941430 - Max tab width is smaller in Australis

Status: VERIFIED FIXED

(Firefox :: Tabbed Browser, defect)

Description

[Bill McCloskey [inactive unless it's an emergency] (:billm)]

At first I thought this was a design issue, but I asked MattN and he doesn't think it is. Before Australis, if I had seven tabs open, they would take up the entire horizontal width of the window. Now there's just a lot of empty space on the right side. It's annoying because I can't see as much of the tab title as before.

Screenshots will follow.

Comment 1

[Bill McCloskey [inactive unless it's an emergency] (:billm)]

Attachment: screenshot-old.png

Before Australis.

Comment 2

[Bill McCloskey [inactive unless it's an emergency] (:billm)]

Attachment: screenshot-new.png

And the after picture.

Comment 3

[Bill McCloskey [inactive unless it's an emergency] (:billm)]

The screenshots have the same number of tabs and the same window size.

Before Australis:
Width of a tab: ~224px
Usable width for text: ~172px

Australis:
Width of a tab: ~181px
Usable width for text: ~115px

So the tab width went down by about 20% and the text width went down by about 30%. I can file the text issue as a separate problem if need be.

Comment 4

[Matthew N. [:MattN]]

(In reply to Bill McCloskey (:billm) from comment #0)
> At first I thought this was a design issue, but I asked MattN and he doesn't
> think it is. Before Australis, if I had seven tabs open, they would take up
> the entire horizontal width of the window. Now there's just a lot of empty
> space on the right side. It's annoying because I can't see as much of the
> tab title as before.

It is part of the design but my point was more that I hadn't seen much discussion around this case. Many people who have been complaining about the titles of tabs have had tab overflow and so the *minimum* width was the issue. In your case you're interested in the *maximum* width.

I double-checked and we have the correct dimensions according to the spec [1].

Stephen, was the reduction in max-width intentional? Would you consider increasing the max-width? If not, please resolve the bug.

[1] https://people.mozilla.org/~shorlander/files/australis-designSpecs/australis-designSpecs-linux-mainWindow.html

Comment 5

[Matthew N. [:MattN]]

Attachment: Screen Shot with spec on OS X

Comment 7

[tabmix.onemen]

Tabs are overlaping by 15px, so maybe we need to set max-width to 265px

Comment 8

[Bill McCloskey [inactive unless it's an emergency] (:billm)]

I just wanted to note that I changed max-width for .tabbrowser-tab:not([pinned]) in my userChrome.css file. The tabs do get wider, but when I start opening and closing tabs, they appear at the wrong places. There must be other places in the code that I would have to change :-(.

Comment 9

[Jared Wein [:jaws] (please needinfo? me)]

(In reply to Bill McCloskey (:billm) from comment #8)
> I just wanted to note that I changed max-width for
> .tabbrowser-tab:not([pinned]) in my userChrome.css file. The tabs do get
> wider, but when I start opening and closing tabs, they appear at the wrong
> places. There must be other places in the code that I would have to change
> :-(.

You may want to see bug 875180.

Comment 10

[(Currently slow to respond) Philipp Sackl [:phlsa] (Firefox UX) please use needinfo]

Attachment: max-width set to 220px

I think we could set the maximum tab width to something like 220px without any aesthetic fallout.

Comment 11

[u428464]

(In reply to Philipp Sackl [:phlsa] from comment #10)
> Created attachment 8385646 [details]
> max-width set to 220px
> 
> I think we could set the maximum tab width to something like 220px without
> any aesthetic fallout.

It's quite bigger than the current tab at 180. AFAIK Google Chrome has more or less the same width.

The Firefox pre-Australis was 250, but the tabs were huge. I think 200 or 210 would be perfect.

Comment 12

[Dão Gottwald [:dao]]

Attachment: patch

Comment 13

[Matthew N. [:MattN]]

Comment on attachment 8386129 [details] [diff] [review]
patch

Review of attachment 8386129 [details] [diff] [review]:
-----------------------------------------------------------------

This may affect TART results so you should probably do a Try build before landing to be safe (even if it's just to be prepared to say that the change in result is fine).

r=me with the changes and assuming no major TART regression.

::: browser/base/content/browser.css
@@ +103,5 @@
>  }
>  
>  .tabbrowser-tab:not([pinned]) {
>    -moz-box-flex: 100;
> +  max-width: 220px;

phlsa found that 210px looked better so please use that max-width

@@ +108,4 @@
>    min-width: 100px;
>    width: 0;
>    transition: min-width 200ms ease-out,
> +              max-width 240ms ease-out;

phlsa says we could leave this as it was

Comment 14

[Dão Gottwald [:dao]]

(In reply to Matthew N. [:MattN] from comment #13)
> This may affect TART results so you should probably do a Try build before
> landing to be safe (even if it's just to be prepared to say that the change
> in result is fine).

I think we can just land this, especially if we keep the transition duration as is -- if there's a reasonably small change due to a few more pixels being drawn, we can disregard that as expected. If there's a larger change, we can still back out / investigate after the fact.

Comment 15

[Matthew N. [:MattN]]

I would prefer to avoid the churn so some other regression doesn't slip in before the backout. It doesn't take that long to do a talos run on try

Comment 16

[Dão Gottwald [:dao]]

Attachment: patch v2

Comment 17

[Dão Gottwald [:dao]]

http://perf.snarkfest.net/compare-talos/index.html?oldRevs=26721e371d9d&newRev=8be8dc1033e2&submit=true

Comment 18

[Dão Gottwald [:dao]]

https://hg.mozilla.org/integration/fx-team/rev/a31dbdb555db

Comment 19

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/a31dbdb555db

Comment 20

[Dão Gottwald [:dao]]

Comment on attachment 8386284 [details] [diff] [review]
patch v2

[Approval Request Comment]
Bug caused by (feature/regressing bug #): bug 826689
User impact if declined: less room for tab labels
Testing completed (on m-c, etc.): landed on m-c
Risk to taking this patch (and alternatives if risky): low
String or IDL/UUID changes made by this patch: none

Comment 21

[Dão Gottwald [:dao]]

https://hg.mozilla.org/releases/mozilla-aurora/rev/b6cfb57e477d