633768 - Strange zoom behaviour with iframe containing only an image

Status: RESOLVED INCOMPLETE

(Core :: Layout, defect)

Description

[Tobias (:Tobbi) Markus]

When zooming via View > Zoom > Zoom In / Zoom Out, the zooming is quite strange:

When zooming in, the image gets minimally SMALLER, when zooming out, the image gets larger and at a specific point resizes to a certain size.

As far as I can see, the issue has to do with the iframe. Because when I open the iframe source site directly, I cannot reproduce the issue.

Comment 1

[Mats Palmgren (inactive)]

Sounds a bit like bug 633164.  Can you check if it's the same regression range?

Comment 2

[Tobias (:Tobbi) Markus]

(In reply to comment #1)
> Sounds a bit like bug 633164.  Can you check if it's the same regression range?

As far as I can tell, it's not the same regression range. 

I have tested nightlies dating back till 2010-12-13 and the bug was already present at that time.

Comment 3

[Timothy Nikkel (:tnikkel)]

I think this might be the "correct" behaviour, however unintuitive it is.

Comment 4

[Boris Zbarsky [:bzbarsky]]

Yeah, the problem is that in this case the image is not being shown at its actual size, but scaled to the size of the frame.  And zooming in actually makes the frame smaller (this may be a bug?).

If you click on the image to show at normal size and then zoom in, the image gets bigger.

Comment 5

[Timothy Nikkel (:tnikkel)]

(In reply to comment #4)
> Yeah, the problem is that in this case the image is not being shown at its
> actual size, but scaled to the size of the frame.  And zooming in actually
> makes the frame smaller (this may be a bug?).

I don't know if the frame gets smaller but the padding around the image gets bigger (because it is zoomed), while the image is resized to fill the space again.

> If you click on the image to show at normal size and then zoom in, the image
> gets bigger.

I thought we had some code to not allow the auto-zooming of images when in iframes, maybe it doesn't get applied to framesets? Maybe it should?

Comment 6

[Boris Zbarsky [:bzbarsky]]

I see nothing dealing with iframes in nsImageDocument.

Comment 7

[Timothy Nikkel (:tnikkel)]

Looks like I was wrong about that. We do do auto-zoom on images in iframes.

Comment 8

[Tobias (:Tobbi) Markus]

Moving testcase to somewhere else to be able to make changes to the front page.

Comment 9

[Alice0775 White]

If I do not misread STR and ACTUAL and EXPECTED RESULT, the regression window is as fllows.

Regression window:
Works:
http://hg.mozilla.org/mozilla-central/rev/123e48c8edf4
Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.3a1pre) Gecko/20091014 Minefield/3.7a1pre ID:20091014045102
Fails:
http://hg.mozilla.org/mozilla-central/rev/04b1081dbb5d
Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.3a1pre) Gecko/20091015 Minefield/3.7a1pre ID:20091015051159
Pushlog:
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=123e48c8edf4&tochange=04b1081dbb5d
Triggered by:
af032ca9e9ec	Robert O'Callahan — Bug 512367. Don't reset fullZoom level when zooming in/out of an image loaded in an <iframe>. r=smaug

Comment 10

[Thomas Ahlblom]

Bug 658333 describes a similar behavior for images opened directly in Firefox, and I tracked it down to af032ca9e9ec, thus the same changeset as in this bug.

Should we duplicate that bug against this one or keep them separate?

Comment 13

[roxleitan]

20160502172042 Mozilla/5.0 (Windows NT 5.1; rv:46.0) Gecko/20100101 Firefox/46.0
20160523030225 Mozilla/5.0 (Windows NT 5.1; rv:49.0) Gecko/20100101 Firefox/49.0

I have tested your issue on latest FF release 46.0.1 and latest Nightly build and could not reproduce it.
Is this still reproducible on your end ? If yes, can you please retest this using latest FF release and latest Nightly build (https://nightly.mozilla.org/) and report back the results ? When doing this, please use a new clean Firefox profile, maybe even safe mode, to eliminate custom settings as a possible cause (https://goo.gl/PNe90E).

Comment 14

[roxleitan]

Considering the fact that I cannot reproduce this I will mark this as Resolved-Worksforme.

If anyone can still reproduce it, feel free to reopen the issue and provide more information.

Comment 15

[Boris Zbarsky [:bzbarsky]]

What testcase did you use to try to reproduce this?

Comment 16

[roxleitan]

20160710030217 Mozilla/5.0 (Windows NT 6.1; rv:50.0) Gecko/20100101 Firefox/50.0

Hi Boris,

I tried to reproduce this issue using testcase from comment 8: http://dreg.bplaced.net/testcases/bug633768.html on latest Nightly build.

Comment 17

[Boris Zbarsky [:bzbarsky]]

> I tried to reproduce this issue using testcase from comment 8:

That link goes to a "there is nothing here" page.  It doesn't have a testcase anymore.  Please try with an actual testcase that represents this bug.

Comment 18

[roxleitan]

Attachment: testcase1.html

20160711034039 Mozilla/5.0 (Windows NT 5.1; rv:50.0) Gecko/20100101 Firefox/50.0

Hi Boris,

I tested this issue using attached testcase on latest Nightly build and could not reproduce it.
Please confirm that this testcase is ok.

Thanks.

Comment 19

[roxleitan]

I don't know why the image is missing from the attached testcase. Here is the link: http://jsbin.com/mabimitono/edit?html,output

Comment 20

[Boris Zbarsky [:bzbarsky]]

> I don't know why the image is missing from the attached testcase.

Because the image is at http://something, the testcase at https://something, and mixed content blocking blocks the image load.  What you want to do generally is attach the image to the bug, then link to it from the testcase.

Anyway, that testcase doesn't have the "iframe gets smaller on zoom" behavior the original testcase in this bug did.  There are various ways of producing that, I think (e.g. an iframe that takes up all the space left over after some text that gets bigger with zoom is laid out) and I expect thsoe ways to still reproduce this bug as reported.  The only question is whether this is wontfix, I think.

Comment 21

[roxleitan]

Marking this as Resolved: Incomplete due to the fact that original testcase is not available anymore and due to the lack of response from the reporter.

If anyone can still reproduce it on latest versions, feel free to reopen the issue and provide more information.