Gajim - 2017-01-23

  1. kalkin lovetox: but you guys plan to implement it?
  2. lovetox yeah
  3. tm wuzzap, pgrep gajim (it is something like python
  4. lovetox its just me though :D
  5. kalkin if I would be you, I would add a hook which for message decoration, and do the timestamp, nick and other stuff over this hook
  6. kalkin lovetox: ahh, I thought asterix is also working on it :)
  7. kalkin s/which//g
  8. kalkin this way you could easily add plugins displaying i.e. the user avatar or bubbles around the msg
  9. lovetox i thought of implementing the icons for unencrypted/encrypted/verified beside each message
  10. lovetox and that the plugin only has to set a var in the event obj, before it passes it to gajim
  11. wuzzap thanks alot tm
  12. lovetox so gajim knows its encrypted and even could save this to the db
  13. lovetox so we know the status also for history messages
  14. kalkin lovetox: I would say this are two different things. The one is decryption and saving the msg to db, the other is the drawing of icon which is a special case of decorating a message. They should be at different “layers”
  15. kalkin msg in -> prerpocess msg -> send to gui layer -> decorate -> display in gui
  16. kalkin you could communicate to the gui layer that it needs to display a encryption lock via a var in the event obj, this is completely fine, but you would simplify a lot of things if you have a msg decorator hook at the GUI level.
  17. kalkin you could communicate to the gui layer that it needs to display a encryption lock via a var in the event obj, this is completely fine, but you would simplify a lot of things if you have a msg decorator hook at the GUI layer.
  18. kalkin (which looks at the set var)
  19. lovetox so that the plugin sets the icon?
  20. kalkin yes
  21. kalkin you could add a general encryption-decorator plugin of course
  22. kalkin it doesn't need to be an actual gajim-plugins plugin, but just something using the decorator hook
  23. lovetox we have such hooks already for the chatwindow, thats not a problem
  24. kalkin AFAIK for window, but not for each message, right?
  25. lovetox its just that this special use case, is needed in generell, for PGP, OTR, etc
  26. lovetox so i would like to make it into gajim core, rather then every plugin setting icons
  27. kalkin right
  28. kalkin let me explain it again from scratch, i think i'm just confusing you :)
  29. lovetox we can do this for every message
  30. lovetox for example gajim writes the timestamp then it passes the location of the cursor to the plugin
  31. lovetox the plugin can add a symbol and gives the cursor back
  32. lovetox gajim writes the name and the rest of the message
  33. kalkin So we could insert a padlock where gajim writes the timestamp and the nick?
  34. wuzzap oh nice i got the remote function to work yeha! omemo is used automatically if its activated before
  35. lovetox yes
  36. lovetox of course the hook has to be added at this location
  37. kalkin lovetox: well then we could add the lock icon at the same time when it's writing timestamp and nick.
  38. lovetox like between timestamp and name
  39. lovetox but thats like 5 lines of code
  40. kalkin lovetox: this would be a special case
  41. kalkin your UI needs a way to decorate a message, so you can add timestamp, nick & whatever
  42. kalkin so in the next version of gajim instead of hardcoding how it adds timestamp & nick, you could use a decorate hook
  43. lovetox thats a lot of work and nobody would use this
  44. kalkin lovetox: think about a plugin displaying avatars
  45. lovetox yeah i know someone could do a lot of nice things with it
  46. lovetox but i wait for someone who actually wants to do this, and asks me to add this functionality
  47. kalkin one of the biggest issues with gajim is that it's code is hacky ;), generalizing special cases is a way to clean up code
  48. lovetox what we want is padlocks
  49. lovetox and this is not complicated and can be added fast
  50. kalkin lovetox: how is the state of gajim master branch? Do you use it in day-to-day?
  51. lovetox yeah
  52. lovetox works for me :)
  53. kalkin So if I would start to hack on it I could actually test it during actual usage?
  54. lovetox yeah it works
  55. kalkin lovetox: it's still python2?
  56. lovetox no master is python3
  57. kalkin nice
  58. lovetox but just so you know, i think you overestimating what is possible with a gtk textview
  59. lovetox like decorating stuff with bubbles around messages etc
  60. kalkin lovetox: bubbles were just an example, i don't need them :)
  61. kalkin I just want to add stuff not inside the message text, so it can't be tampered with
  62. tm btw, what is development model there, i.e., master branch vs gajim_0.16. Where should I submit fixes, especially if they apply to both branches?
