Accueil » Bryce Harrington » Ressources » PasteBin » Release meeting for Inkscape 0.92 - May 6, 2016
  1. May 06 13:11:30 <bryce> ================== Inkscape Release 0.92 Meeting ==================
  2. May 06 13:11:41 <bryce> welcome everyone! ;-)
  3. May 06 13:11:46 <ScislaC> :)
  4. May 06 13:11:53 <Mc-> ^^
  5. May 06 13:11:54 <jabiertxof> !!!
  6. May 06 13:12:12 <Tavmjong> tweenk: There seem to be some performance problems after the GTK3 rendering fix. Maybe su_v can sum up.
  7. May 06 13:12:18 <bryce> ScislaC, thanks for sending me the release process. How many times now have we kicked off 0.92 release? ;-)
  8. May 06 13:12:32 <bryce> I'm going to update that and re-send hopefully today
  9. May 06 13:12:44 <Tavmjong> Tavmjong: This should probably be a blocker.
  10. May 06 13:13:06 * su_v refers to the report mentioned on the mailing list by jazzynico: !bug 1571192
  11. May 06 13:13:17 <Mc-> >> 1571192
  12. May 06 13:13:18 <InkBot> Bug #1571192: "Severe performance regression"
  13. May 06 13:13:18 <InkBot> https://launchpad.net/bugs/1571192
  14. May 06 13:13:55 <su_v> I can share a few details not mentioned in the report about how it exposes to local builds
  15. May 06 13:14:24 <ScislaC> I'm going to be going through the tracker over the next week looking for blockers
  16. May 06 13:14:31 <su_v> specifically, it is related to real-time updates in the GUI, mostly the spinboxes (x, y, w, h) of the select tool controls bar
  17. May 06 13:14:44 <su_v> the severity can change depending on the gtk2 theme (or theme engine)
  18. May 06 13:14:57 <JonCruz> How are we on the Windows front? got this box up to being productive now
  19. May 06 13:15:08 <su_v> Adwaita GTK2 is rather bad, stock Raleigh on the other hand is quite fast
  20. May 06 13:16:04 <su_v> with Quartz backend, basically all parts of the GUI interfere with actions on-canvas (status bar, controls bar, rulers, even the scrollbars)
  21. May 06 13:16:25 <su_v> not relevant though (Quartz is of interest for future GTK3)
  22. May 06 13:16:47 <bryce> JonCruz, it would be great to get cmake working on Windows
  23. May 06 13:16:52 <Tavmjong> It's puzzling that the theme can have such a big impact.
  24. May 06 13:16:58 <su_v> tweenk: FailBit discussed with you certain aspects of changing priorities
  25. May 06 13:16:59 <Mc-> bryce: I think we're close to that
  26. May 06 13:17:22 <bryce> Mc-, yeah I was glad to see that got worked on a bit at the hackfest, I hope we're close
  27. May 06 13:17:29 <Mc-> (thanks to the folks at semiodesk)
  28. May 06 13:17:40 <JonCruz> bryce, I'd gotten a lot of magic working from single-lines via the proper file extension variables
  29. May 06 13:17:53 <su_v> G_PRIORITY_DEFAULT_IDLE vs G_PRIORITY_HIGH_IDLE
  30. May 06 13:17:59 <JonCruz> hard for people who work on platforms where those are always empty strings
  31. May 06 13:19:16 <bryce> there's also some bits in the cmake file which were purportedly for windows related stuff but I have no idea if they're still relevant. cleanup of stuff marked for win32 would be appreciated
  32. May 06 13:20:01 <bryce> ok so we have 1571192; any other bugs that need top priority?
  33. May 06 13:20:20 <ScislaC> JonCruz: https://bugs.launchpad.net/inkscape/+bug/1416674 Since you're a good "resources" guy, maybe you can figure out how to fix this. It's a blocker for 0.92, and I would love to get a fix backported for 0.91.1
  34. May 06 13:20:20 <InkBot> Bug #1416674: "0.91 doesn't show fonts loaded by font manager"
  35. May 06 13:20:21 <InkBot> https://launchpad.net/bugs/1416674
  36. May 06 13:21:01 <JonCruz> did ScislaC just say the "f"-word?
  37. May 06 13:21:03 <JonCruz> "fonts"
  38. May 06 13:21:12 <bryce> I'm going to encourage ScislaC to be very stringent for what he includes on his list. I want to avoid getting us mired down.
  39. May 06 13:21:31 <ScislaC> JonCruz: our crafters and design folk are really hurting on this issue
  40. May 06 13:21:39 <JonCruz> Hmmm... looks like my lunchime wifi access has aged me out.
