Gajim - 2018-09-09

  1. concerto
  2. lovetox everytime?
  3. concerto Can't say just yet, this is just the first time I tried it :)
  4. lovetox hm works for me
  5. lovetox i give the bridge the fault
  6. bot Sophie Herold updated a merge request for _gajim/master_ < >: WIP: Rework Contact Info Dialog
  7. Orbit appcrash when logging in module: -gi-cpython-37m.dll version: Gajim-Portable-Master-64bit-2018-09-09.exe downgrading to 2018-09-02
  8. lovetox Orbit yes its known but not Gajims fault, i reported the bug all we can do is waiting
  9. marmistrz Is there any other way to find out if orientation is enabled in `print_conversation_line` than to add one more parameter to the function and pass it from above?
  10. lovetox what is orientation?
  11. marmistrz I see that the property is set in
  12. marmistrz s/orientation/encryption/
  13. marmistrz xD
  14. marmistrz I'll actually want to read `ChatControlBase.encryption`
  15. lovetox no there is a method in chatcontrolbase
  16. lovetox get_encryption_state()
  17. lovetox if this yields None, there is no encryption activated
  18. marmistrz lovetox, should I just pass it as a parameter to textview.print_conversation_line? I have doubts since the parameter list is atrocious, and I'd add one more parameter
  19. lovetox we have no reference to the chatcontrol in the textview hm
  20. marmistrz well, it'd be best to make a class out of it
  21. marmistrz well, it'd be best to make a class out of the parameter list
  22. lovetox haha, see you in 2 weeks :)
  23. marmistrz Yeah, that's why I'm not doing it :)
  24. marmistrz well, it'd be best to make a class out of the parameter list
  25. marmistrz damn, some correction quirks
  26. marmistrz Yeah, that's why I'm not doing it
  27. marmistrz it may be useful to type-annotate at least some of the arguments here because I have absolutely no idea what type some of them are :P
  28. lovetox no just pass a chatcontrol ref into the Textview on creation
  29. lovetox its useful anyway
  30. marmistrz lovetox, it will be circular deps
  31. lovetox what do you mean? because the child holds a ref to the parent? i dont see a problem with that
  32. marmistrz I know that in python you can get away with it, unless you at some point decide to type-annotate the signature so that other contributors know what to expect
  33. marmistrz I mean, that even if we don't type-annotate now, we shouldn't design the code in a way that will make it really hard later
  34. lovetox there are only these 2 options, either pass a ref on creation or on every method call
  35. lovetox choose one
  36. lovetox and i would not think about type annoting this, alone this module would be probably insane work
  37. lovetox whole Gajim probably a year of work
  38. lovetox i last week tried to annoate a simple utility function, wasted 45 minutes and it didnt work in the end
  39. marmistrz lovetox, yeah, that's why you do it as you go
  40. lovetox
  41. lovetox there annoate this function :D
  42. lovetox i thought it would be easy but was not possible to solve with 30 minutes of google
  43. bot Marcin updated a merge request for _gajim/master_ < >: WIP: Show an icon if a message is not encrypted.
  44. marmistrz Can I force redraw of the conversation window when the global encryption setting is changed?
  45. lovetox no, and this is also not needed
  46. marmistrz lovetox, when one reenables encryption, the crossed padlock needs to appear
  47. lovetox its just so that people that dont use encryption dont have constantly a red warning
  48. lovetox on new messages it does
  49. marmistrz yes, only new messages
  50. marmistrz but that's inconsistent UX
  51. lovetox not really, if i activate encryption i care what happens from now on encryption wise
  52. lovetox not what happend a day ago
  53. lovetox user can close his chatwindow and open it again, then he will see the encryption state of older messages
  54. marmistrz lovetox, but it means that there are two ways of indicating that a message was not encrytped
  55. marmistrz it's confusing for users
  56. lovetox Gajim warns you if something is not as you would expect it
  57. lovetox if you talk encrypted and there are unencrypted messages, you should be warned
  58. marmistrz
  59. lovetox if you talk unencrypted, the expectation is that that the chat is unencrypted
  60. lovetox so there is no need to warn
  61. lovetox and how often do you think people switch encrypted/uencrypted in a conversation?
