Seen this on a few sites lately, with Firefox 0.10. Filing under ImageLib as someone on IRC "thinks" zoom is a Seamonkey feature too. Observable at: http://www.nytimes.com/nytstore/clothing/NSNYTBC.html Click 'view larger image', notice image is slightly zoomed out in pop-up window. Click to bring to 100%, and select image with mouse to see borders. Notice it has plenty of room to fit in the pop-up without zoom.
Mozilla/5.0 (Windows; U; Windows NT 5.1; rv:1.7.3) Gecko/20040913 Firefox/0.10 I see similar behavior, but with some different details. When the popup comes up, if I left-click on the image (with +zoom cursor), I see barely any change. Also, the cursor remains +zoom. If I move the mouse pointer out of the window and back in again, it becomes a -zoom cursor. If I click a second time, the image remains unchanged and the cursor remains -zoom" Moving the mouse out and back in the window changes the pointer to +zoom. I see no scroll borders in any case.
Yes, those are actually the very same "details" I see on Linux. (er, that sounds snippy... only meaning to clarify). Two comments: Yes, the +zoom/-zoom cursor doesn't update immediately, but I only need to move it 1 pixel for it to change, not out of the window and back in. Second, there is "barely any change" because, as you can see if you enlarge the pop-up window, it is only scaled to 99%. I highly suspect this another observable effort of Gecko's poor margin creation when directly viewing images. (see also bug 136002). It seems likely that the margins Gecko actually displays the image with differ from those it "thinks" it's displaying the image with, making Gecko believe scaling is necessary. (or s/Gecko/imglib/, or pr0nlib, or whatever the backend responsible for this is called nowadays).
13 years ago
Nominating for 1.0. This is really, really poor (non IE parity) layout we're doing on a pretty common web site "feature". I think someone should at least see what a fix would involve--maybe it's just a px value adjustment.
mostly looking for finding and fixing of regressions at this point... do you know if this is a recent regression? renominate if so..
*** This bug has been marked as a duplicate of 224293 ***