1327195 - HiDPI: caret in urlbar isn't visible in the end of long text

Status: NEW

(Core :: Layout, defect, P3)

Description

[arni2033]

>>>   My Info:   Win7_64, Nightly 52, 32bit, ID 20161028030204 (2016-10-28)
STR_1:
0. Set DPI -> 125% :
 A) Set DPI level -> 125% in your OS   [OR]
 B) Set layout.css.devPixelsPerPx -> 1.25
1. Select all text on this page (Ctrl+A), copy it to clipboard (Ctrl+C)
2. Open new tab (Ctrl+T), paste text in urlbar (Ctrl+V)

AR:  Text caret isn't visible in urlbar
ER:  Text caret should be visible

This is regression. Good old Firefox 28 is unaffected, <> 52 is affected

Comment 1

[Emil Pasca [:emilpasca]]

Narrowed inbound regression window from [6c37be9c, 79755182] (3 r
evisions) to [6c37be9c, b2354d42] (2 revisions) (~1 steps left)
Oh noes, no (more) inbound revisions :(
Last good revision: 6c37be9cee51e14e1f04ebfb96ab58cc5113c477
First bad revision: b2354d420c2ca6b8ecef1c8cc7a1ed17cca6e1bf
Pushlog:
https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=6c37be9cee51e14e1f04ebfb96ab58cc5113c477&tochange=b2354d420c2ca6b8ecef1c8cc7a1ed17cca6e1bf

Comment 2

[Timothy Nikkel (:tnikkel)]

(In reply to Emil Pasca [:emilpasca] from comment #1)
> Narrowed inbound regression window from [6c37be9c, 79755182] (3 r
> evisions) to [6c37be9c, b2354d42] (2 revisions) (~1 steps left)
> Oh noes, no (more) inbound revisions :(
> Last good revision: 6c37be9cee51e14e1f04ebfb96ab58cc5113c477
> First bad revision: b2354d420c2ca6b8ecef1c8cc7a1ed17cca6e1bf
> Pushlog:
> https://hg.mozilla.org/integration/mozilla-inbound/
> pushloghtml?fromchange=6c37be9cee51e14e1f04ebfb96ab58cc5113c477&tochange=b235
> 4d420c2ca6b8ecef1c8cc7a1ed17cca6e1bf

Are you sure? The changesets in there seem very unlikely to affect this.

Comment 3

[castiel743]

I think I have the correct data.

This is regression from bug 1012752. Regression range:
https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=4b7cd5b39cca85efc4b8350520303605cb59efbe&tochange=b1dbce81bf3b6124577fb46414f811ec5f45f4e0

Comment 4

[Andrew Overholt [:overholt]]

Markus, seems like maybe bug 1012752 caused this. Can you take a quick look?

Comment 5

[Markus Stange [:mstange]]

Ah. This might be tricky to fix. I'll assign it to myself so that I won't lose track of it but I can't promise a quick turnaround here.

Comment 6

[Nathan Froyd [:froydnj]]

Markus, should we just fix-optional this for 54?  i.e. It should be classified as backlog, rather than something we're planning on fixing in the near future?

Comment 7

[Markus Stange [:mstange]]

Yeah, probably a good idea. This is an unfortunate bug, but bug 1012752, which caused this, was much worse.