Gajim - 2018-01-03

  1. Citizen Zibb oi I missed a commit
  2. Citizen Zibb ?
  3. Citizen Zibb byy the way, on fedora im getting this error upon trying to start gajim
  4. Citizen Zibb Traceback (most recent call last):
  5. Link Mauve Yeah, it’s a super bad idea to rely on that, I reported it some time ago but got told I “greatly overestimate the problem”. ^^
  6. Link Mauve See
  7. Citizen Zibb Link Mauve, I've been seeing this issue quite often
  8. Citizen Zibb it's supremely irritating as for a while I didn't know what the problem was as the program would exit silently
  9. Citizen Zibb running in terminal yielded that information
  10. Link Mauve Citizen Zibb, a possible reason would be if you’re using a projector, those don’t have any intrinsic size in millimeters, so some compositors will set both width and height to 0mm.
  11. Citizen Zibb no projector here, friend
  12. Citizen Zibb just a simple, albeit very crappy, monitor
  13. Citizen Zibb though, on one of my virtual machines, this issue pops up, maybe because of the guest resizing the window
  14. Link Mauve Virtual machines also have this issue of not knowing the physical size of the non-physical monitor.
  15. Link Mauve This is why using GdkScreen is a bad solution.
  16. Link Mauve Citizen Zibb, you may want to comment on this issue to add your problem.
  17. Link Mauve Or open a new one and link it there, that may be better.
  18. Citizen Zibb I don't want to seem like I'm greatly exaggerating the issue
  19. Citizen Zibb Link Mauve, maybe phil thinks its more trouble than its worth to fix
  20. Citizen Zibb but at least a workaround so I can run the goddamn thingf
  21. concerto Hola. What directories do I need to copy to backup Gajim on Linux?
  22. concerto Anything other than ~/.local/share/gajim and ~/.config/gajim/ ?
  23. Link Mauve You may want to also keep your cache, in ~/.cache/gajim, or don’t and let it be regenerated the next time you start it.
  24. Link Mauve I personally don’t backup anything from ~/.cache.
  25. concerto Cool, thanks :)
  26. bot Jeffrey Baitis updated a merge request for _python-nbxmpp/master_ <>: Detect invalid file handles when passed to SelectIdleQueue
  27. bot Jeffrey Baitis updated a merge request for _python-nbxmpp/master_ <>: Detect invalid file handles when passed to SelectIdleQueue
  28. bot Jeffrey Baitis updated a merge request for _python-nbxmpp/master_ <>: Detect invalid file handles when passed to SelectIdleQueue
  29. bot Jeffrey Baitis updated a merge request for _python-nbxmpp/master_ <>: Detect invalid file handles when passed to SelectIdleQueue
  30. rom1dep lovetox: eh, because of you, now I end up pressing ctrl+n frequently when using Skype for business at work. And it crashes/freezes it. Good work over there!
  31. Marzanna Does image plugin work in gtk3 Gajim?
  32. lovetox Link Mauve, i used gajim on almost all systems, that get screen_height_mm() returns 0 is broken in recent versions of gtk, and yields now on any VM zero
  33. lovetox it seems even on real monitors
  34. lovetox so yeah this got now a higher priority
  35. lovetox albeit its still not clear how to replace it, as the topic is quite complex and needs someone who knows in and outs about DPI and conversion of units such as pt
  36. lovetox Marzanna, no idea just try it
  37. Marzanna lovetox, I don't see a button
  38. lovetox any errors in the conosle?
  39. Link Mauve lovetox, it’s been deprecated for years, I’m surprised it got broken before 4.0, but I don’t believe any VM or projector could work at all before.
  40. lovetox it did, i run gajim exclusively in avm
  41. Link Mauve Which values was it returning before, for the physical size?
  42. lovetox but i fix it now, after reading 15 mins into dpi stuff
  43. lovetox i dont know, it did not throw an exception though
  44. lovetox and NOBODY sends xhtml with css
  45. lovetox so this code i never tested
  46. Link Mauve Hi, here is some CSS.
  47. lovetox you have to set a font-size attribute in "em" or "points"
  48. lovetox and make it big :9
  49. Link Mauve Test in em and in points
  50. lovetox nice works :)
  51. lovetox
  52. Link Mauve Looks pretty big though.
  53. lovetox the em to pixel calc is probably wrong ^^
  54. Link Mauve Here is how your web browser renders it:
  55. lovetox 1 em is the default font size
  56. int lovetox, is there a simple solution for me to shrink the "cellpadding" between the chat tabs?
