Closed Bug 1607441 Opened 4 years ago Closed 4 years ago

Pinned tabs order are reversed every time Nightly is restarted

Categories

(Firefox :: Tabbed Browser, defect, P1)

74 Branch
defect

Tracking

()

VERIFIED FIXED
Firefox 74
Tracking Status
firefox-esr68 --- unaffected
firefox72 --- unaffected
firefox73 --- unaffected
firefox74 + verified

People

(Reporter: Fanolian+BMO, Assigned: emmamalysz)

References

(Regression)

Details

(Keywords: nightly-community, regression, reproducible)

Attachments

(1 file, 1 obsolete file)

User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:74.0) Gecko/20100101 Firefox/74.0
Build ID: 20200107095722

Steps to reproduce

  1. In a new profile, pin 2 or more tabs, preferably with different favicons to differentiate them easier.
  2. Restart, or close & reopen Nightly.

Actual result

Order for pinned tabs are reversed.

Expected result

Order remains unchanged.

This is regressed by bug 1242912. Is bug 1606633 related?

[Tracking Requested - why for this release]:

Has Regression Range: --- → yes
Has STR: --- → yes
Regressed by: 1242912
Summary: Pinned tabs order are reversed every time Nightly is restarted/session is restored → Pinned tabs order are reversed every time Nightly is restarted

Thanks for filing this.

(In reply to Fanolian from comment #0)

This is regressed by bug 1242912. Is bug 1606633 related?

No, that bug is just about the performance of pinned tabs. Because of how pinning works, pinned tabs get inserted immediately, whereas all the other tabs get inserted as a single document fragment. bug 1606633 is about doing the same with pinned tabs, which was hairy because of the implementation of pinning.

This is a pretty odd bug as they come; we pin tabs in the order in which they come in, so I'm not sure where the order gets reversed... Emma, can you take a look?

Flags: needinfo?(emalysz)
Assignee: nobody → emalysz
Flags: needinfo?(emalysz)

This started right with today's update for me: 20200107095722

User Agent Mozilla/5.0 (X11; Linux x86_64; rv:74.0) Gecko/20100101 Firefox/74.0

I've had four starts with 20200107095722 and each time the pinned tab order reversed.

For those who don't see the issue, what is your configuration? This doesn't seem to be OS related as the original reporter is on Windows and I'm using Ubuntu.

(In reply to B.J. Herbison from comment #4)

For those who don't see the issue, what is your configuration?

Nobody on this bug has suggested they cannot reproduce.

Can we land a test for this?

Flags: in-testsuite?
Pushed by gijskruitbosch@gmail.com:
https://hg.mozilla.org/integration/autoland/rev/7109d6215ce1
ensure pinned tabs order is correct by updating their position attribute r=Gijs

(In reply to Julien Cristau [:jcristau] from comment #8)

Can we land a test for this?

We should, but let's do it in a follow-up. Given Emma and I are 8 timezones apart another review cycle would delay fixing this on nightly.

Blocks: 1607698

(In reply to :Gijs (he/him) from comment #10)

(In reply to Julien Cristau [:jcristau] from comment #8)

Can we land a test for this?

We should, but let's do it in a follow-up. Given Emma and I are 8 timezones apart another review cycle would delay fixing this on nightly.

Nightly has been my primary browser for years and I put up with minor annoyances to help improve Firefox. For this fix (and almost anything other than a crash or major usability problem) I'd rather wait a day and have a new or improved test included with the fix. I have no doubt about your intentions, but sometimes a "follow-up" can drift into the pile of unfixed work.

(In reply to B.J. Herbison from comment #11)

(In reply to :Gijs (he/him) from comment #10)

(In reply to Julien Cristau [:jcristau] from comment #8)

Can we land a test for this?

We should, but let's do it in a follow-up. Given Emma and I are 8 timezones apart another review cycle would delay fixing this on nightly.

Nightly has been my primary browser for years and I put up with minor annoyances to help improve Firefox. For this fix (and almost anything other than a crash or major usability problem) I'd rather wait a day and have a new or improved test included with the fix.

I appreciate the vote of confidence.

I have no doubt about your intentions, but sometimes a "follow-up" can drift into the pile of unfixed work.

I'm writing the test right now. :-)

(In reply to B.J. Herbison from comment #11)

(In reply to :Gijs (he/him) from comment #10)

(In reply to Julien Cristau [:jcristau] from comment #8)

Can we land a test for this?

We should, but let's do it in a follow-up. Given Emma and I are 8 timezones apart another review cycle would delay fixing this on nightly.

Nightly has been my primary browser for years and I put up with minor annoyances to help improve Firefox. For this fix (and almost anything other than a crash or major usability problem) I'd rather wait a day and have a new or improved test included with the fix. I have no doubt about your intentions, but sometimes a "follow-up" can drift into the pile of unfixed work.

I disagree, this bug is hugely annoying for me and gets in the way of getting work done. Thanks for landing the fix asap.

Component: Session Restore → Tabbed Browser
Priority: -- → P1

(In reply to zak.mcgregor from comment #14)

Just to confirm that it happens for me, 3 times now. Reversing the reversed pinned tabs made it appear correct, but the latest restart reversed them again.

74.0a1 (2020-01-08) (64-bit) on Fedora 28.

A fix is already on autoland and, all being well, will be in tonight's nightly (approx. 8-10 hours from now). Please do not leave more "I see this too" comments.

Status: NEW → RESOLVED
Closed: 4 years ago
Resolution: --- → FIXED
Target Milestone: --- → Firefox 74

Possibly related, but I just updated and all my pinned tabs were gone except for the previously focused one.

Regressions: 1608036

Possibly related, but I just updated and all my pinned tabs were gone except for the previously focused one.

Reported from many places, instead of fixing things this is a dataloss bug now. Filed the new regression as bug 1608036.

It would be great to also get a test for this particular situation. As it looks like it would have to be a Marionette one given that a restart is involved, right?

(In reply to Henrik Skupin (:whimboo) [⌚️UTC+1] from comment #20)

It would be great to also get a test for this particular situation. As it looks like it would have to be a Marionette one given that a restart is involved, right?

No, we can test session store without restarting, and a test is already in the dep bug and has r+, but there's room for more tests so I haven't landed it yet - I'm also more concerned with the regressions about losing pinned tabs on nightly right now.

Attachment #9119645 - Attachment is obsolete: true
Flags: qe-verify+

Confirmed issue with 74.0a1 (2020-01-07) on Windows10.
Fix verified with 74.0b4 on Windows 10, macOS 10.13, Ubuntu 18.04.

Status: RESOLVED → VERIFIED
Flags: qe-verify+
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: