GNOME Bugzilla – Bug 409728
[verified] Orca repeats mouse-selected text numerous times in Gecko
Last modified: 2008-07-22 19:27:53 UTC
Steps to reproduce: 1. Launch Thunderbird and press Control N to create a new message 2. Tab to the body of the message and type "This is a test." 3. Use the mouse to select the text you just typed. Expected results: Orca would either say nothing (because at this time it typically does not announce text that was selected via the mouse) OR it would speak the selected text once. Actual results: It repeats the selected text quite a few times. Full debug.out to follow.
Created attachment 82908 [details] aforementioned debug.out
Left off an important word in the description, namely "thunderbird." Sorry for the spam.
Created attachment 83702 [details] Another debug.out I just built Thunderbird 3.0 (alpha) from Mozilla CVS and tried this with latest Orca on my Ubuntu Feisty system. I get a slightly different result. It speaks "T" 18 times. I thought it be the number of characters in the test message, but it's more than that.
Created attachment 83743 [details] [review] Patch to fix the problem. Assuming we don't want to speak/braille anything for mouse activated events.
I've reassigned this to at-spi since it really appears to be a problem with at-spi and not Orca.
Aargh! I hate Bugzilla and the way it shoves unassociated links in your face. Wrong bug.
I can also reproduce this when selecting text with the mouse in a FF document frame. For example, I selected the "Additional" from "Additional Comments:" on this page and heard "Additional Comments" spoken many times. Retitling and also setting the blocks field.
Created attachment 90307 [details] [review] patch to hopefully solve the problem We already have an RFE (bug 409731) regarding the need to speak text that is selected by the mouse therefore, what needs doing here is to stop the extreme chattiness that is resulting from all of the caret-moved events. Rich's patch accomplishes this nicely, but it should be in Gecko.py's onCaretMoved() instead as the problem is also in Firefox. Will okay to commit?
> We already have an RFE (bug 409731) regarding the need to speak text that is > selected by the mouse therefore, what needs doing here is to stop the extreme > chattiness that is resulting from all of the caret-moved events. As long as we have a bug/RFE to get us to the desired behavior, this sounds good. > Will okay to commit? Yep! Thanks!
Done. Thanks. Moving to pending.
I don't think I can verify this one.
Can I verify it? :-)