/irc-logs / mozilla / #accessibility / 2014-02-26 / end

Options:

  1. # Session Start: Wed Feb 26 00:00:01 2014
  2. # Session Ident: #accessibility
  3. # [00:02] * Quits: Justin_o_ (uid14648@moz-E77DEB21.irccloud.com) (Quit: Connection closed for inactivity)
  4. # [00:02] <@tbsaunde> cabanier: mach mochitest-a11y I think
  5. # [00:02] <cabanier> tbsaunde: yeah but I wanted to debug at the same time
  6. # [00:03] <cabanier> tbsaunde: I guess I can attach
  7. # [00:06] <@tbsaunde> cabanier: I think passing -d gets you a debugger
  8. # [00:07] <@tbsaunde> or maybe -d gdb
  9. # [00:07] <cabanier> tbsaunde: ok. with XCode I can tell it to attach as soon as it launches so that should work too
  10. # [00:28] * Quits: Gijs (gijs@moz-45AC6916.range86-164.btcentralplus.com) (Quit: sleep)
  11. # [00:30] * Joins: kushagra (kushagra@C620A800.DF268AF7.520CDC98.IP)
  12. # [00:32] * Quits: kushagra (kushagra@C620A800.DF268AF7.520CDC98.IP) (Ping timeout)
  13. # [00:42] <@surkov> cabanier: you can attach the debugger from Xcode UI if add --no-autorun
  14. # [00:43] <@surkov> tbsaunde: what a review status of bug 969532 is?
  15. # [00:44] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=969532 nor, --, ---, surkov.alexander, ASSI, add TextRange stub
  16. # [00:45] <cabanier> surkov: ah. nice!
  17. # [00:46] <cabanier> surkov: having xcode attach automatically wasn't working
  18. # [00:46] <@surkov> ok, glad it was helpful
  19. # [01:09] <cabanier> surkov: is there a way to set up a11y on the mac so it queries firefox about the region that the mouse is over?
  20. # [01:10] <@surkov> cabanier: accessibility inspector
  21. # [01:11] <cabanier> surkov: thanks!
  22. # [01:12] <@surkov> yw!
  23. # [01:13] <@tbsaunde> surkov: I'm waiting for a patch that doesn't have useless refcounting
  24. # [01:14] <@surkov> tbsaunde: I'm still not sure decent idea, can you find couple good examples in Gecko for me pls?
  25. # [01:16] <cabanier> surkov: does the inspector call into Accessible::ChildAtPoint?
  26. # [01:16] <@surkov> cabanier: it should I think
  27. # [01:16] <cabanier> surkov: hmm
  28. # [01:17] <@surkov> cabanier: doesn't seem so?
  29. # [01:17] <cabanier> surkov: no. Maybe I'm not using it correctly
  30. # [01:18] <@tbsaunde> surkov: I'd be fine if you just passed reference to already existing TextRange into methods too
  31. # [01:18] <@tbsaunde> surkov: but I'm not really sure what you want, other than some people who use nsAutoPtr
  32. # [01:19] <@surkov> tbsaunde: I didn't see a lot of use cases of nsAutoPtr to pass/return objects and it didn't look nice
  33. # [01:20] <@surkov> tbsaunde: I didn't get first idea, put some details into bug probably?
  34. # [01:23] <@tbsaunde> surkov: about 5 seconds with grep find nsISHEntry::forgetEditorData
  35. # [01:25] <@surkov> tbsaunde: do you want to see all method prefixed by 'forget' word?
  36. # [01:26] <@surkov> tbsaunde: copied of the arrays of auto pointers also might be not so evident as refptred stuff
  37. # [01:27] <@surkov> copied->copiing
  38. # [01:28] <@tbsaunde> surkov: given that you just created the object I don't think it should start with forget
  39. # [01:28] <@tbsaunde> and why are you copying arrays in the first place? that's slow
  40. # [01:29] <@surkov> tbsaunde: xcpom and internal use different arrays of different objects
  41. # [01:30] <@surkov> tbsaunde: that feels sounds kind of dangerous because semantically we don't have a difference between say Accessible* ChildAtPoint() and TextRange* RangeFromChild()
  42. # [01:31] <@tbsaunde> surkov: sure, it would be nice if someone made returning nsAutoPtr work
  43. # [01:32] <@tbsaunde> surkov: but refcounting stuff when it has only one owner seems worse
  44. # [01:32] <@surkov> tbsaunde: maybe we could go with refptr approach for now?
