932419 - don't make NS_ASSERT's pop up a modal dialog

Status: VERIFIED FIXED

(Toolkit :: General, defect)

Description

[Timothy Nikkel (:tnikkel)]

Attachment: tabassertrem

STR: Install FavIconize Tab extension. Faviconize a tab. Close the tab. Hit modal assert that interrupts your workflow.

This bites me constantly every single day, eating away at my productivity.

If someone is actually going to look into the problem and fix it then fine, but as it stands right now this thing bugs me all the time and no one is doing anything about it. So we get no positive effect from having this assertion.

Maybe the extension is to blame, I don't care at this point, I need a solution.

Comment 1

[Dão Gottwald [:dao]]

Comment on attachment 824162 [details] [diff] [review]
tabassertrem

This is indeed a bug in the extension and needs to be reported to its authors so they can fix it. Apparently FavIconizeTab is on github, so presumably you could also contribute a patch there: https://github.com/ktakayama/faviconizetab/

Comment 2

[Tim Taubert [:ttaubert] (inactive)]

Isn't this just an add-on that implements pinned tabs? The description mentions Fx 2.0 and it was last updated two years ago. It modifies max- and min-width of faviconized tabs and thus pretty clearly breaks the animation. Or rather, we don't know that there is no animation.

Comment 3

[Timothy Nikkel (:tnikkel)]

(In reply to Tim Taubert [:ttaubert] from comment #2)
> Isn't this just an add-on that implements pinned tabs? The description
> mentions Fx 2.0 and it was last updated two years ago. It modifies max- and
> min-width of faviconized tabs and thus pretty clearly breaks the animation.
> Or rather, we don't know that there is no animation.

No, it just shrinks their width. It doesn't move them to a special area or anything.

Comment 4

[Timothy Nikkel (:tnikkel)]

I don't really have time to debug someone's extension.

Surely there is a better solution here then popping up a modal dialog?

Comment 5

[Timothy Nikkel (:tnikkel)]

(In reply to Tim Taubert [:ttaubert] from comment #2)
> Isn't this just an add-on that implements pinned tabs? The description
> mentions Fx 2.0 and it was last updated two years ago. It modifies max- and
> min-width of faviconized tabs and thus pretty clearly breaks the animation.
> Or rather, we don't know that there is no animation.

I don't really care about a tab closing animation. It's not really expected that a shrunken tab has a tab closing animation.

Comment 6

[Dão Gottwald [:dao]]

(In reply to Timothy Nikkel (:tn) from comment #5)
> I don't really care about a tab closing animation. It's not really expected
> that a shrunken tab has a tab closing animation.

Then the extension should make it so that the animation doesn't happen in the first place (like we do for pinned tabs). The assert is there because the animation starts but we never see it finish.

(In reply to Timothy Nikkel (:tn) from comment #4)
> Surely there is a better solution here then popping up a modal dialog?

We could change the behavior in http://mxr.mozilla.org/mozilla-central/source/toolkit/modules/debug.js.

Comment 7

[Timothy Nikkel (:tnikkel)]

(In reply to Dão Gottwald [:dao] from comment #6)
> We could change the behavior in
> http://mxr.mozilla.org/mozilla-central/source/toolkit/modules/debug.js.

Oh nice, that code considers nightly and beta as non-release builds, but aurora is a release build.

Comment 8

[Timothy Nikkel (:tnikkel)]

Attachment: modalassert

Comment 9

[:Gavin Sharp [email: [e-mail]]]

See also bug 914092. Can you resummarize this bug?

Comment 10

[Timothy Nikkel (:tnikkel)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/a96aea73b2cf

Comment 11

[Phil Ringnalda (:philor)]

https://hg.mozilla.org/mozilla-central/rev/a96aea73b2cf

Comment 13

[Ioana (away)]

Reproduced this bug with Firefox 28.0a1. Verified as fixed with Firefox 28.0b4 on windows 7 64bit, Mac OS X 10.8.5 and Ubuntu 13.04 32bit (the message is now displayed in the browser console).