Gajim - 2016-08-09

  1. lovetox arune
  2. lovetox the windows tray icon problem
  3. lovetox you can simply go to the tray menu and tell windows to always show it
  4. lovetox it feels weird to make an application that ignores this from the start
  5. lovetox what if everyone would do this
  6. lovetox then the whole feature that micrsoft integrated would have no sense
  7. lovetox on building gajim, i can build on a platfrom called appveyor
  8. lovetox you could just create a bitbucket acc, import gajim there (2 clicks)
  9. lovetox make your changes
  10. lovetox go to appveyor, add the bitbucket account, and click build
  11. lovetox 5 minutes later you have an installer for download
  12. frainz Hi, I'm looking for a way to set archiving preferences (for xep-0313, mam) on the server. Could you tell me, where to do this in gajim?
  13. kalkin frainz: I think you need to setup this on the server
  14. kalkin AFAIK you can't setup this from your client
  15. SouL I think the user can configure what messages (or what kind of messages) wants to store, kalkin.
  16. SouL «This module will archive all messages that match the simple rules setup by the user, and allow the user to access this archive.»
  17. kalkin SouL: Thanks TIL
  18. Holger Gajim doesn't have a knob for that, some other clients do.
  19. Holger If all else fails you could throw the desired request into the XML console:
  20. frainz does Gajim interfere with settings specified by other clients?
  21. SouL kalkin, this is how Conversations show that
  22. Holger No. You can use another client to configure the settings for your account and Gajim won't change them.
  23. kalkin SouL: awesome thanks
  24. frainz that's good :) thank you
  25. arune lovetox: yes, a user can go to the tray menu and always show it. But in my case where I have 60 users, even if they are engineers, most of them even software engineers, they won't do this, they don't notice the icon is hidden or think that the icon will show once they get a message
  26. arune lovetox: I'll try that bitbucket/appveyer
  27. kalkin also my wife just didn't got it with status icon in the gnome message bar
  28. kalkin she just always reopens a new instance if she can't find Gajim
  29. kalkin May be Gajim shouldn't open a new instance if another is running? At least not if you open it via .desktop file
  30. kalkin It should just show the current instance
  31. arune kalkin: normally gajim will tell you it's already running. But I agree it should show you the running instance
  32. kalkin arune: well it will and the user will ignore the message and just click on the button to just open another instance, because the user can't find it
  33. kalkin arune: well it will and the user will ignore the message and just click on the button to just open another instance, because the user can't find the original instance
  34. kalkin arune: well it will and the user will ignore the message and click on the button to just open another instance, because the user can't find the original instance
  35. arune lovetox: where's your daily build of the default branch for win? The link should be added to the trac
  36. tmolitor arune, lovetox: the windows behaviour of putting tray icons into an overflow area is useful for all those icons that you don't usually interact the openoffice quicklauncher or the various drivers that put icons in the systray that you never need in your whole live...
  37. tmolitor putting those away makes sense, but it doesn't make sense to put interactive icons away, such like skype or gajim or such...
  38. tmolitor because of this I would propose to put the registry "hack" into the installer so that gajim is always visible...
  39. tmolitor kalkin: your suggestion to just open the already running instance would be helpful, too (but those two things should both be implemented)...
  40. Link Mauve With gtk3, GtkApplication can handle the uniqueness for you, btw.
  41. kalkin tmolitor: best way would be if ./ would have some kind of switch --resume-if-exists, which is used when called by the Gajim.desktop file. The possibility to start multiple instances should be kept
  42. kalkin Also Gtk StatusIcon shouldn't be used because it's deprecated. Just use notifications o_O
  43. tmolitor kalkin: yes, exactly :)
  44. kalkin how does gajim + gtk3 solves this issue, with the statusicon
  45. tmolitor I don't know...