  45. # [01:33] <@tbsaunde> I'd really rather you did something else
  46. # [01:33] <@surkov> tbsaunde: I agree but nothing easy and nice to replace
  47. # [01:33] <@tbsaunde> I'd say there are several reasonably nice things
  48. # [01:33] <@surkov> tbsaunde: after all reftpr approach shouldn't hit us
  49. # [01:33] <@tbsaunde> so "no" I think
  50. # [01:38] * Quits: victorporof (victorporo@moz-FFB76956.cpe.net.cable.rogers.com) (Quit: victorporof)
  51. # [01:38] * Joins: victorporof (victorporo@moz-FFB76956.cpe.net.cable.rogers.com)
  52. # [01:40] <@surkov> tbsaunde: are you suggesting to follow "T* getSomething" approach and return a raw memory object hoping that called will free memory or something else?
  53. # [01:42] <@tbsaunde> surkov: returning T* is probably what I'd do for now
  54. # [01:42] <@tbsaunde> but passing in T& seems pretty reasonable
  55. # [01:44] <cabanier> surkov: it seems that Accessible::ChildAtPoint is already working
  56. # [01:45] <@surkov> cabanier: awesome
  57. # [01:45] <cabanier> surkov: :-)
  58. # [01:45] <@surkov> cabanier: but what do you mean by already?
  59. # [01:45] <@surkov> you mean you have a patch?
  60. # [01:45] <cabanier> surkov: it calls into the code that gets the bounds
  61. # [01:45] <cabanier> surkov: and that code returns the hit region
  62. # [01:45] <@surkov> ah, interesting
  63. # [01:46] <@surkov> cabanier: so we need a test
  64. # [01:46] <cabanier> > xul.dll!mozilla::a11y::Accessible::GetBoundsRect(nsRect & aTotalBounds, nsIFrame * * aBoundingFrame) Line 928 C++
  65. # [01:46] <cabanier> xul.dll!mozilla::a11y::Accessible::GetBounds(int * aX, int * aY, int * aWidth, int * aHeight) Line 976 C++
  66. # [01:46] <cabanier> xul.dll!mozilla::a11y::Accessible::ChildAtPoint(int aX, int aY, mozilla::a11y::Accessible::EWhichChildAtPoint aWhichChild) Line 876 C++
  67. # [01:46] <cabanier> xul.dll!mozilla::a11y::AccessibleWrap::accHitTest(long xLeft, long yTop, tagVARIANT * pvarChild) Line 972 C++
  68. # [01:47] <cabanier> should I look at accessible/test/mochitest/hittest and go from there?
  69. # [01:49] <@surkov> cabanier: that's right, it'd be good to test nested hit regions
  70. # [01:49] <@surkov> tbsaunde: I don't like T* approach since it's potentially memory leaky, T& thing is sort of semantically strange since GetTextRange(TextRange&) assumes TextRange is created by caller and that Get just change the given TextRange
  71. # [01:49] <cabanier> surkov: ok, I will look at the mochitest and let you know if I have questions
  72. # [01:50] <@surkov> tbsaunde: let's save that cleaning up issue for future
  73. # [01:50] <@tbsaunde> surkov: lets not add bad in the first place
  74. # [01:50] <@surkov> cabanier: thanks!
  75. # [01:50] <@tbsaunde> surkov: we already have other things that take object to fill data in on
  76. # [01:50] <@surkov> tbsaunde: I would agree if I was convinced we have nicer alternative
  77. # [01:51] <cabanier> surkov: the test should have a canvas with a couple of fallback elements with hit regions. It should query a couple of positions and check if the correct info is returned.