  62. lovetox you can come up with a better idea, but fact is, if i dont use encryption i dont want to have in ALL my chats constantly warnings that i dont talk encrypted
  63. lovetox even worse, giving the user the feeling something is wrong, when in fact there is nothing wrong with talking unencrypted
  64. marmistrz rarely, indeed, so it's not a dealbreaker, but still it's bad UX for me. What Conversations does when the encryption gets enabled is to mark red all unencrypted messages, even the old oens
  65. lovetox yes thats ok for me
  66. marmistrz I still think it'd be better to reprint the textview - it's not going to happen very often and will improve the UX of such situations
  67. lovetox but its not doable in a elegant fashion
  68. lovetox you can try, but i tell you it will not work out like you want it to
  69. marmistrz ok, so I'll leave it now and open an issue about it, ok?
  70. marmistrz it may be a thing to keep in mind if the chat view is redesigned at some point in time
  71. lovetox yes
  72. marmistrz and how do you feel about those misaligned timestamp if the encrypted and non-encrypted messages are mixed (as on the screenshot above)?
  73. lovetox i dont care
  74. zuglufttier The timestamps have to go somewhere else, I think.
  75. marmistrz +1
  76. lovetox but again, you can try to change that and see for yourself
  77. marmistrz I'm fine with that, since it improves the UX for encrypt-all and doesn't matter for no-encrypt
  78. marmistrz (I'm fine with the misalignment here)
  79. bot Sophie Herold updated a merge request for _gajim/master_ < >: WIP: Rework Contact Info Dialog
  80. lovetox Link Mauve,
  81. lovetox Wrong timezone defintion: 2018-09-09T12:07:37.014717
  82. Link Mauve Ugh, this isn’t fixed yet?!
  83. Link Mauve poezio, why you do this?
  84. lovetox i think you fixed it but wrong :D
  85. Link Mauve Oh?
  86. lovetox or maybe that was another client i dont know anymoe
  87. lovetox last time a client added +:0000
  88. lovetox ah that was yaxim i remember
  89. Link Mauve lovetox, which iq is that?
  90. lovetox time
  91. lovetox <!-- Incoming 09.09.2018 14:14:08 --> <iq xmlns="jabber:client" from="" id="b36d8c2a-dba9-4d8c-b56c-138e78a57ee9" type="result" to=""> <time xmlns="urn:xmpp:time"> <utc>2018-09-09T12:14:09.039811</utc> <tzo>Z</tzo> </time> </iq>
  92. lovetox its stupid to add the Z to the utc tag if we have a tzo tag
  93. lovetox but thats what the XEP says
  94. bot Sophie Herold updated a merge request for _gajim/master_ < >: WIP: Rework Contact Info Dialog
  95. bot Sophie Herold updated a merge request for _gajim/master_ < >: WIP: Rework Contact Info Dialog
  96. bot Sophie Herold updated a merge request for _gajim/master_ < >: WIP: Rework Contact Info Dialog
  97. grey Hello
  98. bot Philipp Hörist pushed 1 commit to branch _refs/heads/master_ of _gajim_ < >: *a4970519* < > Add Builder class
  99. grey I'm using Gajim 1.0.3 and it will not let me add any bookmarked groupchats in any way. The buttons to do it are either greyed out or don't respond.
  100. grey What should I do?
  101. lnostdal grey: hm?
  102. grey I dunno.
  103. bot Philipp Hörist pushed 1 commit to branch _refs/heads/master_ of _gajim-plugins_ < >: *37105a88* < > [openpgp] Bugfixes - dont pass Path instance to sqlite3.connect() - add keydate to devicelist publish
  104. lovetox grey, which button? in the gropuchat menu?
  105. bot Philipp Hörist pushed 1 commit to branch _refs/heads/master_ of _gajim_ < >: *43270400* < > Request blocking list on connect
  106. bot Philipp Hörist closed an issue in _gajim_ < >: #9321: < Cannot allow contact to see my status >
  107. bot Philipp Hörist pushed 1 commit to branch _refs/heads/master_ of _gajim_ < >: *f82cd95f* < > Fix subscribing to contact Fixes #9321
  108. bot Sophie Herold updated a merge request for _gajim/master_ < >: WIP: Rework Contact Info Dialog
  109. bot Sophie Herold updated a merge request for _gajim/master_ < >: WIP: Rework Contact Info Dialog
  110. lovetox Maranda, need help with debugging
  111. lovetox i want to add my contact on another server
  112. lovetox if i send a presence i get back
  113. lovetox <presence xmlns="jabber:client" to="" from="" type="error" id="db95c057-e43c-44c7-92f2-89e345ce07c7"><error type="auth"><not-authorized xmlns="urn: etf:params:xml:ns:xmpp-stanzas" /></error></presence>
  114. lovetox do you run some spam protection?