  57. lovetox where the fk should i get the default font size from
  58. lovetox not really int sorry but im working on new tabs that allow for more space and scrolling
  59. lovetox Link Mauve, em is totally variable
  60. lovetox it depends what your browser uses for a font, and what the default font size is for that
  61. int lovetox, good to know.. i need more space :D always have to scroll and oversee new messages
  62. Link Mauve lovetox, of course, it depends on your inherited size.
  63. lovetox i would say you can test my MR, but the resolving problem is not fixed
  64. lovetox although i got a response on the bug report at gnome
  65. Link Mauve pt you can’t really guess correctly.
  66. Link Mauve But em you should be able to translate it correctly.
  67. Link Mauve lovetox, what are the units you can already use?
  68. lovetox ex, px, pt, em
  69. lovetox though em seems hard to implement correctly
  70. lovetox because i would have to know if there is a parent to the css element that maybe sets a different size
  71. lovetox i dont want to write a full fledged css parser
  72. lovetox you even would need a html parser for that
  73. lovetox too complicated
  74. lovetox or Link Mauve ? if i dont parse the parent and look for font size it makes no sense to support "em"
  75. lovetox i could set a default size from where it calculates, but i doubt that makes much sense
  76. Link Mauve lovetox, you mean treat it as rem?
  77. lovetox i mean i ignore that it depends on the parent, and set a resonable default value
  78. Link Mauve That’s the rem unit.
  79. lovetox ah k didnt know
  80. lovetox its the same with %
  81. lovetox which gajim also supports which also depends on parent
  82. lovetox i just take user configed default font size
  83. lovetox nobody will really do that complicated css stuff
  84. b_b hi
  85. b_b Asterix ftr i've proposed small changes for fr translation
  86. b_b
  87. b_b absent(e) => absent⋅e
  88. b_b same for occupé⋅e
  89. b_b more inclusive ;)
  90. Link Mauve b_b, that’s exactly as inclusive, though. :p
  91. Link Mauve There are a few places in Gajim using “his” to talk about a contact, though.
  92. b_b not sure if you are saying that (e) is the same that ⋅e Link Mauve ?
  93. Link Mauve Yes.
  94. b_b sorry, but no :p
  95. Link Mauve Why?
  96. b_b cause putting the feminism between () is not inclusive
  97. b_b there's plenty of text about it available on the web
  98. Marzanna lovetox, yes. It has errors File "/home/marzanna/.local/share/gajim/plugins/image/", line 81, in __init__ actions_hbox.pack_start(self.button, False, False , 0) AttributeError: 'NoneType' object has no attribute 'pack_start'
  99. b_b i let you searc & read about that ;)
  100. Link Mauve Ok.
  101. b_b a godd start is romy article
  102. b_b and a longer one
  103. mathieui the second one doesn’t advocate against parenthesis at all
  104. mathieui (and it actually says that using it is inclusive)
  105. ivucica Totally feels like authoritative and not an opinion of one person. (I don't speak French, so I'm just guessing what this is all about.)
  106. ivucica Anyway, #off. Bye
  107. mathieui bye
  108. b_b funny to see how a small change like this can generate comments :)
  109. b_b anyway, i don't care, but it's caracter less and more easy to read :p
  110. mathieui well, I don’t care either, because I find both equivalent ^^ (but I don’t see a reason to reject the change either)
  111. b_b nice, we will see if it get merge
  112. ThibG I'm slightly in favor of that change, personally
  113. b_b maybe i should have proposed the change with a logged account...
  114. Link Mauve b_b, I don’t know how it is with Pootle, but with Weblate you can change a translation right away when you are logged in, like on Wikipedia or OpenStreetMap.
  115. b_b i have just tilted when i've seen Absent(e) (that i see every day since years), so i've proposed right away without more reflexion
  116. Link Mauve There are a bunch more sexist language around in Gajim, “his status” is something I remember seeing.
  117. Link Mauve If you want to fix it that’d be nice.
  118. Link Mauve Maybe directly in the original strings.
  119. mathieui Link Mauve, I guess b_b is here to fix french :p
  120. b_b yep, i'm native french so i will not try to fix en strings :p
  121. Link Mauve Imo it’d be nicer to fix English first, or at least to tell where bugs are in the original strings.
  122. b_b and we all know that en is really != fr for male/female differenciation
  123. b_b a boat is a girl in en
  124. lovetox Link Mauve, if you change the original string, that invalidates all translation
  125. Link Mauve Otherwise the translation will be wrong.
  126. b_b in french it's a guy
  127. Link Mauve lovetox, yes.
  128. Citizen Zibb in latest git compile, chat roster is blank
  129. lovetox then restart
  130. Citizen Zibb the app or computer
  131. Citizen Zibb they show up for jdev chat room
  132. lovetox gajim
  133. b_b the app, your computer, your box, and maybe the electricity Citizen Zibb
  134. b_b => []
  135. b_b ^^
  136. Citizen Zibb Traceback (most recent call last):
  137. Citizen Zibb lovetox, it locks up and I have to ctrl-alt-delete out of it
  138. Citizen Zibb (btw running as portable exec if it makes a difference)
  139. lovetox delte cache.db in the userdata folder
  140. lovetox did you upgrade from gajim 0.16?