  78. # [01:51] <@tbsaunde> I might agree if I wasn't reasonably happy with what I suggested
  79. # [01:53] <@surkov> Trev, but I'm not :) I'm pretty fine to have a follow up to find a nicer solution but I wouldn't happy if the issue delayed further UIA work
  80. # [01:54] <@tbsaunde> then just do one of the things I suggested ;)
  81. # [01:55] <@surkov> yeah, but I wouldn't feel happy either
  82. # [01:55] <@surkov> you know I always fix the patch when I totally agree with suggestion
  83. # [01:55] <@surkov> but when it doesn't look so clear with me then I prefer to deal with it later
  84. # [01:56] <@surkov> does it fair? :)
  85. # [01:56] <@surkov> does -> is
  86. # [01:58] <@tbsaunde> I think it'll be much harder to take refcounting things out than it will be to not add it
  87. # [01:59] <@tbsaunde> is returning T* really that big a deal? just always assign return value to auto ptr
  88. # [02:02] <@surkov> tbsaunde: it's not wide practice in a11y (and possibly Gecko) but the caller must know that this object is not refcounted while others are, so I would go with approach
  89. # [02:03] <@surkov> tbsaunde: it shouldn't be really harder to change refptr to something else because primarily I plan to focus on TextRange implementation rather than it's usage
  90. # [02:05] <@tbsaunde> surkov: that's what documentation is for
  91. # [02:05] <@surkov> tbsaunde: document is good but that's some extra, people don't always read and just follow common used practice
  92. # [02:05] <@surkov> common -> commonly
  93. # [02:07] <@tbsaunde> surkov: RelationByType is actually another example, I'm pretty sure if you do nothing with its return value we leak
  94. # [02:07] <@tbsaunde> besides just add MOZ_COUNT_CTOR / DTOR and then testing will help
  95. # [02:09] <@surkov> tbsaunde: I like RelationByType but it's protected from mistakes, testing is really good but it's not always 100% protection as it's human dependent
  96. # [02:10] <@tbsaunde> surkov: how is it protected?
  97. # [02:12] <@surkov> tbsaunde: I don't really remember all magic we had but I think you don't need to care about return value
  98. # [02:14] <@tbsaunde> surkov: well anyway if you ignore return value that's clearly crazy since that's whole point of method
  99. # [02:14] <@surkov> agree
  100. # [02:15] <@tbsaunde> surkov: if it'll make you happier I'll fix returning nsAutoPtr
  101. # [02:16] <@tbsaunde> but I think the current patch is still closer to r- land
  102. # [02:16] <@surkov> tbsaunde: It would make me happier of course, closer to r- because of refptr stuff?
  103. # [02:21] <@tbsaunde> surkov: yes
  104. # [02:24] <@surkov> tbsaunde: it's not even design issue, it might be cleaning up issue, why are you so hard on it?
  105. # [02:25] <@tbsaunde> surkov: how is it not a design issue?
  106. # [02:26] <@surkov> tbsaunde: it's sort of small in the project to be a designing issue
  107. # [02:27] <@surkov> (my laptop track pad seems go crazy preventing me from working)
  108. # [02:29] <@tbsaunde> surkov: I think how lifetimes are determined is pretty important
  109. # [02:29] <@tbsaunde> anyway /me -> home
  110. # [02:31] * Joins: kushagra (kushagra@C620A800.DF268AF7.520CDC98.IP)
  111. # [02:32] <@surkov> tbsaunde: ok, let's continue tomorrow
  112. # [02:33] * Quits: @surkov (surkov@moz-DF24A6EA.cpe.pppoe.ca) (Quit: surkov)
  113. # [02:33] * Quits: kushagra (kushagra@C620A800.DF268AF7.520CDC98.IP) (Ping timeout)
  114. # [02:46] * khuey is now known as khuey|away
  115. # [03:21] <@firebot> New Core - Disability Access APIs bug 976895 filed by kwierso@gmail.com.
  116. # [03:45] * khuey|away is now known as khuey
  117. # [03:50] * Joins: alinah (Adium@E521DFFF.DDFDBF93.E30FDA0B.IP)
  118. # [03:52] * Quits: alinah (Adium@E521DFFF.DDFDBF93.E30FDA0B.IP) (Ping timeout)
  119. # [04:23] * Quits: maxli (maxli@moz-EE42E0E.student.cs.uwaterloo.ca) (Ping timeout)
  120. # [04:33] * Joins: davidb (davidb@moz-F8A6F0DC.dsl.bell.ca)
  121. # [04:33] * ChanServ sets mode: +qo davidb davidb
  122. # [04:33] * Quits: @davidb (davidb@moz-F8A6F0DC.dsl.bell.ca) (Quit: blast off!)