  46. kalkin I'm currently need to implement a custom statusicon with menu ala gnome networkmanager, but without using gnome shell JS api, because it needs to work in XFCE & KDE
  47. kalkin And I'm kind of thinking about "just" creating an appindicator rpm (and hope it works besides ubunut). Or just use Xlib
  48. kalkin And I'm kind of thinking about "just" creating an appindicator rpm (and hope it works besides ubunut). Or "just" use Xlib
  49. kalkin And I'm kind of thinking about "just" creating an appindicator rpm (and hope it works besides ubunut), or "just" use Xlib
  50. kalkin And I'm kind of thinking about "just" creating an appindicator rpm (and hope it works besides ubuntu), or "just" use Xlib
  51. kalkin Also using StatusIndicator doesn't allow to have custom Menu & MenuItem objects *sigh*
  52. kalkin It would be really nice if gtk and gnome would be more separate and they don't enforce UI decisions on the framework level, but on the DE level.
  53. kalkin While I really like Gnome Desktop and hate using KDE with passion, QT looks just more sane.
  54. kalkin May be the grass is greener on the other side
  55. arune I kind of get the feeling that all focus should currently be put in making the gtk3-version out
  56. tmolitor arune: lovetox and I do exactly that: focus only on the default branch...
  57. Link Mauve kalkin, having to write a few Qt-using programs, I can tell you GTK+ is much saner. :/
  58. tmolitor arune: we hope the nightly builds will bring more attention to the default branch and more help from other developers...
  59. kalkin Link Mauve: one of this days i will write my own ui framework with black jack & …
  60. Link Mauve Please no. D:
  61. kalkin also documentation will be in lojban, just for lulz :)
  62. mathieui remember to use ncurses as a backend
  63. mathieui in order to have sane code
  64. mathieui (not)
  65. kalkin mathieui: we could just abstract away the ui so it can be displayed in terminal in ncurses and also in X11 :D
  66. Link Mauve For example, it seems that today I can’t use any shortcut anymore in Qt programs…
  67. kalkin Link Mauve: Probably sun flares in aether because of the metheor showers this week.
  68. kalkin :)
  69. Link Mauve I’m going to bet on that instead: [2016-08-03 17:00] [ALPM] upgraded qt5-base (5.7.0-1 -> 5.7.0-2)
  70. kalkin Ok the gajim status icon doesn't work in Gnome at all
  71. kalkin If you close gajim there is NO WAY to reopen an already running instance.
  72. kalkin There is no icon for it in notification area and also no in the gnome shell menubar at the top
  73. kalkin (Centos 7)
  74. vorner Not that it would solve the bug, but there may be another way. gajim-remote toggle_roster_appearance
  75. kalkin vorner: awesome thanks
  76. kalkin so i just need something like this in my Gajim.desktop python ./src/ toggle_roster_appearance || ./
  77. Link Mauve kalkin, have you had a look at GtkApplication instead?
  78. kalkin Nop. I thought it's something different, will have a look at it
  79. kalkin Link Mauve: not sure if it solves _all_ my problems, but this is much more usefull.
  80. kalkin Link Mauve: thanks
  81. Link Mauve :)
  82. kalkin Hmm this adds an application menu. But i need something that implements the Tray Protocol Specification\
  83. kalkin Hmm this adds an application menu. But i need something that implements the Tray Protocol Specification
  84. Link Mauve Eww, X11 crap.
  85. Link Mauve kalkin, nobody is going to use that anymore.
  86. mathieui (“after 2025”)
  87. Link Mauve mathieui, I mean, at least GNOME, KDE and Unity have stopped using this.
  88. kalkin Link Mauve: what is the current way to add some tray icon to some panel without beeing specific if it's gnome shell or an xfcepanel? Is there at all?
  89. Link Mauve I don’t think the concept of tray icon or tray panel is wanted anymore by anyone.
  90. Link Mauve Even Microsoft has wanted to kill it ever since Windows 95.
  91. Link Mauve Then people copied the bad parts of it on other systems.
  92. kalkin Link Mauve: tray icon especially if it's at the top fits very good in our workflow
  93. Link Mauve It doesn’t exist anymore in modern DEs.
  94. kalkin Link Mauve: Have you ever used gnome shell networkmanager/shutdown/stuff menu in gnome?
