GNOME Bugzilla – Bug 618484
Heuristic method of handling Thunderbird's flooded text insertions
Last modified: 2010-09-20 10:54:40 UTC
When using Thunderbird and replying to a large message, sometimes Thunderbird will send many GTK text insertion events instead of one large event, causing Orca to lag and, if using speech, to be impossible to silence until it catches up to the events. Steps to reproduce: Reply to a large message on a single-core system. This does not appear to happen on multi-core systems due to threading. This may have to be tried several times, as it doesn't always occur. Proposed solution: Ideally, Mozilla themselves would fix this issue. In the mean time, a heuristic approach to determine which text insertion events should be interpreted and which should be ignored is probably a reasonable improvement. Note that I cannot attach a debug.out as such would contain personal information about not only myself but people I'm conversing with.
Created attachment 161106 [details] [review] proposed fix (In reply to comment #0) > When using Thunderbird and replying to a large message, sometimes Thunderbird > will send many GTK text insertion events instead of one large event, Huh. Now that I'm looking at it, I'm seeing only one large event rather than a bunch of smaller ones. I'm using today's build: Mozilla/5.0 (X11; U; SunOS i86pc; en-US; rv:1.9.3a5pre) Gecko/20100514 Shredder/3.2a1pre What are you using? > Orca to lag and, if using speech, to be impossible to silence until it catches > up to the events. If you are indeed seeing a bunch of events, I cannot reproduce it. > Note that I cannot attach a debug.out as such would contain personal > information about not only myself but people I'm conversing with. Totally understood. But that might make solving this problem a bit trickier seeing as how what you describe having in your debug.out doesn't match what I have in mine. Having said that, I took a stab at it anyway (attached). If that doesn't solve the problem for you, and you cannot get me a full debug.out, well, we'll have to figure something out....
Jacob, have you had the opportunity to test this yet? Thanks!
Well, even though Jacob hasn't commented on this, I do believe it is safe and I do believe it will stop the problem. And getting it into master would cause it to get some much-needed user testing. Alejandro, if you are looking for a means to test your new git access.... :-) Since I didn't know at the time I created this patch that someone else might be committing it, I just did a git diff rather than a format patch. If you do commit it, please use the format we've adopted (i.e. commit line: 'Fix for bgo#12345 - blah blah blah') and set the author to me. Thanks!
Comment on attachment 161106 [details] [review] proposed fix Sorry for the spam. Setting to 'accepted-commit_now' to get it off of the 'patches without a response' list. And the reason I'm not committing my own patch myself is so that Alejandro will have some opportunities to get familiar with our git repository.
Works for me. Pushed to master http://git.gnome.org/browse/orca/commit/?id=fa871a38dbe8c5f66b2df52887faa8517799e4af