mozilla :: #accessibility

17 Mar 2017
13:14davidbtbsaunde: all i did was look at what you changed, and the reorder-reorder case, and try to figure out what we used to do. I spent about 25 minutes hurting my head.
13:15davidbbut I did get the same scribbles, trying twice, so I might revisit for my own sanity.
13:24tbsaundeheh, linked lists are hard I guess
13:24tbsaunde?
13:50MarcoZThis is all waaaaaayyyyyy too complicated. ;)
13:51* MarcoZ is beginning to like the idea of getting rid of buffers and doing a proper dynamic web solution on Windows more and more.
14:02MarcoZBTW: Weekend now! Parents come for a visit, later a theatre visit.
15:16tbsaundedavidb: whatever happens after the null deref I'm going to push that patch in a bit
15:31davidbsgtm
15:50davidbi clarified that with a comment
17:05firebotNew Core - Disability Access APIs bug 1348340 filed by jdiggs@igalia.com.
17:05firebothttps://bugzil.la/1348340 NEW, nobody@mozilla.org Reliable crash when using atspi_table_cell_get_position() on ARIA table cell
17:12joanietbsaunde: mind looking at the above bug?
17:13tbsaundejoanie: sure
17:13joanieTurns out that using the automated tool I'm working on for ARIA 1.1. testing crashes your browser reliably and during every test run if we actually try to test aria-colindex and aria-rowindex implementation.
17:15tbsaundeheh, I doubt it'll be hard to fix
17:15joaniecool
17:17joanieit's proving surprising hard to *test* though. First I had to get at it via atspi methods, one of which didn't quite exist. That was fixed by implementing the right/preferred method instead. And now that I can test it via atspi, the findings are reliable segfault.
17:17joanienot sure if that means implemented or not </troll>
17:18joanieanyhoo, since I&#39;m officially not working today.... /me waves and says thanks and wishes tbsaunde a great weekend :)
17:20tbsaundeyeah you too, maybe I&#39;ll have it fixed for next week
17:25tbsaundejoanie: fwiw, the stack in that report is pretty bogus looking, but I can probably guess what&#39;s happening anyway
17:36tbsaundejoanie: actually I do want to see a better stack if you can when you are working for real
17:40firebotNew Core - Disability Access APIs bug 1348351 filed by aryx.bugmail@gmx-topmail.de.
17:40firebothttps://bugzil.la/1348351 NEW, nobody@mozilla.org Intermittent test_navigation.py TestBackForwardNavigation.test_frameset_after_navigating_in_frame |
17:46davidbjoanie: thanks for pinging us on a not working day :)
19:26tbsaundeaklotz: is it a safe assumption new version of the handler patches are coming and I can cancel the reviews?
19:26aklotztbsaunde: yeah
19:26tbsaundethanks
19:35tbsaundedavidb: going to get to bug 1348148 at some point?
19:35firebothttps://bugzil.la/1348148 NEW, nobody@mozilla.org fix assertion failure in Accessible::HandleAccEvent() that there is an ipc doc
19:39davidbthanks for ping
19:39davidblooking
19:58joanietbsaunde: https://bug1348340.bmoattachments.org/attachment.cgi?id=8848609 How&#39;s that?
19:59firebothttps://bugzil.la/1348340 NEW, nobody@mozilla.org Reliable crash when using atspi_table_cell_get_position() on table cell
20:11tbsaundejoanie: well, that&#39;s a various stack, looking into what&#39;s going on, but it looks like the source of the crash is at-spi2-atk doing something dbus doesn&#39;t like
20:11tbsaundewtf I mean curious not various
20:18tbsaundejoanie: ok, I don&#39;t claim to understand this dbus code without reading a bunch of docs, but I really suspect something is wrong with impl_get_Position() in at-spi2-atk/atk-adaptors/adaptors/table-cell.c
20:19tbsaundefor one thing that code doesn&#39;t really seem similar to other functions there that return multiple ints
18 Mar 2017
No messages
   
Last message: 100 days and 9 hours ago