  123. # [04:51] * icaaq|afk is now known as icaaq
  124. # [04:52] * Joins: maxli (maxli@moz-F47DD19B.student.cs.uwaterloo.ca)
  125. # [05:00] * Joins: sawrubh (uid6719@moz-5F4AA75A.irccloud.com)
  126. # [05:06] * icaaq is now known as icaaq|afk
  127. # [05:19] * Quits: victorporof (victorporo@moz-FFB76956.cpe.net.cable.rogers.com) (Quit: victorporof)
  128. # [05:24] * Joins: yzen (yzen@moz-A36A7FD4.cpe.pppoe.ca)
  129. # [05:24] * ChanServ sets mode: +o yzen
  130. # [05:25] * Quits: maxli (maxli@moz-F47DD19B.student.cs.uwaterloo.ca) (Input/output error)
  131. # [05:25] * Joins: victorporof (victorporo@moz-FFB76956.cpe.net.cable.rogers.com)
  132. # [05:25] * khuey is now known as khuey|away
  133. # [05:26] * Joins: maxli (maxli@moz-F47DD19B.student.cs.uwaterloo.ca)
  134. # [05:38] * Quits: maxli (maxli@moz-F47DD19B.student.cs.uwaterloo.ca) (Input/output error)
  135. # [05:40] * Joins: maxli (maxli@moz-F47DD19B.student.cs.uwaterloo.ca)
  136. # [06:32] * Quits: maxli (maxli@moz-F47DD19B.student.cs.uwaterloo.ca) (Input/output error)
  137. # [06:32] * Joins: maxli (maxli@moz-B19F68ED.student.cs.uwaterloo.ca)
  138. # [06:46] * Quits: maxli (maxli@moz-B19F68ED.student.cs.uwaterloo.ca) (Input/output error)
  139. # [06:48] * Joins: maxli (maxli@moz-B19F68ED.student.cs.uwaterloo.ca)
  140. # [06:54] * Quits: maxli (maxli@moz-B19F68ED.student.cs.uwaterloo.ca) (Ping timeout)
  141. # [06:57] * Joins: maxli (maxli@moz-EE42E0E.student.cs.uwaterloo.ca)
  142. # [07:04] * Parts: sawrubh (uid6719@moz-5F4AA75A.irccloud.com)
  143. # [07:31] * icaaq|afk is now known as icaaq
  144. # [07:42] * Quits: @yzen (yzen@moz-A36A7FD4.cpe.pppoe.ca) (Ping timeout)
  145. # [07:57] * Joins: kushagra (kushagra@E76FFEEC.1E9542C7.6EAD5CF4.IP)
  146. # [07:59] * Quits: kushagra (kushagra@E76FFEEC.1E9542C7.6EAD5CF4.IP) (Ping timeout)
  147. # [07:59] * Joins: kushagra (kushagra@E76FFEEC.1E9542C7.6EAD5CF4.IP)
  148. # [08:01] * Quits: kushagra (kushagra@E76FFEEC.1E9542C7.6EAD5CF4.IP) (Ping timeout)
  149. # [08:03] * Joins: kushagra (kushagra@E76FFEEC.1E9542C7.6EAD5CF4.IP)
  150. # [08:06] * Joins: kushagra_ (kushagra@E76FFEEC.1E9542C7.6EAD5CF4.IP)
  151. # [08:06] * Quits: kushagra (kushagra@E76FFEEC.1E9542C7.6EAD5CF4.IP) (Connection reset by peer)
  152. # [08:08] * Quits: kushagra_ (kushagra@E76FFEEC.1E9542C7.6EAD5CF4.IP) (Client exited)
  153. # [08:44] * Quits: maxli (maxli@moz-EE42E0E.student.cs.uwaterloo.ca) (Ping timeout)
  154. # [08:45] * Joins: victorporof_ (victorporo@moz-FFB76956.cpe.net.cable.rogers.com)
  155. # [08:45] * Quits: victorporof (victorporo@moz-FFB76956.cpe.net.cable.rogers.com) (Ping timeout)
  156. # [08:59] * Joins: maxli (maxli@moz-B19F68ED.student.cs.uwaterloo.ca)
  157. # [09:00] * Quits: maxli (maxli@moz-B19F68ED.student.cs.uwaterloo.ca) (Ping timeout)
  158. # [10:11] * Joins: agibson (agibson@moz-2C643250.gate.cable.virginm.net)
  159. # [10:20] * Quits: victorporof_ (victorporo@moz-FFB76956.cpe.net.cable.rogers.com) (Quit: victorporof_)
  160. # [10:42] * Joins: Gijs (gijs@moz-45AC6916.range86-164.btcentralplus.com)
  161. # [10:49] * Joins: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net)
  162. # [11:26] * Joins: ginn (ginn@moz-15DF0CD.oracle.com)
  163. # [11:52] * Quits: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net) (Quit: ChatZilla 0.9.90.1 [Firefox 27.0.1/20140212131424])