  41. May 06 13:21:46 <Mc-> the other blocken mentionned on the ml was >> 1389723
  42. May 06 13:21:47 <InkBot> Bug #1389723: "Documents created before experimental branch merge get scaled by 90/96"
  43. May 06 13:21:47 <InkBot> https://launchpad.net/bugs/1389723
  44. May 06 13:21:58 <bryce> I'm going to try to package up an alpha this weekend, monday at the latest.
  45. May 06 13:21:59 <Mc-> blocker*
  46. May 06 13:22:20 <JonCruz> l8r. see people later tonight
  47. May 06 13:22:26 <doctormono> thanks JonCruz
  48. May 06 13:22:27 <bryce> I'm also going to try and make an effort to post pre-releases on a regular cadence, maybe every week or two
  49. May 06 13:22:36 <bryce> JonCruz, cya! thanks for coming!!
  50. May 06 13:22:42 <su_v> ScislaC: re blockers, and other regressions in 0.91 of matter or interest for a next stable release: I'm not yet at a point to resume the same level of activity in the bug tracker; I'd be interested to resume partially, i.e. in teamwork with release manager with focus on release-relevant reports.
  51. May 06 13:22:57 <su_v> s/i.e./e.g./
  52. May 06 13:23:24 <doctormono> i.e. is probably right there.
  53. May 06 13:23:24 <tweenk> Hmm. So it looks like we might need two rendering methods, one for GTK2 and another for GTK3. I wonder why scrollbars affect performance though
  54. May 06 13:23:47 <Tavmjong> Re 1389723, I need test files and more details to work with.
  55. May 06 13:23:56 <doctormono> It's number of events, widgets must be putting out more events.
  56. May 06 13:24:05 <DASPRiD> Mc-, sup?
  57. May 06 13:24:07 <su_v> tweenk: only with Quartz backend, and there they are the components which contribute the least to the lag
  58. May 06 13:24:09 <Mc-> tweenk: in some circumstances their length get recomputed a LOT
  59. May 06 13:24:37 <tweenk> DASPRiD: you made a change to lpe-jointype.cpp, are you OK with tat being GPL v2+?
  60. May 06 13:24:41 <tweenk> *that
  61. May 06 13:24:57 <su_v> with quartz, the only really usable state right now is with all GUI hidden (Shift+F11), performance-wise. but again - that's a special case not relevant for 0.93
  62. May 06 13:25:01 <su_v> 0.92*
  63. May 06 13:25:28 <doctormono> My time here at the café is coming to a close.
  64. May 06 13:25:28 <DASPRiD> tweenk, tweenk i don't mind
  65. May 06 13:25:38 <tweenk> DASPRiD: cool!
  66. May 06 13:25:38 <doctormono> Thanks everyone, great meeting.
  67. May 06 13:25:57 <DASPRiD> what is this about anyway? :)
  68. May 06 13:26:07 <crogers> hehehe
  69. May 06 13:26:32 <tweenk> DASPRiD: that file is one of 2 files in the entire codebase which are GPL v2 only and we want to change to GPLv2+ and potentially GPLv3+ in the future
  70. May 06 13:26:50 <DASPRiD> ah, i see :)
  71. May 06 13:27:04 <bryce> doctormono, thanks!
  72. May 06 13:27:26 <bryce> su_v, :-)
  73. May 06 13:27:34 <DASPRiD> tweenk, i suppose it would be a dick-move to decline that, when one only contributed a single line to it anyway ;)
  74. May 06 13:28:04 <bryce> DASPRiD, thanks :-0
  75. May 06 13:28:06 <bryce> :-)
  76. May 06 13:28:48 <bryce> guessing ScislaC is tied up with something
  77. May 06 13:28:59 <ScislaC> Yo, what's up?
  78. May 06 13:29:05 <ScislaC> Did I miss a question?
  79. May 06 13:29:22 <bryce> I've noted 1571192, 1416674, and 1389723 as nominated for ScislaC's list
  80. May 06 13:29:52 <bryce> ScislaC, oh su_v made an offer to assist us with release bug tracking.
  81. May 06 13:30:11 <jabiertxof> :)
  82. May 06 13:30:28 <ScislaC> those are all marked
  83. May 06 13:30:32 <bryce> su_v, that's great, thanks
  84. May 06 13:30:36 <bryce> ok cool
  85. May 06 13:30:44 <su_v> tweenk: btw, here's a minor regression related to the renderer refactoring (in case you missed the notification from the bug tracker): !bug 1572189
  86. May 06 13:30:44 <InkBot> Bug #1572189: "Outline view of clipped path not fully cleared in normal view mode (rev >= 14821)"