  63. lovetox gajim_0.16 is the one that is currently released
  64. lovetox its python2/gtk2
  65. lovetox in master branch we try to port gajim0.16 to gtk3/python3
  66. lovetox master branch works mostly, not everything though
  67. lovetox if you have fixes for both, make MR against both branches
  68. lovetox because gtk3 offers much new stuff, at some points the branches differ greatley
  69. lovetox master branch was never released
  70. lovetox so if you have new ideas or new features, or maybe see a way to do things alltogether in another way, you should do that in master
  71. tm ok, what about OMEMO specifically, given that gtk3 version is a little behind. Does it make any sense to send patches to gtk3 version at this point?
  72. andrey.g Once I tried to compare gajim_0.16 and master I wanted also to ask about the policy of committing. As I understand the both branches should be ideally totally equal with exception of python3/gtk3 stuff. But for it seems some commits landed in master and some in 0.16 which are actually not p3/gtk3 related. I'm wondering is it possible to review and cleanup this...
  73. lovetox tm, no i would only commit to gtk2 plugin branch
  74. lovetox to port the plugin to gtk3/python3 i need maybe an hour
  75. lovetox nobody really uses the master branch right now because its not released
  76. lovetox so i think its not too useful to spend to much time into applying each patch to both branches
  77. andrey.g I mean that with time it will be even harder to sync branches to ensure they differ only in p3/gtk3 but not in small fixes
  78. lovetox both branches should be ideally totally equal with exception of python3/gtk3 stuff.
  79. lovetox gajim is a UI application
  80. lovetox really much of the code is gtk3
  81. lovetox gtk3 breaks with many things and concepts of gtk2
  82. andrey.g Yes, that's what I mean. So ideally patches should be applied to both.
  83. lovetox of course if they are not gtk related
  84. lovetox and if they are security fixes, or stuff that fixes a bug
  85. lovetox but if you want to add something totally new
  86. andrey.g not p3/gtk3. And ideally p3/gtk3 patches should be distinct without any other changes (even small fixes / identations)
  87. lovetox we want that to be in gtk3
  88. tm lovetox, thanks for clarifying the situation
  89. lovetox ideally we want to ditch the gtk2 branch, only supply security fixes to it
  90. lovetox but because the gtk3 branch lacks stuff, and has some bugs, and missing features that are in gtk2 we cant release it
  91. andrey.g I'd say this would be realistic once first stable release of gtk3 comes out?
  92. lovetox yeah we should make a list with issues that hinder us to release the gtk3 branch
  93. lovetox and get this done
  94. lovetox because like it is right now its horrible
  95. lovetox any help is welcome :)
  96. andrey.g Similar problems are also in atril (gtk2->gtk3) vs evince (gtk3) pdf/... viewers. There is much work could be done. But despite of 7 Mrd. people still not enough to deal with it.
  97. tm I was actually interested in contributing to omemo plugin (especially the gui parts), so this state of the matter is quite unfortunate. Oh well.
  98. lovetox what did you want to do?
  99. lovetox just work on the gtk2 plugin, it is used by many people and they can benefit
  100. lovetox as i said, once we release the gtk3 branch i can port the plugin in an hour
  101. lovetox your work is not lost
  102. tm change the fingerprints window for example, to somewhat similar design to what is used in conversations
  103. andrey.g And this is a main advantage of working on top of 0.16. The things will get tested.
  104. lovetox tm i see no problem why you shouldnt do this :)
  105. andrey.g tm, unless you like to use some gtk3 widget for that
  106. tm gtk3 widgets are part of the story :)
  107. andrey.g and what widget?
  108. lovetox gtkswitch?
  109. tm yeah, something like gtk switch would be more convenient (rather than separate button), though there are more than two states
  110. lovetox yeah i thought about this, and how to solve that problem?
  111. andrey.g CheckButton?
  112. lovetox if i have time i bring the omemo plugin up to speed in the gtk3 branch, but first we will release a new version with the pening MRs
  113. lovetox *pending
  114. lovetox im going to bed, have a good night
  115. bot Yann Leboulanger pushed 1 commit to branch _refs/heads/gajim_0.16_ of _gajim_ <>: *744ee063* <> fix bilding installer with correct translations
  116. bot Yann Leboulanger pushed 1 commit to branch _refs/heads/gajim_0.16_ of _gajim_ <>: *23443819* <> Fix postable installer with correct languages list
  117. elmo Hi! I was asking about getting Gajim to run fine on Mac, yesterday. I was linked to a guide and I got everything to run -- except half of icons don't display.
