the -PLUS name advertised by the server), then the client MUST NOT use an "n" gs2-cbind- flag.
  146. lovetox but they dont say what the client should use instead
  147. lovetox Maranda,
  148. Maranda lovetox, hm No not what I meant
  149. lovetox the question to me is now, if i see a server without PLUS
  150. lovetox what should i send
  151. Maranda lovetox, I said that if the server advertises plus and you don't want to use plus you can't send y
  152. Maranda lovetox, in that case you can send y
  153. lovetox why what does y mean
  154. Maranda That you support channel binding :P
  155. lovetox ok server doesnt support channel binding = y
  156. lovetox but what is if the server supports channelbinding but i dont want to use it
  157. Maranda send n :P
  158. lovetox or is this stupid
  159. Maranda I told you
  160. lovetox kk
  161. Maranda I'm not sure if it's stupid or not, but that's how you should do if you don't want a server supporting channel binding to drop you thinking you have been MITM'ed :P
  162. lovetox then the last question to me is
  163. lovetox how should i choose/sort
  164. lovetox is 512 better then 256-PLUS
  165. Maranda lovetox, disregard 384 and 512 (for now)
  166. lovetox ok
  167. Maranda lovetox, they apparently aren't officially approved by the IETF I got tricked into pulling 'em in because googling I actually seen 'em being used
  168. Maranda and also the IANA register listed SCRAM*
  169. lovetox although they write to determine from the mechs if cb is supported
  170. lovetox ssl offers a method for that
  171. concerto Daniel: thanks!