Hubert, could you take a look at this?

Hubert, do you still want to work on that?

Yes, I would love to but I'm a bit overwhelmed with my teaching. I know this feature is important. What's the timeframe for FF 20? I already had a look at how it could be implemented. I think a weekend would do the trick (including tests...) but I won't be able to do it before the 21 dec. What do we do?

No hurry. No deadline. > What do we do? We wait until you have time :)

Why did I ask? Mozilla I love you!!

Sorry Hubert, I couldn't wait :D I just picked this for my first bugfix as it seemed pretty easy. As I've said, it's my first bugfix here, please evaluate the patch and tell me what I have done wrong :D PS: I don't know how to send a binary file(png) along with the patch, so in order to this patch work correctly, the file browser/themes/<platform>/devtools/responsive-vertical-resizer.png should be cloned, rotated(90°) and renamed as responsive-horizontal-resizer.png. Sorry for the inconvenience :(

Just in case someone needs the png file.

Hey André, no worries. I've been pretty busy these days. Thx for your submission...

Comment on attachment 772454 [details] [diff] [review] Patch that solves the bug André, thanks a lot for your help. To add a file (binary or not), just do "hg add path/to/the/file". Instead of creating a new file though, I'd suggest that you use responsive-vertical-resizer.png and you apply a CSS transform to it (transform:rotate(90deg)). You might want want to rename this file to "responsive-bar-resizer.png". You use translate(12,12). You want to use translate(-12,12) (to align horizontally the image). In the function `onDrag`, you should do: `if (this.ignoreX) deltaX = 0`. We don't want to change the width with this resizer is used.

Paul, thanks for the directions. I did as you said, but the results of css transform are not nice. I think it's a good idea to reuse a resource, but in this case there's a deterioration of the image quality. Check the attachment.

Comment on attachment 772963 [details] Image vs CSS Transform comparison This is weird. And it should not happen. Can I ask you to try to reproduce this bug within a web page?

I've setup a small test, please take a look. Same behavior under Chromium 26.

Interesting. What happens if you add `image-rendering: -moz-crisp-edges`?

The issue is because of odd number of pixels in width. And yes, `image-rendering: -moz-crisp-edges` fixes the issue.

Great, now it looks nice. What do you think about "rotate(90deg) translate(20px, -12px)", these values were based on trial and error, do think there's a better way to do it? Latter I'll prepare another patch. Once again thanks.

You should use `translate(12px, -12px) ` to match what we do for the other resizer. But start with translate then rotate `translate(12,-12) rotate(90deg)`. I think this will work.

Once again something strange happened. I've followed your instructions and `translate(12,-12) rotate(90deg)` doesn't work as it should(see attachment). Then I tried `rotate(90deg) translate(20px, 5px)` and worked just fine under Linux. Then I hacked the onmi.ja under Windows and while dragging the bar it becomes weird(see attachment). At work's machine, I also hacked the onmi.ja, but it seems ok. Any guess? My box: Archlinux | Windows 64-bit Ultimate - Core2Duo + NVidia GTX460 Work's box: Windows 64-bit Professional - Phenom II + Radeon HD 4250

Can you attach the patch?

This patch uses the css transform rotate. Works, but only on my PC has shown some bugs.

This patch creates new rotated image. Works fine on every PC I've tested.

Comment on attachment 777560 [details] [diff] [review] Patch new-image Let's go for this approach. Thank you André.

This orange (bug 866671) is triggered on 4 platforms with this patch. Let's try with a more recent tree: https://tbpl.mozilla.org/?tree=Try&rev=c3646bda31c8

That's better.

s/12px/-12px/ https://hg.mozilla.org/integration/fx-team/rev/eb7ff2beda97

While testing the nightly, I noticed a problem when resizing with the new resizer and then rotate, it causes the resizer to move to an incorrect position. Patch attached.

(In reply to André Miranda from comment #29) > Created attachment 790619 [details] [diff] [review] > patch-fix-rotate > > While testing the nightly, I noticed a problem when resizing with the new > resizer and then rotate, it causes the resizer to move to an incorrect > position. > Patch attached. Thank you! Can I ask you to file a new bug and attach this patch?

Done: bug 906025