897626 - Investigate performance impact of non-autohiding menubar on Windows XP

Status: RESOLVED WONTFIX

(Firefox :: Toolbars and Customization, defect)

Description

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

Windows XP suffers from a higher tpaint cost than the other Windows versions. One of the key differences is the presence of the visible menubar, whereas Vista and 7 autohide the menubar by default.

I'm not sure why we are keeping the menubar visible on XP now that we have switched away from the orange Firefox button.

The attached patch should keep the menubar autohidden on all Windows platforms.

Comment 1

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

Attachment: 897626.patch

Comment 2

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

Pushed to tryserver https://tbpl.mozilla.org/?tree=Try&rev=830354564f0c

Comment 3

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

Baseline try push, https://tbpl.mozilla.org/?tree=Try&rev=fdf4d28c6198

Comment 4

[Dão Gottwald [:dao]]

(In reply to Jared Wein [:jaws] from comment #0)
> Windows XP suffers from a higher tpaint cost than the other Windows
> versions. One of the key differences is the presence of the visible menubar,
> whereas Vista and 7 autohide the menubar by default.

Of course that wouldn't explain the tpaint regression on the UX branch compared to m-c, as the menu bar is visible on m-c too.

Considering how autohiding works, I'm also skeptical that it makes significant a difference for tpaint, but I guess you try run will clarify that.

> I'm not sure why we are keeping the menubar visible on XP now that we have
> switched away from the orange Firefox button.

See bug 879599.

Comment 5

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

Comment on attachment 780543 [details] [diff] [review]
897626.patch

The numbers that came back weren't conclusive enough to make this change worthwhile. Averaging the tpaint numbers they went from 147.8 to 152.4 (actually a decrease in performance).

Gijs, could I have done something wrong with this patch?

Comment 6

[:Gijs (he/him)]

(In reply to Jared Wein [:jaws] from comment #5)
> Comment on attachment 780543 [details] [diff] [review]
> 897626.patch
> 
> The numbers that came back weren't conclusive enough to make this change
> worthwhile. Averaging the tpaint numbers they went from 147.8 to 152.4
> (actually a decrease in performance).
> 
> Gijs, could I have done something wrong with this patch?

I don't think so, it looks fine to me.