  87. May 06 13:30:45 <InkBot> https://launchpad.net/bugs/1572189
  88. May 06 13:30:56 <ScislaC> any help from su_v is always welcomed by me :)
  89. May 06 13:31:17 <bryce> oh one other topic to discuss relating to the release is whether to fork an experimental branch
  90. May 06 13:31:38 <su_v> (I don' recall at the moment whether masking is similarly affected - I tested, but forgot to take notes :/ )
  91. May 06 13:31:50 <crogers> doctormo is having connection probs, btw.
  92. May 06 13:31:55 <bryce> I'm open either way. It seemed to do the trick last release but I know there's pros and cons
  93. May 06 13:32:29 <bryce> another option is just to let folks develop on their own side branches, and then postpone integration until we've forked the release branch
  94. May 06 13:32:51 <bryce> anyone have opinions on this?
  95. May 06 13:33:06 <tweenk> su_v: looks like incorrect dirty area, should be simple to fix
  96. May 06 13:33:37 <tweenk> su_V: I'll try to work on those during the weekend
  97. May 06 13:33:38 <su_v> tweenk: thx, it's really minor, but would be nice to have fixed
  98. May 06 13:33:55 <ScislaC> bryce: I think side branches and staggered merges are a better approach if we commit to pushing the release through as quickly as we can so there isn't a huge amount of time people would need to manage branches.
  99. May 06 13:34:02 <bryce> the one other topic was when to start the feature freeze
  100. May 06 13:34:10 <tweenk> su_v: the second one should be easy but the first one will depend on whether I can get Inkscape to compile on my Windows machine
  101. May 06 13:34:13 <bryce> ScislaC, yeah
  102. May 06 13:34:38 <su_v> bryce: I think experimental would have made more sense before the hackfest (e.g. for the GTK3-related stuff, which affects GTK2 for next stable)
  103. May 06 13:34:39 <bryce> ScislaC, I suppose we can do that, and always create an experimental branch later if/when necessary
  104. May 06 13:34:45 <tweenk> su_v: I recently received a Mac work laptop so eventually I may be able to help with OSX problems too
  105. May 06 13:35:01 <Tavmjong> Feature freeze in a month?
  106. May 06 13:35:11 <bryce> su_v, has anything landed recently that is especially unstable?
  107. May 06 13:35:12 <ScislaC> I am not opposed to an experimental branch though if we know we'll be locked out of trunk for long... mainly because it helps get stuff tested.
  108. May 06 13:35:44 <Mc-> I'm not convinced by the need for experimental branch
  109. May 06 13:35:45 <FailBit> and pushed through official channels
  110. May 06 13:35:53 <su_v> bryce: I'm not in a position to determine that on a global state of current trunk
  111. May 06 13:35:59 <bryce> mm ok
  112. May 06 13:36:19 <Tavmjong> Let's wait on forking an experimental branch until someone has something compelling to put into it.
  113. May 06 13:36:31 <su_v> I do think that the changes for the renderer under GTK3 might be a bit unfortunate so close to planning to cut a stable release based on gtk2
  114. May 06 13:36:35 <bryce> Tavmjong, do you feel we need to give it a month? Is there feature work currently under way? I was wondering if maybe we could do a week or two?
  115. May 06 13:37:07 <bryce> su_v, I'm wondering if we should consider rolling back any recent bits like that
  116. May 06 13:37:17 <su_v> bryce: I will resume more testing once the performance of unchanged trunk is better again with GTK2
  117. May 06 13:37:27 <Tavmjong> Could be sooner, but we should announce our intention on inkscape-devel and get some feedback.
  118. May 06 13:37:43 <bryce> Tavmjong, ok that seems fair. I'll do that.
  119. May 06 13:38:25 <bryce> maybe we can propose 1 week, and if there's pushback go to a longer period
  120. May 06 13:38:36 <ScislaC> bryce: jabiertxof has some things he wanted reviewed if I recall correctly
  121. May 06 13:39:01 <jabiertxof> yes
  122. May 06 13:39:06 <bryce> ScislaC, is there ever a time that that isn't the case? ;-)
  123. May 06 13:39:16 <bryce> jabiertxof :-)