  164. # [12:28] * agibson is now known as agibson|afk
  165. # [12:45] * Quits: ginn (ginn@moz-15DF0CD.oracle.com) (Client exited)
  166. # [13:29] * Joins: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net)
  167. # [13:41] * agibson|afk is now known as agibson
  168. # [14:22] * Joins: victorporof (victorporo@moz-FFB76956.cpe.net.cable.rogers.com)
  169. # [14:42] * Joins: Justin_o_ (uid14648@moz-E77DEB21.irccloud.com)
  170. # [14:47] * Quits: richardschwerdtfeger (richardsch@moz-B4F71088.lightspeed.austtx.sbcglobal.net) (Quit: richardschwerdtfeger)
  171. # [14:50] * Gijs is now known as Gijs_away
  172. # [15:00] * Joins: MarcoZ (marco.zehe@13F2CEC5.7672369.D8E68FF6.IP)
  173. # [15:00] * ChanServ sets mode: +ao MarcoZ MarcoZ
  174. # [15:08] * Quits: victorporof (victorporo@moz-FFB76956.cpe.net.cable.rogers.com) (Quit: victorporof)
  175. # [15:10] * Joins: yzen (yzen@13F2CEC5.7672369.D8E68FF6.IP)
  176. # [15:10] * ChanServ sets mode: +o yzen
  177. # [15:16] * Joins: jongunderson (chatzilla@moz-238203DB.dres.uiuc.edu)
  178. # [15:18] * Joins: jwei (jwei@13F2CEC5.7672369.D8E68FF6.IP)
  179. # [15:24] * Joins: richardschwerdtfeger (richardsch@11EF93FD.9DD4DBBF.6A7A197.IP)
  180. # [15:35] * Joins: victorporof (victorporo@13F2CEC5.7672369.D8E68FF6.IP)
  181. # [15:54] * Joins: clown (clown@67828CC7.C1A51174.9D42CF23.IP)
  182. # [15:58] * Joins: surkov (surkov@moz-DF24A6EA.cpe.pppoe.ca)
  183. # [15:58] * ChanServ sets mode: +o surkov
  184. # [16:21] * agibson is now known as agibson|afk
  185. # [16:31] * Joins: kushagra (kushagra@9E4F728B.A50065F5.C28326FD.IP)
  186. # [16:34] * Joins: davidb (davidb@13F2CEC5.7672369.D8E68FF6.IP)
  187. # [16:34] * ChanServ sets mode: +qo davidb davidb
  188. # [16:38] * Gijs_away is now known as Gijs
  189. # [16:39] * agibson|afk is now known as agibson
  190. # [16:40] * Quits: SteveF (chatzilla@moz-6F24D0BD.cable.virginm.net) (Client exited)
  191. # [16:55] * Joins: kushagra_ (kushagra@CE5FF735.9C102D3A.520CDC98.IP)
  192. # [16:55] * Quits: kushagra (kushagra@9E4F728B.A50065F5.C28326FD.IP) (Ping timeout)
  193. # [16:57] <@davidb> jwei: how's now?
  194. # [16:57] <jwei> davidb: Sounds good.
  195. # [16:57] <@davidb> omw
  196. # [17:08] * Quits: victorporof (victorporo@13F2CEC5.7672369.D8E68FF6.IP) (Quit: victorporof)
  197. # [17:10] * Joins: victorporof (victorporo@13F2CEC5.7672369.D8E68FF6.IP)
  198. # [17:14] * Quits: jongunderson (chatzilla@moz-238203DB.dres.uiuc.edu) (Quit: ChatZilla 0.9.90.1 [Firefox 27.0.1/20140212131424])
  199. # [17:44] * khuey|away is now known as khuey
  200. # [17:47] * Quits: @yzen (yzen@13F2CEC5.7672369.D8E68FF6.IP) (Ping timeout)
  201. # [17:51] * Quits: @surkov (surkov@moz-DF24A6EA.cpe.pppoe.ca) (Quit: surkov)
  202. # [17:57] * clown is now known as clown_mtg
  203. # [17:58] * Quits: kushagra_ (kushagra@CE5FF735.9C102D3A.520CDC98.IP) (Ping timeout)
  204. # [17:58] * Joins: surkov (surkov@moz-DF24A6EA.cpe.pppoe.ca)
  205. # [17:58] * ChanServ sets mode: +o surkov
  206. # [18:16] * Joins: yzen (yzen@13F2CEC5.7672369.D8E68FF6.IP)
  207. # [18:16] * ChanServ sets mode: +o yzen
  208. # [18:17] * Quits: @MarcoZ (marco.zehe@13F2CEC5.7672369.D8E68FF6.IP) (Ping timeout)
  209. # [18:17] * Quits: @yzen (yzen@13F2CEC5.7672369.D8E68FF6.IP) (Ping timeout)
  210. # [18:19] <@firebot> New Core - Disability Access APIs bug 977170 filed by surkov.alexander@gmail.com.
