Holding Shift while choosing a "Switch to Tab" result still works correctly visually: It changes the action text from "Switch to tab" to the URL; but then pressing Enter still switches to tab instead of opening the selected url at the current tab as expected.

Last good revision: 7fe3ee0cf8be (2014-04-18) First bad revision: 582b2d81ebe1 (2014-04-19) Pushlog: https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=7fe3ee0cf8be&tochange=582b2d81ebe1 I'm currently guessing the cause is bug 754265. Inbound bisection in progress

Inbound: Last good revision: 7c5d076dd141 First bad revision: 5f558788549e Pushlog: https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=7c5d076dd141&tochange=5f558788549e Bug 754265 seems most likely so tentatively blocking.

Tracking because it is a regression.

Marco, could you have a look to this bug? Thanks

I will try to, this should probably be prioritized in the firefox backlog for the next iteration.

this should fix the problem, but I must ensure it's not regressing the expected behavior and it needs a test.

this should be safer, I also included a test. Try run proceeding https://tbpl.mozilla.org/?tree=Try&rev=86e10785e7e2

Comment on attachment 8433374 [details] [diff] [review] patch v1 Review of attachment 8433374 [details] [diff] [review]: ----------------------------------------------------------------- note to myself: I should promiseClearHistory before exiting the test, just in case.

Comment on attachment 8433374 [details] [diff] [review] patch v1 [Approval Request Comment] Bug caused by (feature/regressing bug #): bug 754265 User impact if declined: switch to tab can't be overridden through alt/shift key Testing completed (on m-c, etc.): m-c, automated test Risk to taking this patch (and alternatives if risky): low impact oneliner patch with a test String or IDL/UUID changes made by this patch: none

Hi Florin, do you know who could be assigned as the QA Contact for verification of this bug?

I've verified this bug using the following environment: FF 32 Build Id: 20140606030206 OS: Win 7 x65, Mac Os X 10.8.5, Ubuntu 12.10 x86 I found 2 strange behaviors: 1. On Win and Mac when using Alt + Enter the page is opened in a new tab, however on Ubuntu your are switched to the original tab. 2. On windows holding Shift or Alt for 2 seconds will override "switch to tab" even after you let go of the key.

mak, when I tried this in Mac OSX 10.8.5 with 32.0a1 (2014-06-09), holding down the shift key + Enter on a "Switch to New Tab" link, I got a new window (not just a new tab). Is that intentional?

(In reply to Liz Henry :lizzard from comment #15) > mak, when I tried this in Mac OSX 10.8.5 with 32.0a1 (2014-06-09), holding > down the shift key + Enter on a "Switch to New Tab" link, I got a new window > (not just a new tab). Is that intentional? It shouldn't open a new window or tab when you are holding shift, it should *switch* to the existing tab (which may happen to be in another window). Are you talking about some steps where you are not holding shift? :mak, see Catalin's issues in comment 13.

(In reply to Catalin Varga [QA][:VarCat] from comment #13) > I found 2 strange behaviors: > 1. On Win and Mac when using Alt + Enter the page is opened in a new tab, > however on Ubuntu your are switched to the original tab. I just tried on Win and with Alt enter the page is loaded into the current tab, that is expected, not a new tab. It's actually not really nice cause Alt also opens the menubar and you might end up with fancy behavior (I guess we should only support shift on Win). No idea about ubuntu, the code doesn't seem to differentiate any platform. > 2. On windows holding Shift or Alt for 2 seconds will override "switch to > tab" even after you let go of the key. There is no specific code comment for this, but the implementation using keydown and a counter looks suspicious to me. It actually looks like it was implemented this way on purpose, since IIRC keydown may be fired multiple times if you keep the button pressed, on some platform. I think it's a very hidden feature. I think you should test those behaviors into three situations: - before bug 754265 - before bug 1003461 (this one) - current nightly I suspect those behaviors are not really regressions, and thus should be filed apart. (In reply to Liz Henry :lizzard from comment #15) > mak, when I tried this in Mac OSX 10.8.5 with 32.0a1 (2014-06-09), holding > down the shift key + Enter on a "Switch to New Tab" link, I got a new window > (not just a new tab). Is that intentional? you should not even get a new tab, it should load into the current tab, afaik. I don't recall if Mac has some special behavior for shift, but I'm sure shift+click is handled as open in a new window, this makes me suspect some other code is handling the event. Please verify that is actually a regression of bug 754265 or this bug, otherwise we are just speaking of a totally unrelated bug, that should be filed apart.

(In reply to Marco Bonardo [:mak] from comment #17) > > 2. On windows holding Shift or Alt for 2 seconds will override "switch to > > tab" even after you let go of the key. > > There is no specific code comment for this, but the implementation using > keydown and a counter looks suspicious to me. It actually looks like it was > implemented this way on purpose, since IIRC keydown may be fired multiple > times if you keep the button pressed, on some platform. I think it's a very > hidden feature. > > I think you should test those behaviors into three situations: > - before bug 754265 > - before bug 1003461 (this one) > - current nightly I tested this and is reproducible on all 3 builds it's not a regression. I'll fill a different bug for this to see if it's a feature or a bug. > > (In reply to Liz Henry :lizzard from comment #15) > > mak, when I tried this in Mac OSX 10.8.5 with 32.0a1 (2014-06-09), holding > > down the shift key + Enter on a "Switch to New Tab" link, I got a new window > > (not just a new tab). Is that intentional? > > you should not even get a new tab, it should load into the current tab, > afaik. I don't recall if Mac has some special behavior for shift, but I'm > sure shift+click is handled as open in a new window, this makes me suspect > some other code is handling the event. Please verify that is actually a > regression of bug 754265 or this bug, otherwise we are just speaking of a > totally unrelated bug, that should be filed apart. I tested this issue and it seems a Mac behavior both Firefox and Chrome are acting the same when you Shift+Enter an item from the history display form the address bar it will open it in a new page.

Based on the comment 17 and comment 18 marking the bug as verified. The bug was also verified as fixed on Beta: FF 31.02 Build Id:20140616143923 OS: Windows 7 x64, Ubuntu 12.10 x32, Mac Os X 10.8.5