  95. kalkin For me it's the same as a tray icon, at least from workflow
  96. Link Mauve Those are GNOME-Shell plugins, not tray icons.
  97. Link Mauve But no, I haven’t used them.
  98. mathieui Link Mauve, nm-applet is still a tray icon
  99. Link Mauve I highly doubt that it works that way in GNOME.
  100. mathieui I have a fbpanel with a tray on my desktop
  101. mathieui because ubuntu hates me
  102. kalkin Link Mauve: there is no difference between this kind of Gnome Shell plugin and a tray icon, at least at the workflow level
  103. Link Mauve I highly doubt it works that way in GNOME.
  104. Link Mauve kalkin, but there is at the development level.
  105. Link Mauve Also, it’s for user interaction, not for notifications like Gajim currently does.
  106. kalkin Link Mauve: also the Tray Protocol is ideal for something like Qubes. This allows us to have the rending, all the logick and other operations which can be exploited are done in a VM but the xfce-panel in dom0 can integrate the tray icon
  107. kalkin Link Mauve: also the Tray Protocol is ideal for something like Qubes. This allows us to have the rending, all the logic and other operations which can be exploited are done in a VM but the xfce-panel in dom0 can integrate the tray icon
  108. Link Mauve Wat, are you expecting people to use X11 alongside sandboxing?
  109. kalkin Link Mauve: QubesOS
  110. Link Mauve To give sandboxed programs full access to the X11 session?
  111. kalkin Link Mauve: I don't need to fear that my gajim reads the password I'm entering in some other vm
  112. kalkin We have X11 security
  113. kalkin like for real real, not for play play
  114. arune tmolitor: 👌👌👌
  115. Link Mauve kalkin, do you have a link? I’m very interested in that.
  116. Link Mauve Are you running a second Xorg server?
  117. kalkin Link Mauve: there is some old post from joanna
  118. kalkin Link Mauve: more details here
  119. kalkin Link Mauve: actually i have an Xserver in each of my domains and one in dom0
  120. kalkin Link Mauve: actually i have a Xserver in each of my domains and one in dom0
  121. zak Hi there. I always get several (up to about 4) messages for some contacts from a specific server: item-not-found
  122. zak Does anybody have an idea what this is?
  123. zak It's rather annoying because I always have to close this messages for every contact after login.
  124. Asterix zak, enable the XML logs before you connect to catch the XML that causes that
  125. zak Hi Asterix. Does this help? <message to="me@jabberserver" from="contact@jabberserver" id="me@jabberserver__contact@jabberserver__5J62h" type="error"> <event xmlns=""> <items node=""> <item id="0" node=""> <mood xmlns=""/> </item> </items> </event> <error code="404" type="continue"> <item-not-found xmlns="urn:ietf:params:xml:ns:xmpp-stanzas"/> </error> </message>
  126. Link Mauve zak, are you using an older version of Ejabberd?
  127. zak @Link Mauve: It may be that the specific server does, yes.
  128. zak No, actually, it seems to be Openfire 3.9.1
  129. arune Yuck! 👎
  130. lovetox you should search a good xmpp server
  131. zak Yes that server is not my main server, I just keep it for some old contacts (who are seldom online). I was just wondering about these messages.
  132. zak By the way, I think this horrible diversity of server (and client!) support for XMPP features is probably the biggest flaw of XMPP in general! The casual user just cannot be sure if something works, he is often confused why something does not work and there are so many cases where the answer just is "the server (or client) does not support it/is not configured for it".
  133. Link Mauve zak, I think it’s the biggest strength.
  134. Link Mauve If you had an unique implementation or an unique deployment, it would be impossible for people from the outside to make things move forward.
  135. Link Mauve Compare for example the terrible state of with modern servers.
  136. zak of course you should move things forward, bit there are just too many servers and clients who just stay where they are
  137. Link Mauve That’s still much better than if you had a single server or client that would stay where it was.
  138. zak Well, agreed. Then let me perhaps rephrase my critic: Too many (XMPP-)servers and clients just stay where they are.