  118. elmo -- screenshot
  119. elmo The console output is also printing things like
  120. elmo (gajim:45664): Gtk-WARNING **: Error loading theme icon 'system-run' for stock: Icon 'system-run' not present in theme Adwaita
  121. elmo Aha! Installing adwaita icon theme helped.
  122. lovetox elmo can you tell how exactly you installed the icon theme
  123. lovetox then i could add it to the guide
  124. elmo Sure.
  125. elmo brew install adwaita-icon-theme
  126. elmo Is the Gajim's end-to-end encryption supposed to be compatible with ChatSecure on iOS?
  127. lovetox no
  128. elmo :( Aren't both OMEMO?
  129. lovetox there are many encryptions
  130. lovetox gajim supports omemo with plugin
  131. lovetox to install it you have to install some dependencys
  132. lovetox i think this has never been tried on mac
  133. elmo I installed pycrypto, and "End to End message encryption" turned on in Features.
  134. elmo It kind of works, I can see my messages to test account being encrypted, in XMPP console
  135. elmo So Gajim→Gajim works fine
  136. mathieui End-to-End message encryptions probably doesn’t mean omemo
  137. mathieui but rather the deprecated stuff that gajim uses by default
  138. lovetox elmo you need the gajim-omemo plugin
  139. lovetox for that you have to install various dependencys
  140. elmo ← sighs. Do I care enough to dig for it?
  141. lovetox i dont know
  142. lovetox tell me if you know :)
  143. elmo Another question. Whither will a message get delivered if all resources are of equal priority?
  144. elmo To all of them?
  145. lovetox on modern servers the priority will not have an impact on what resource gets what message
  146. lovetox if you server supports, CarbonCopys and MAM, all resources will get all messages all the time
  147. elmo Aha! Thanks!
  148. bot Philipp Hörist pushed 2 commits to branch _refs/heads/gajim_0.16_ of _gajim_ <>: *afb2fb1f* <> Include msg kind when merging nicknames When two users had the same nickname merging was not working correctly *243ab148* <> Merge branch 'gajim_0.16' into 'gajim_0.16' Fix chat_merge_consecutive_nickname See merge request !34
  149. bot Philipp Hörist pushed 5 commits to branch _refs/heads/gajim_0.16_ of _gajim_ <>: *dccbd1d1* <> Store passwords in Windows Credential Vault *f023426f* <> Add keyring dependency to windows build *d691ccbe* <> Add config setting for windows keyring *f715de36* <> Update available features on windows On Windows we can now save account passwords in the Windows Credential Vault *83c9bdb8* <> Merge branch 'gajim_0.16' into 'gajim_0.16' Store passwords in Windows Credential Vault See merge request !35
  150. bot Philipp Hörist pushed 4 commits to branch _refs/heads/master_ of _gajim_ <>: *5a2a83a3* <> Update available features on windows On Windows we can now save account passwords in the Windows Credential Vault *e41a751c* <> Add config setting for use of keyring *f59fd73f* <> Improve fallback for windows keyring *2be4b0a5* <> Merge branch 'master' into 'master' Improvements for windows keyring storage See merge request !38
  151. arune Keyring for 0.16? Nice
  152. bot Yann Leboulanger pushed 2 commits to branch _refs/heads/master_ of _gajim_ <>: *eb55d81e* <> Use python executable located by configure when running tests. *9b5ec09c* <> Merge branch 'test-python-version' into 'master' Use python executable located by configure when running tests. See merge request !37
  153. tm ail
  154. lovetox hey :)
  155. lovetox tm i will try to port all changes from omemo gtk2 to the gtk3 branch tomorrow
  156. lovetox then you can focus on gtk3 if thats more to your liking :)
  157. tm lovetox: ok, great :)
  158. tm I am currently playing with shortcuts window, they look pretty cool
  159. lovetox yeah something like that would be great
  160. lovetox i took me one year to find out how to correct a message :D
  161. tm hmm, should creating a new line with shift-enter work (instead of ctrl-enter)?
  162. lovetox yeah i think its buggy in gtk3
  163. lovetox on gtk2 branch i can create a new line with
  164. lovetox shift+enter, and ctrl+enter
  165. lovetox but there is also a option in the advanced config editor
  166. lovetox to send message on ctrl+enter
  167. lovetox test
  168. lovetox so if this is activated then Enter should make a new line
  169. anotheragency Need help with the last few of these strings
  170. anotheragency Explanations and fixes are in order