  115. bot Sophie Herold modified an issue in _gajim_ < >: #9321: < Cannot allow contact to see my status >
  116. bot Sophie Herold modified an issue in _gajim_ < >: #9321: < Cannot allow contact to see my status >
  117. marmistrz lovetox: once you're done with saving the encryption state, mention me on this channel (alternatively, you can tell me which file needs patching and I'll add the change myself)
  118. bot Philipp Hörist pushed 1 commit to branch _refs/heads/master_ of _gajim-plugins_ < >: *820cec8c* < > [openpgp] Bugfixes - Parse datetime correctly and convert to int - Query contact keys on openpgp activation
  119. Maranda lovetox, yes
  120. Maranda lovetox, but you can disable it via adhoc too
  121. lovetox i already found it thanks
  122. Maranda lovetox, np
  123. bot Philipp Hörist pushed 1 commit to branch _refs/heads/master_ of _gajim-plugins_ < >: *9b3e23be* < > [openpgp] Update manifest.ini
  124. Maranda who does translations here?
  125. Maranda pokes guybrush88
  126. Maranda ‎[20:06:10] ‎errore durante l'invio *di* messaggio ( not-authorized )
  127. Maranda s/di/del/
  128. guybrush88 Maranda, è da un sacco che non ne contribuisco, domani mi riscarico il file delle traduzioni e risistemo tutto
  129. Maranda guybrush88, ok gimbo
  130. Maranda 👍
  131. lovetox marmistrz, what infos should we save on messages that we send?
  132. lovetox highest trust, and the fingerprint which we used to send the message?
  133. marmistrz lovetox, on sent messages we need to know if it was encrypted and which fingerprint was used
  134. marmistrz on received too
  135. marmistrz we definitely don't want to store the trust as it may change over time
  136. lovetox yeah but not for that message
  137. marmistrz lovetox, we already store the trust for fpr
  138. lovetox and about what trust do you think im talking about?
  139. marmistrz lovetox, which trust?
  140. lovetox there is only one trust, the one for the fingerprint
  141. marmistrz why should we double-store it?
  142. marmistrz it's already stored and shown in the OMEMO plugin settings
  143. lovetox because gajim cant access the database for the plugin
  144. lovetox so if you want gajim to show it, you have to give it to gajim
  145. lovetox and if you want that gajim shows it after a restart, you have to save it somewhere
  146. marmistrz So we either need the plugin to pass it to gajim on init or save a copy for gajim and not forget updating it anywhere in the plugin (!).
  147. lovetox i think you misunderstanding something
  148. marmistrz I'd rather go for the former as it's less error-prone - as long as it's not hard to implenet
  149. lovetox if we receive a message from a fingerprint, at that time we receive it, there is a certain trust for that fingerprint
  150. lovetox that will never change
  151. lovetox it will always be that trust at that point in time for that fingerprint
  152. lovetox and that message
  153. marmistrz lovetox, yes, but we rather want to display the current trust to the fpr
  154. marmistrz not the trust that was at the moment of reception
  155. lovetox why would someone look at the current trust of a fingerprint on old messages?
  156. marmistrz if the fpr was compromised or the fpr was trusted tentatively (we need communication, we'll verify it later as it's better than unencrypted)
  157. lovetox only because i untrust a fingerprint now because the device was lost, does not mean the messages that i wrote a week ago are untrusted
  158. lovetox so why would i show them as untrusted?
  159. marmistrz or because the message was the first from this fpr and I didn't mark it as trusted yet
  160. lovetox yes, if you dont save the trust you cant display that this was the first message
  161. marmistrz lovetox, but the device might've been stolen, so the messages may no longer be confidentional
  162. marmistrz lovetox, for the first messages the initial trust is unknown - so we *don't* trust the messages, and a warning *should* be displayed
  163. lovetox yeah but for that you need the trust level saved
  164. marmistrz sometimes we also temporarily mark the fpr as untrusted until we have an opportunity to live verify it
  165. lovetox because you cant save it as "unencrypted" because thats simply not true
  166. marmistrz lovetox, what does unencrypted have to do?
  167. lovetox you want a warning
  168. lovetox the only warning you have now is "unencrypted"
  169. marmistrz btw. I'd rather keep the behavior consistent with Conversations, which has it done really well.
