1582532 - "open containing folder" runs zip command

Status: RESOLVED DUPLICATE of bug 1584539

(Firefox :: Downloads Panel, defect)

Description

[madhu-GG]

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Firefox/60.0

Steps to reproduce:

1. Open Firefox
2. Click the downloads button on the top right.
3. Select "Open Containing Folder" for any download that is completed.

Actual results:

A "compressing files" dialog with progress bar opens.

On checking the output of ps aux | grep zip, found the following:

zip -q -r /tmp/thunar-sendto-email.7Z3HhC/Downloads.zip Downloads

/usr/lib/x86_64-linux-gnu/Thunar/thunar-sendto-email /home/<user>/Downloads

Expected results:

Should have opened file-explorer (thunar).

Comment 1

[Alice0775 White]

[Tracking Requested - why for this release]:UX regression

Regression window:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=bf2586dc82562136449ebe14bbc14b609c20c245&tochange=1572223bd5ecd9ae000be6c766cece6533972735

Regressed by:
1572223bd5ecd9ae000be6c766cece6533972735 Christoph Kerschbaumer — Bug 1497200: Apply Meta CSP to about:downloads. r=Gijs

Comment 2

[Alice0775 White]

STR
0. Associate ".zip" file to 7-zip etc..

1. Open http://ftp.mozilla.org/pub/firefox/nightly/2004/02/2004-02-10-08-trunk/MozillaFirebird-win32.zip
2. Save it.
3. Click the downloads button in NavBar after download finished.
4. Click Folder Icon at the right side of list item

Actual results:
File Explorer and zip associated application will open.

Expected Results:
Only File Explorer will open.

Comment 3

[Christoph Kerschbaumer [:ckerschb]]

Gijs, I can fix, but i would need some guidance. Any suggestion on how to fix the problem here? Comment 2 provides great STRs. What do you think?

Comment 4

[:Gijs (he/him)]

(In reply to Christoph Kerschbaumer [:ckerschb] from comment #3)

Gijs, I can fix, but i would need some guidance. Any suggestion on how to fix the problem here? Comment 2 provides great STRs. What do you think?

I'm not actually clear on why this is breaking now when it wasn't before. I suspect that something was stopping the event from the inner button (which opens the folder) from reaching the "outer" download item (which opens the item), or the outer listener did something where it threw out the events from the inner item. I don't know what that thing was, or why it isn't happening anymore. It'd be useful to understand that better, and model the solution on that (to avoid new regressions, and/or fix potential other problems that could be caused by this change in behaviour).

I suspect ultimately something along the lines of calling stopPropagation from the inner button's handler or checking the target of the click in the outer handler is what will be needed.

What I don't understand is how the regression and comment #0 relate - comment #0 is filed using Fx60, not nightly, and was filed 7 days ago, before bug 1497200 landed. We should fix the regression, but I actually suspect there's a separate problem that was being reported. :-\

Can we split the new problem off to a new bug?

Comment 5

[Christoph Kerschbaumer [:ckerschb]]

(In reply to :Gijs (he/him) from comment #4)

What I don't understand is how the regression and comment #0 relate - comment #0 is filed using Fx60, not nightly, and was filed 7 days ago, before bug 1497200 landed. We should fix the regression, but I actually suspect there's a separate problem that was being reported. :-\

Alice could you help us answer that? Maybe find a different regression range?

Comment 6

[Alice0775 White]

(In reply to Christoph Kerschbaumer [:ckerschb] from comment #5)

(In reply to :Gijs (he/him) from comment #4)

What I don't understand is how the regression and comment #0 relate - comment #0 is filed using Fx60, not nightly, and was filed 7 days ago, before bug 1497200 landed. We should fix the regression, but I actually suspect there's a separate problem that was being reported. :-\

Alice could you help us answer that? Maybe find a different regression range?

Sorry. hijacking this bug.

I cannot reproduce the issue with STR comment#2 on 60.9.0ESR Linux Mint and 60.9.0ESR Windows10.
And also I got a same regression window on Linux Mint as follows.
43:55.65 INFO: No more inbound revisions, bisection finished.
43:55.65 INFO: Last good revision: bf2586dc82562136449ebe14bbc14b609c20c245
43:55.65 INFO: First bad revision: 1572223bd5ecd9ae000be6c766cece6533972735
43:55.65 INFO: Pushlog:
https://hg.mozilla.org/integration/autoland/pushloghtml?fromchange=bf2586dc82562136449ebe14bbc14b609c20c245&tochange=1572223bd5ecd9ae000be6c766cece6533972735

So, I misunderstand STR comment#0....

Comment 7

[:Gijs (he/him)]

OK, so clearly there was some confusion. Let's try and actually get to the bottom of the issue here - I filed bug 1584539 for the recent regression.

Comment 8

[:Gijs (he/him)]

Reporter:

Can you reproduce the issue with a more recent version of Firefox? Firefox 60 is quite old at this point; no more updates will be released.

If so, can you try:

1. open the developer tools console (tools > web developer > console)
2. open the settings (icon with the three dots on the top right of the devtools, then click settings)
3. tick "Enable browser chrome and add-on debugging toolboxes" (bottom right in the settings)
4. open the browser console (ctrl-shift-j)
5. evaluate:

Services.dirsvc.get("Desk", Ci.nsIFile).reveal()

do you get a file explorer (thunar) window, or does something else happen?

If you create a temporary new profile ( https://support.mozilla.org/kb/profile-manager-create-and-remove-firefox-profiles ) does the same thing happen?

Comment 9

[Marco Bonardo [:mak]]

Please confirm which version of firefox you are using

Comment 10

[:Gijs (he/him)]

It's been 3 weeks since I asked for more details - bug 1584539 is fixed, so let's assume it's the same issue for now, and we can reopen if more details emerge later.