  211. # [18:19] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=977170 nor, --, ---, nobody, NEW, can't query IAccessible from HTML area element having ISimpleDOMNode pointer
  212. # [18:24] * Joins: maxli (maxli@moz-EE42E0E.student.cs.uwaterloo.ca)
  213. # [18:29] * Quits: richardschwerdtfeger (richardsch@11EF93FD.9DD4DBBF.6A7A197.IP) (Quit: richardschwerdtfeger)
  214. # [18:36] * Quits: khuey (khuey@moz-DB4A9C19.scl3.mozilla.com) (Ping timeout)
  215. # [18:39] * Joins: khuey (khuey@moz-DB4A9C19.scl3.mozilla.com)
  216. # [18:48] * Joins: yzen (yzen@13F2CEC5.7672369.D8E68FF6.IP)
  217. # [18:48] * ChanServ sets mode: +o yzen
  218. # [18:52] * Joins: MarcoZ (marco.zehe@13F2CEC5.7672369.D8E68FF6.IP)
  219. # [18:52] * ChanServ sets mode: +ao MarcoZ MarcoZ
  220. # [19:16] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Quit: scott_gonzalez)
  221. # [19:18] * Joins: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net)
  222. # [19:23] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Quit: scott_gonzalez)
  223. # [19:38] <@firebot> jwei@mozilla.com requested needinfo from trev.saunders@gma il.com on bug 977170.
  224. # [19:38] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=977170 nor, --, ---, nobody, NEW, can't query IAccessible from HTML area element having ISimpleDOMNode pointer
  225. # [19:48] * Joins: kushagra (kushagra@E8FDF2F6.613DE165.520CDC98.IP)
  226. # [19:52] <@tbsaunde> surkov: are you sure nsAutoPtr<T> Foo(); nsAutoPtr<T> x(Foo()); doesn't work? it looks like it should
  227. # [19:52] <@surkov> tbsaunde: I bet it works
  228. # [19:53] <@tbsaunde> surkov: then what's wrong with just doing that?
  229. # [19:54] <@surkov> tbsaunde: you need to consider how it's supposed to work with passing T as argument and how we will deal with arrays
  230. # [19:55] <@tbsaunde> surkov: when would you pass T as an argument?
  231. # [19:55] <@surkov> tbsaunde: when I construct xpcom object
  232. # [19:55] <@tbsaunde> surkov: x.forget() or Move(x) whichever you prefer
  233. # [19:56] <@surkov> tbsaunde: I'm pretty sure if we spent some time for investigations then we will figure out how we can deal with it but spending time for it isn't probably worth
  234. # [19:58] <@tbsaunde> surkov: I really don't care what you do as long as objects can't have multiple owners
  235. # [20:00] <@surkov> tbsaunde: why is it so important for you?
  236. # [20:02] <@surkov> tbsaunde: also could you please outline your position in the bug, to state why refptr is a stopper for the patch
  237. # [20:02] <@tbsaunde> surkov: its bad design and allows people to old onto things in ways they shouldn't, so it'll be much harder to fix later
  238. # [20:03] * Joins: richardschwerdtfeger (richardsch@11EF93FD.9DD4DBBF.6A7A197.IP)
  239. # [20:05] <@surkov> tbsaunde: will you?
  240. # [20:08] * agibson is now known as agibson|afk
  241. # [20:11] <@tbsaunde> surkov: in a minute
  242. # [20:11] <@surkov> thank you
  243. # [20:12] * Joins: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net)
  244. # [20:24] <@firebot> continuation@gmail.com changed the Component on bug 976895 from Disability Access APIs to DOM.