  124. May 06 13:39:17 <ScislaC> lol
  125. May 06 13:39:29 <jabiertxof> :)
  126. May 06 13:39:34 <ScislaC> bryce: it's one of those things that I like about him :)
  127. May 06 13:39:39 <bryce> but yeah let's help him wrap up current efforts
  128. May 06 13:40:24 <bryce> jabiertxof, can you email me a list of links to items you need reviewed? I probably won't review them myself but would like to keep track, and help with finding reviewers if needed.
  129. May 06 13:40:56 <jabiertxof> tweenk is going to review some of this, I send to both and doctormon the list
  130. May 06 13:41:13 <jabiertxof> also there is a item to discuss in a patch
  131. May 06 13:41:21 <Tavmjong> I'm going to have to sign off.... great discussions!
  132. May 06 13:41:32 <bryce> jabiertxof, and only items that you feel are ready for release; if you have items that are still work in progress / experimental maybe hang onto those a bit until post-release
  133. May 06 13:41:33 <jabiertxof> is a ScislaC feature of convert item to paths
  134. May 06 13:41:47 <bryce> Tavmjong, thanks!
  135. May 06 13:41:49 <jabiertxof> ok bryce
  136. May 06 13:41:56 <ScislaC> thanks Tavmjong!
  137. May 06 13:42:00 <tweenk> I'll have to sign off to - back to work :)
  138. May 06 13:42:04 <tweenk> *too
  139. May 06 13:42:04 <jabiertxof> Bye Tavmjong
  140. May 06 13:42:14 <bryce> alright, I think we've covered enough, shall we close here?
  141. May 06 13:42:15 <Tavmjong> Good night all!
  142. May 06 13:42:17 <jabiertxof> bye tweenk
  143. May 06 13:42:27 <tweenk> Bye everyone & thanks
  144. May 06 13:42:34 <bryce> if anyone has other release bits to discuss, ping me or post to the mailing list thread
  145. May 06 13:42:57 <bryce> I don't know if we need another release meeting, but can decide that later.
  146. May 06 13:43:12 <bryce> ===================== End of Meeting ====================
  147. May 06 13:43:13 <ScislaC> bryce: any thoughts on a name for this? https://bugs.launchpad.net/inkscape/+bug/1556592
  148. May 06 13:43:14 <InkBot> Bug #1556592: "Stroke to fill retaining fills"
  149. May 06 13:43:14 <InkBot> https://launchpad.net/bugs/1556592
  150. May 06 13:43:31 <jabiertxof> ok I drop you some lines thanks bryce and ScislaC
  151. May 06 13:43:52 <jabiertxof> thanks ScislaC :)
  152. May 06 13:44:05 <bryce> ScislaC, hmm, not offhand
  153. May 06 13:44:40 <ScislaC> bryce: it converts strokes, fills, markers, etc to separate paths and groups them. Look at my last comment in the report to see if you think either of those looks decent.
  154. May 06 13:44:53 <jabiertxof> also retain paint order
  155. May 06 13:44:53 <bryce> ScislaC, idly I wonder if feature naming might be a thing the new UI group might be able to weigh in on?
  156. May 06 13:45:23 <bryce> Deconstruct object sounds pretty cool
  157. May 06 13:45:42 <ScislaC> Do they have their own mailing list or irc channel or is everything discussed in the dev places?
  158. May 06 13:45:46 <jabiertxof> and the old stroke to path is removed?
  159. May 06 13:46:05 <jabiertxof> #inkscape-ui
  160. May 06 13:46:15 <jabiertxof> but not so much populated
  161. May 06 13:46:29 <jabiertxof> mailing list with [ui]
  162. May 06 13:46:32 <ScislaC> jabiertxof: I think we can leave it, this is technically a more complete feature, but no need to surprise people if they're used to the old workflow.
  163. May 06 13:46:39 <bryce> ScislaC, also see https://inkscape.org/en/paste/9553/ for their charter/scope
  164. May 06 13:47:22 <jabiertxof> Ok ScislaC
  165. May 06 13:47:37 <ScislaC> thanks jabiertxof and bryce
  166. May 06 13:47:48 <jabiertxof> you are welcome!
 
 

38

 

796

Release meeting for Inkscape 0.92 - May 6, 2016

par Bryce Harrington

-

PasteBin

Lignes
166
Mots
2448
Taille
14,4 Kio
Créé le
Type
text/x-irclog
Creative Commons Attribution Share-Alike 3.0 (CC-BY-SA 3.0)
Connectez-vous pour ajouter un commentaire !

Inkscape est un logiciel libre et open source sous licence GPL.