  170. marmistrz yes, untrusted fpr will come later
  171. marmistrz as a next changeset
  172. marmistrz currently I only need whether it's encrypted or not, but latter I'll also need the fpr and ways of getting the current trust
  173. lovetox and how do you display that a fingerprint is not trusted, if we dont save the trust
  174. lovetox ?
  175. marmistrz check the current trust, as Conversations does
  176. lovetox so if you untrust a fingerprint in conversations, all messages before are flagged untrusted?
  177. marmistrz yes, exactly as in Conversations
  178. lovetox thats really not the behavior i imagine :D
  179. lovetox thats just bad
  180. marmistrz see, the fact that I received a message when I didn't trust the message yet doesn't make it untrusted
  181. lovetox yes thats why there are multiple trust levels
  182. lovetox that can be saved
  183. marmistrz lovetox, wait, what?
  184. lovetox and then you can display it accordingly later
  185. marmistrz I thought I can only save yes/no
  186. marmistrz I can't even go back to unknown
  187. lovetox we can save anything we want if we write it
  188. lovetox the point is i dont want to change history
  189. lovetox if i trust a message when i receive it, that should be documented
  190. marmistrz again: I may not yet trust the fingerprint, adding more trust levels will make it even more confusing for users
  191. marmistrz again: I may not **yet** trust the fingerprint, adding more trust levels will make it even more confusing for users
  192. marmistrz if we calculate it while printing the history, we don't need to rewrite history
  193. marmistrz we just display it accordingly
  194. lovetox and when you restart the cleint?
  195. lovetox that is rewriting history
  196. lovetox if you calculate it while loading
  197. marmistrz lovetox, you just get the current trust (which is already saved), the fingerprint (which will be stored) and display it accordingly
  198. marmistrz we don't change the old entries in the DB
  199. lovetox but you still dont show accuratly THE history
  200. lovetox if i write a year long messages which i trust
  201. lovetox then i change my key, and untrust it
  202. lovetox suddenly all these messages show up as untrusted
  203. lovetox sounds bad, i dont know what you gain from that
  204. marmistrz well, but if I mistrust an identity, I don't know which of the messages are valid and which not, and I don't trust any of them that's not a meritoric argument, but even my mom says that the way Conversations does it is the right one :P
  205. marmistrz and making it work completely different as in Conversations will make the users completely confused. I know that's not a meritoric argument, but, well, Conversations was first and formed the habits
  206. lovetox you are right thats not an argument
  207. marmistrz I talked with another friend of mine, and he also said that for him all of the past messages should me marked as untrusted, possibly with an additional information that they were trusted back then, but still warned about.
  208. marmistrz but still, the possibly compromised confidentiality is a good enough reason for me to mark those messages as not secure
  209. marmistrz (if it's the reason I mistrust the fpr)
  210. marmistrz alternatively, we can let the users choose between one of those two modi operandi, still, I want to implement my approach :)
  211. marmistrz alternatively, we can let the users choose between one of those two modi operandi, still, I want to implement my approach :) but you can add your approach as an alternative, to be chosen by those who don't like mine
  212. lovetox how generous of you
  213. marmistrz xD
  214. marmistrz btw. note that trust in PGP for mails also works the same - we display the current trust
  215. tms has anyone successfully used gajim over tor? the proxy option doesn't seem to work with .onion domains
  216. tms and neither does torify
  217. lovetox i dont think this will work with onion domains
  218. lovetox dns querys dont run through TOR, so the resolution will fail on the domains, and we dont have a ip to connect to
  219. tms yeah when running un-torified
  220. tms just curious why torsocks is failing
  221. tms most programs work just fine through it
  222. lovetox start gajim with -v so we can get some debug data
  223. tms 1536531421 WARNING torsocks[16043]: [connect] Connection to a local address are denied since it might be a TCP DNS query to a local DNS server. Rejecting it for safety reasons. (in tsocks_connect() at connect.c:192)
  224. tms makes sense, kind of. gotta do some more research
  225. lovetox Gajim needs the IP of the server
  226. lovetox if it cant get it via DNS resolution, you can specify it in the connection settings in the Account window
  227. lovetox under custom hostname
  228. tms wait a tic, perhaps I should disable the local tor proxy when using torsocks :o
  229. tms ah good idea
  230. lovetox im out, night
  231. tms hrm still being troublesome. but at least the regular traffic can be tor'd, only the dns lookup is clearnet
  232. tms perhaps I should just get myself a laptop that isn't crap and run tails in a vm
  233. bluemoon when is the new gajim supposed to come out?