  139. Link Mauve Their maintainers should either be given some help, or declare these abandonned. :)
  140. zak Almost all of my peers just use Pidgin for example. And development there is... well I guess the code is unmaintainable. I cannot explain this to me any way else.
  141. Link Mauve Pidgin people are actually quite active on their 3.0 release, but it takes a lot of time since they are switching a lot of their code to newer versions of their libraries.
  142. zak It's probably the most common open messenger on desktop and almost everytime I look at some ticket it just isn't happening anything there anymore.
  143. Link Mauve For example moving from gtk2 to gtk3 is a huge upgrade that takes very long to get to a satisfying situation.
  144. zak Yeah... I suppose.
  145. Link Mauve But yeah, their support of XMPP is… sub-par.
  146. zak And after they are finished, gtk4 is live.
  147. Link Mauve Heh, GNOME people decided at their latest hackfest to release gtk4 pretty soon. :p
  148. zak You see my point
  149. Link Mauve Yes.
  150. Link Mauve But that’s what you get with a critically low amount of developers with free time. :/
  151. zak To me it seems that some things are developed with some "vision" to a specific version, and at the time where this is almost completed, they start all over, leaving the fine working situation as it is.
  152. zak ...meaning all the developers who learned and accepted this piece of software need to learn again, adapt their code (again) and so on
  153. zak Sometimes I think, for some developers, software is just a way for them to try something out, then throw it away and try something different ;)
  154. Link Mauve Maintainers don’t break APIs for the sake of it.
  155. zak No, they think it is bad and needs to be redone.
  156. Link Mauve gtk2 to gtk3 was a necessary step.
  157. zak Well I don't know. You have a concrete example?
  158. Link Mauve It was mostly an update on top of the existing code, nothing that didn’t need to has been broken during it.
  159. zak Then why do all the gtk-applications need to be redone for gtk3?
  160. Link Mauve Hmm, off the top of my head, the low-level rendering code (GDK) needed to be changed a lot to remove some X11 assumptions, and to allow for better ports to Wayland, OS X, Windows, even to the web.
  161. Link Mauve They don’t need to be redone, the user-facing API is still pretty similar.
  162. zak Then I don't understand why I so often read "no with gtk3 support" or something like this
  163. Link Mauve In Gajim’s case it was decided to also move to python3 and gobject-introspection and gstreamer1.0 at the same time.
  164. zak s/no/now/
  165. Link Mauve Because even if the API is similar, it’s not identical, and it rarely makes sense to try to support both 2.x and 3.x.
  166. zak exactly
  167. Link Mauve Especially in Gajim’s case which changed its language and bindings at the same time.
  168. Link Mauve So the conventions changed.
  169. Link Mauve It’s maybe unfortunate, I don’t know.
  170. zak I admit that in some way these changes are important, but it seems to me that it is happening too fast.
  171. Link Mauve But it was done in the users’ and maintainers’ best interest again.
  172. Link Mauve I’d say the opposite. :p
  173. Link Mauve It’s been happening way too slowly, if they started porting at the beginning of the 3.x branch this would be a story from the past right now.
  174. zak who do you mean "they"?
  175. zak pidgin
  176. zak ?
  177. Link Mauve Pidgin, Gajim.
  178. Link Mauve They are both in the same boat here.
  179. zak Well I wouldn't jump to something like gtk3 right after it has been released. Usually there is so much afterwards happening
  180. zak Then you write the code, and 6 month later there is gtk3.1 and some new functions are introduced which helps for other stuff that then would have been done differently
  181. zak At a developers perspective I agree with you to some point.
  182. zak But in the end, the users are the ones who just wait that the base code is adapted to some new framework while other important features are just stalled
  183. Link Mauve But that’s what’s happening right now. :)
  184. zak Exactly :)
  185. zak But don't get me wrong, I appreciate all the work!