Gajim - 2018-08-08

  1. mrDoctorWho new new history manager is great!
  2. mrDoctorWho the new, sorry
  3. lovetox ProfP30, what Gajim version?
  4. lovetox Daniel, does this problem still exist?
  5. ProfP30 lovetox: current, 1.0.3
  6. lovetox it has nothing to do with my MR, but i dont think after restarting Gajim formatting is kept
  7. lovetox and from where did you download plugins?
  8. ProfP30 lovetox: [gajim / gajim-plugins · GitLab]
  9. ProfP30 lovetox: and following the instructions there
  10. lovetox try downloading the zip from here and installing with the button
  11. lovetox
  12. bot Jan Ulrich Hasecke created an issue in _gajim_ < >: #9281: < AttributeError: 'ActionOption' object has no attribute 'action' >
  13. ProfP30 lovetox: thank you very much!
  14. Link Mauve Oh, there is now an early guide for GTK+ 4.0 migration:
  15. Link Mauve It’s mostly about not using deprecated stuff anymore though.
  16. Link Mauve And Gajim seems to be doing fairly well for now.
  17. lovetox Link Mauve, im surley not doing the same mistake like last time with GTK3
  18. lovetox If we port to GTK4 then it will be fast in one go
  19. lovetox and this can only happen if GTK4 is final, and on every system available
  20. Link Mauve Sure, but it also won’t be that different either.
  21. lovetox so i consider this not happening for the next 2 years
  22. Link Mauve You can easily support both 3.x and 4.0 in a same application, not so much for 2.x and 3.0.
  23. lovetox no we cannot
  24. Link Mauve Why?
  25. lovetox GTK4 is major relase it will break many things
  26. Link Mauve It mostly boils down to stop using deprecated functions and properties and signals from 3.x.
  27. lovetox no it does not
  28. Link Mauve How so?
  29. lovetox they redo big parts
  30. Daniel lovetox, I don't know if that is still an issue, it just came to my mind while looking at your MR (some fonts parts)
  31. lovetox changing concepts
  32. Link Mauve Sure, internal parts.
  33. lovetox no, please readup on the roadmap
  34. Link Mauve But the overall API is still the same.
  35. lovetox no not at all
  36. lovetox because then you would not need a major release
  37. Link Mauve Sure you would, a major release is the only point where they can drop deprecated symbols.
  38. pep. lovetox: I could very well add a comment to the code and make a major release out of it if I wanted
  39. Link Mauve There are some changes in the API, but it’s primarily an internal redesign which is causing the bump in major version.
  40. lovetox pep. where did i say the CANT do a major release?
  41. lovetox i said they dont need it
  42. Link Mauve I don’t understand what you are trying to say.
  43. lovetox Its makes no sense to support 2 GTK versions
  44. lovetox for what?
  45. Link Mauve It kind of does, their new GPU acceleration will remove many current bottlenecks for instance.
  46. Link Mauve And keeping support for 3.x helps people still on obsolete distributions.
  47. lovetox you just make things up now
  48. lovetox Gajim need no GPU acceleration 😃
  49. Link Mauve Haha.
  50. Link Mauve Have you ever profiled it?
  51. lovetox have you?
  52. Link Mauve Yes, when you were asking me about my roster.
  53. Link Mauve Turns out drawing it was taking a long time.
  54. Link Mauve Same for MUCs.
  55. lovetox not because of a bottleneck in GTK3
  56. lovetox you know, developer has some little part in what gets drawn how often
  57. Link Mauve Of course, you can either optimise it (but you haven’t done so in the past few years), or you can throw it to the GPU and let it do what it does best.
  58. Link Mauve gtk4 chose the latter option.
  59. lovetox and how is this of any priority?
  60. lovetox do you think on a list with all things, drawin a roster faster is anywhere in the top 100 of things Users want in Gajim?
  61. Link Mauve You are the one chosing your priorities, I’m not telling you to do anything.
  62. Link Mauve I’ve heard this complaint a few times, yeah.
  63. lovetox point is there is no need to use GTK4 for that
  64. Link Mauve What else would you use? OpenGL directly?
  65. lovetox there is no need to support a moving target
  66. lovetox that breaks apis at each version
  67. Link Mauve The only point of this discussion was that porting to gtk4 shouldn’t be very hard, not that you should priorise that or anything.
  68. lovetox point is there is no need to support a dev version that is in deployed in different version on each desktop
  69. Link Mauve lovetox, it’s not a moving target to first remove your usage of deprecated APIs.
  70. Link Mauve Of course not.
  71. lovetox there is no need to remove deprecated apis if you dont switch to gtk3
  72. lovetox *gtk4
  73. lovetox GTK4 is not a thing
  74. lovetox its a dev version
  75. Link Mauve They are deprecated for a reason, but I see your point.
  76. Link Mauve lovetox, and this is relevant because?
  77. lovetox wait 1 year, let it be a release
  78. lovetox wait another year to get all bugs fixed and distributed to desktops
  79. lovetox then it makes sense to swithc
  80. Link Mauve And I’m telling you there is no point in switching.
  81. lovetox it just makes no sense time/cost wise
  82. Link Mauve You can use both at the same time pretty easily.
  83. Link Mauve Be it now or in two years.
  84. Link Mauve There are already very few deprecated functions being used in Gajim.
  85. lovetox they will break the api in big ways
  86. Link Mauve They did not, and will not.
  87. lovetox its not finished you know
  88. Link Mauve There are some changes for things which couldn’t stay the same without contorsions in the internal APIs, but overall the user API stays mostly the same, with some very welcome additions though.