  245. # [20:24] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=976895 nor, --, ---, nobody, NEW, Intermittent name/test_list.html | application terminated with exit code 2147483651 | application cr
  246. # [20:28] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Quit: scott_gonzalez)
  247. # [20:34] * Joins: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net)
  248. # [20:42] <@davidb> hurrah a review request!
  249. # [20:43] * @davidb feels inadequate
  250. # [20:45] * agibson|afk is now known as agibson
  251. # [20:45] <@tbsaunde> surkov: so actually I really think you should do the GetFooTextRange&) thing because it gets rid of allocations
  252. # [20:46] * @davidb likes references
  253. # [20:46] <@surkov> tbsaunde: and then do you want to copy it for xpcom object?
  254. # [20:47] <eeejay> references are weak
  255. # [20:48] <@surkov> eeejay: in what sense?
  256. # [20:48] <@tbsaunde> surkov: no, I want to do class XPCTextRange { TextRange mRange; };
  257. # [20:48] <eeejay> surkov, a bad attempt at a pun :)
  258. # [20:49] <@davidb> don't talk about resumes, my heart can't take it
  259. # [20:50] <@tbsaunde> surkov: then you can do GetFoo(xpcTextRange.mRange)
  260. # [20:50] <@tbsaunde> -> I guess but whatever
  261. # [20:54] <@surkov> tbsaunde: what about arrays?
  262. # [20:55] * Quits: maxli (maxli@moz-EE42E0E.student.cs.uwaterloo.ca) (Ping timeout)
  263. # [20:57] <@tbsaunde> surkov: pass in nsTArray<TextRange> *
  264. # [20:57] * Joins: maxli (maxli@moz-EE42E0E.student.cs.uwaterloo.ca)
  265. # [20:58] <@tbsaunde> then copy to the xpcTextRange objects you create or use Move ctors if you can
  266. # [20:58] <@surkov> tbsaunde: so the caller have to make sure to free memory of all objects of that nsTArray?
  267. # [20:58] <@tbsaunde> surkov: no, nsTArray dtor should handle it
  268. # [20:59] <@tbsaunde> I'm saying the nsTArray should be on the stack if that wasn't clear
  269. # [21:00] <@surkov> I don't feel objects coping is better than reference counting
  270. # [21:01] <@tbsaunde> what? why?
  271. # [21:01] <@tbsaunde> the real point is less allocations
  272. # [21:02] <@surkov> tbsaunde: how to transform nsTArray<TextRange> to nsIArray of xpcTextRange?
  273. # [21:06] <@surkov> anyway I don't feel like we have well-baked solution so I would leave the patch as it is and keep discussion as follow up
  274. # [21:07] <@tbsaunde> surkov: just with a loop more or less the same way you do now
  275. # [21:07] * @tbsaunde has clue why this is nearly so hard
  276. # [21:07] <@tbsaunde> *no clue
  277. # [21:09] <@surkov> tbsaunde: if the solution is really so clear then I don't really stop you to put a patch on top and actually that'd be great
  278. # [21:09] <@davidb> yeah why don't you guys do that
  279. # [21:09] <@tbsaunde> surkov: fine
  280. # [21:10] <@surkov> tbsaunde: thank you!
  281. # [21:10] <@surkov> tbsaunde: how do we want proceed from technical prespective?
  282. # [21:11] <@tbsaunde> surkov: what do you want to know exactly?
  283. # [21:11] <@surkov> tbsaunde organizational perspective
  284. # [21:12] <@tbsaunde> surkov: orgainization of what exactly?
  285. # [21:12] <@surkov> tbsaunde: I meant we have a patch having r- from your and r? from davidb, should we land it before you create your one?
  286. # [21:12] <@davidb> is there urgency here?
  287. # [21:12] <@tbsaunde> surkov: I'd prefer you didn't
  288. # [21:13] <@surkov> tbsaunde: so how do you like to proceed with it?
  289. # [21:15] <@surkov> davidb: not really urgency but it's not good for on-top work in terms of merging
  290. # [21:16] * Quits: agibson (agibson@moz-2C643250.gate.cable.virginm.net) (Quit: )
  291. # [21:18] <@tbsaunde> surkov: I guess I'm just write a patch for that bug today since you have some huge problem with doing it yourself
  292. # [21:19] <@davidb> seems like a reasonable way forward
  293. # [21:20] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Ping timeout)
  294. # [21:20] * Joins: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net)
  295. # [21:20] <@surkov> tbsaunde: do you want qfold them?