  141. Citizen Zibb lovetox: didn't work
  142. bot Jeffrey Baitis updated a merge request for _python-nbxmpp/master_ <>: Detect invalid file handles when passed to SelectIdleQueue
  143. baitisj
  144. baitisj hi lovetox. Sorry for my newb mistakes. hopefully I've fixed stuff now.
  145. lovetox yeah looks good
  146. lovetox did you dev that patch against 0.16?
  147. baitisj lovetox, once the merge request is satisfied and the changes included, do /you/ close it?
  148. baitisj lovetox, the patch is against the master branch
  149. lovetox the MR it closes automatically
  150. lovetox once its merged
  151. baitisj Got it. And then, from there, I can incorporate upstream changes, and generate a new merge if needed. I think i understand how gitlab works now, and have a much better idea of how I should initiate merge requests in the future
  152. baitisj lovetox, did you want me to try versus 0.16?
  153. baitisj actually it's probably better if I check to see if this patch resolves my windows client issues
  154. lovetox generally you dont need to include upstream changes
  155. lovetox gitlab checks if upstream changes break the merge request
  156. lovetox for example if you do a = 1 and i do a commit upstream a = 2
  157. lovetox gitlab doesnt know what is the correc tone
  158. lovetox then it flags the MR as not mergeable
  159. baitisj at that point I would guess it's best to cherry-pick upstream commits and resolve the conflicts within your MR?
  160. lovetox in that case you would rebase your MR and incorporate upstream changes and fix your MR
  161. baitisj Ah
  162. baitisj okay, that makes perfect sense
  163. lovetox but if your MR is flaged as "mergable" (green)
  164. lovetox then upstream changes have nothing to do with your changes, and do not create problems
  165. baitisj thanks for taking the time to teach
  166. lovetox still devs could ask you before merge to rebase and incorporate all upstream changes because gitlab merges the commit to the date where you commited to your fork
  167. lovetox if your MR is open months
  168. lovetox this would mean your commit will be merged way back in time not showing up on the top
  169. baitisj is that mainly for accounting purposes so that it remains visible in the log, effectively, in a chronological manner that better reflects reality?
  170. lovetox yeah, also it could be that you open a MR, devs commit some stuff, release a new version, then later merge your commit, but it would show now up back in time before the version release
  171. lovetox but in fact its not in the released version
  172. lovetox also the MR is for review, discussion, changes, etc, so it doesnt have to be right the "first time" there is nothing wrong in adding stuff, removing, whatever
  173. baitisj I rather like how gitlab makes this clarification. I'm not sure if I've seen the same granularity in discussion in github
  174. baitisj wow, I really like the interaction on gitlab.
  175. bot André updated a merge request for _gajim/master_ <>: Port music_track_listener to GTK dbus
  176. lovetox somehow our runners for the tests are down
  177. lovetox i tell asterix once this is fixed and the test run through we can merge
  178. baitisj lovetox, I think our conversation has convinced me that I should make a checking routine for the GlibIdleQueue as well. That way on my FreeBSD system I wouldn't need to use a hack that forces the return of the SelectIdleQueue
  179. lovetox yes would be appreciated. But we can add this in a future MR, this works now and should be merged soon
  180. baitisj Acknowledged. I'll create another branch for the future MR and get my ideas down since the iron is hot right now
  181. lovetox also did you not try to do that in the first place?
  182. baitisj I didn't do that in the first place because the implementation of gobject is opaque whereas the SelectIdleQueue is obvious as to how it interacts with OS-level calls
  183. lovetox but the check should be the same, the fds are the same
  184. baitisj I agree :-)
  185. lovetox Glib probably will not throw a OSError though
  186. lovetox i dont know what it throws or if it does, only someone with that issue can find out :)
  187. baitisj I think it's worth trying and finding out :-)
  188. baitisj hence I'll branch and give it a shot
  189. baitisj sorry to have to ask this, but is the pythonic convention that _method_name is akin to a C++ virtual function ? ie must be implemented by subclass
  190. baitisj or is it that _method_name is akin to C++ private function definition?
  191. lovetox no
  192. lovetox yes
  193. lovetox it meant to be private
  194. baitisj excellent. thanks
  195. lovetox but it is only naming convention, its not private like in C
  196. bot Philipp Hörist closed an issue in _gajim_ <>: #8491: <Remove any usage of Gdk.Screen>
  197. bot Philipp Hörist pushed 3 commits to branch _refs/heads/master_ of _gajim_ <>:
  198. lovetox ahhh another Link Mauve topic closed, now i can go to bed