  89. lovetox lets see when its finished 🙂
  90. Link Mauve lovetox, well, just look at the roadmap and tell me which part of the API used by Gajim will be horribly broken by this change?
  91. lovetox for example status icon removal
  92. Link Mauve That one has been deprecated for years now.
  93. lovetox alone for that we should wait as long as possible to swtich, and only do it if there are MAJOR advantages
  94. Link Mauve Note that many users still can’t use it.
  95. Link Mauve Note that many users already can’t use it.
  96. lovetox yeah and? many can
  97. lovetox im not saying everyone has to use a status icon
  98. lovetox there is missing a replacement cross platform
  99. lovetox Gio has no notification backend for windows
  100. Link Mauve Is that still not fixed? :|
  101. growingupagain Hello! How can gajlm send subscription request presence when get message from third user?
  102. growingupagain I mean how to do it without gui extention points?
  103. bot Sophie Herold proposed a new merge request for _gajim/master_ < >: WIP: Replace menubar and appmenu with popovers
  104. MattJ lovetox, does Gajim handle the new 333 status code in MUC? (the "error leave" one)
  105. lovetox no
  106. lovetox is that in the xep?
  107. MattJ Yes:
  108. MattJ The example shows a 307 in there as well, but we removed that in Prosody today
  109. lovetox hm but would you lose then backwards compatibility?
  110. lovetox so Gajim does not know anymore if someone gets kicked now?
  111. lovetox unless i add this ?
  112. bot Sophie Herold updated a merge request for _gajim/master_ < >: WIP: Replace menubar and appmenu with popovers
  113. lovetox i guess i will add this then soon
  114. MattJ lovetox, no
  115. MattJ lovetox, there are three reasons someone might leave a room: 1) they leave normally 2) a moderator kicks them 3) the server removes them (due to an error, like s2s problems)
  116. MattJ (1) does not need a status code, (2) is 307, (3) is 333
  117. MattJ But 333 is new. Prosody 0.10 used to show (3) as a normal leave, but when our MUC code got rewritten in trunk, the person who wrote the code felt that showing it differently was better, and added 307 to show that it was a kick (by the server, not a moderator)
  118. MattJ But then people thought that people were being kicked from a MUC by a moderator, which was not true
  119. MattJ 333 is a new code that makes it clear what is happening
  120. MattJ poezio is being updated from "UserX has left" to "UserX has left due to an error" when 333 is present
  121. MattJ So 307 -> kick by a moderator (like it has always been historically), 333 -> removed by the server (new feature)
  122. lovetox k thanks for the explanation, yeah adding this is not a problem, will look if i have time this evening
  123. Link Mauve lovetox, is now fixed, poezio won’t reply with invalid iq:version anymore.
  124. lovetox nice
  125. mdosch Gajim says http_upload unavailable on my prosody trunk (updated ~1h ago). But it works (confirmed with conversations in a MUC). I am running Debian Testing. gajim/experimental,experimental,now 1.1~git20180728-1 all
  126. lovetox yeah we fixed that problem already
  127. mdosch Ah OK. So I'll just wait for new experimental package. THX for the info.
  128. bot Sophie Herold proposed a new merge request for _gajim/master_ < >: Mass adjust .po files to new labels without colons
  129. bot MaxXor created an issue in _gajim_ < >: #9282: < Python 3.7: Error when clicking on a link in chat >
  130. bot Philipp Hörist closed an issue in _gajim_ < >: #9282: < Python 3.7: Error when clicking on a link in chat >
  131. bot Philipp Hörist pushed 1 commit to branch _refs/heads/master_ of _gajim_ < >: *aa5fca86* < > Align vcard dialog labels with profile dialog
  132. bot Philipp Hörist merged a merge request for _gajim/master_ < >: Align vcard dialog labels with profile dialog
  133. bot Philipp Hörist pushed 2 commits to branch _refs/heads/master_ of _gajim-plugins_ < >: *575eb729* < > [omemo] Show all fingerprints in one list *4e1fe417* < > [omemo] Update manifest.ini
  134. zuglufttier Nice, all fingerprints in one list!
  135. zuglufttier I stumbled upon that today ;)
  136. bot Philipp Hörist closed an issue in _gajim-plugins_ < >: #306: < InvalidKeyException: Invalid signature on device key! >
  137. bot Philipp Hörist closed a merge request for _gajim/master_ < >: Mass adjust .po files to new labels without colons
  138. bot Philipp Hörist closed an issue in _gajim_ < >: #9274: < Save translations from getting lost by trivial msgid changes >
  139. bot Philipp Hörist pushed 1 commit to branch _refs/heads/master_ of _gajim_ < >: *f438a27a* < > Handle new MUC status code 333
  140. lovetox MattJ, ^