  296. # [21:21] <@tbsaunde> surkov: I'll just apply your patch and then fix it up
  297. # [21:22] <@surkov> tbsaunde: ok
  298. # [21:28] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Ping timeout)
  299. # [21:28] * Joins: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net)
  300. # [21:50] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Ping timeout)
  301. # [21:51] * Joins: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net)
  302. # [21:53] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Ping timeout)
  303. # [21:54] * Joins: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net)
  304. # [21:54] * Quits: scott_gonzalez (scott_gonz@moz-91C81A39.hrbgpa.fios.verizon.net) (Quit: scott_gonzalez)
  305. # [21:57] <@davidb> MarcoZ: you are very prolific this week
  306. # [22:00] <@davidb> (which is awesome)
  307. # [22:05] <@tbsaunde> surkov: can things like EnclosingRange() ever return null with real implementation?
  308. # [22:06] <@surkov> tbsaunde: nothing prevents RangeFromPoint to return null
  309. # [22:09] <@tbsaunde> surkov: how about other methods though?
  310. # [22:10] <@tbsaunde> surkov: valid ranges always have a root right?
  311. # [22:11] <@surkov> tbsaunde: EnclosingRange shouldn't have null range
  312. # [22:12] <@surkov> tbsaunde: yes
  313. # [22:12] <@surkov> RangeFromChild could have null range I guess
  314. # [22:16] <@davidb> http://webaim.org/projects/screenreadersurvey5/
  315. # [22:18] <@tbsaunde> surkov: you mean yes range always has root?
  316. # [22:18] <@tbsaunde> surkov: also, you don't have to do foo<bar<blah> > you can do foo<bar<blah>> now
  317. # [22:23] * Quits: @davidb (davidb@13F2CEC5.7672369.D8E68FF6.IP) (Quit: davidb)
  318. # [22:25] * Joins: jongund (chatzilla@moz-97EEA285.rehab.uiuc.edu)
  319. # [22:28] <@firebot> eitan@monotonous.org changed the Assignee on bug 976142 from nobody@mozilla.org to eitan@monotonous.org.
  320. # [22:29] <@firebot> Bug https://bugzilla.mozilla.org/show_bug.cgi?id=976142 nor, --, ---, eitan, NEW, [AccessFu] Create useful debug event string for vc change events
  321. # [22:41] * Quits: kushagra (kushagra@E8FDF2F6.613DE165.520CDC98.IP) (Client exited)
  322. # [22:43] * Quits: @surkov (surkov@moz-DF24A6EA.cpe.pppoe.ca) (Quit: surkov)
  323. # [23:02] * Quits: @yzen (yzen@13F2CEC5.7672369.D8E68FF6.IP) (Ping timeout)
  324. # [23:04] * Parts: clown_mtg (clown@67828CC7.C1A51174.9D42CF23.IP)
  325. # [23:21] * Joins: surkov (surkov@moz-DF24A6EA.cpe.pppoe.ca)
  326. # [23:21] * ChanServ sets mode: +o surkov
  327. # [23:28] * Quits: @surkov (surkov@moz-DF24A6EA.cpe.pppoe.ca) (Quit: surkov)
  328. # [23:28] * Quits: icaaq (icaaq@moz-6A20EAD4.ias.bredband.telia.com) (Quit: 10-79 Notify coroner, over and out)
  329. # [23:36] * Quits: jwei (jwei@13F2CEC5.7672369.D8E68FF6.IP) (Quit: Leaving.)
  330. # [23:37] * Quits: victorporof (victorporo@13F2CEC5.7672369.D8E68FF6.IP) (Quit: victorporof)
  331. # [23:41] * Joins: kushagra (kushagra@E8FDF2F6.613DE165.520CDC98.IP)
  332. # [23:43] * Quits: kushagra (kushagra@E8FDF2F6.613DE165.520CDC98.IP) (Ping timeout)
  333. # [23:43] * Quits: jongund (chatzilla@moz-97EEA285.rehab.uiuc.edu) (Quit: ChatZilla 0.9.90.1 [Firefox 27.0.1/20140212131424])
  334. # Session Close: Thu Feb 27 00:00:00 2014

The end :)