160 bytes, text/html
101 bytes, text/html
5.47 KB, patch
Dan M: review+
Christopher Aillon (sabbatical, not receiving bugmail): approval-aviary1.0.1+
Christopher Aillon (sabbatical, not receiving bugmail): approval1.7.6+
|Details | Diff | Splinter Review|
I can confirm the behavior with Mozilla 1.8b build 2005011805 on WinNT4 and as stated by the reporter I don't see it with Firefox 1.0. Steps to reproduce: 1. open the given URL 2. click on the image left to the second project called " Flügeleien". I see the following warning in the JS console: Warning: variable Seite hides argument Source File: http://www.uni-bayreuth.de/departments/kunsterziehung/didaktik/scripts/global.js Line: 39, Column: 6 Source Code: var Seite = "../scripts/zoomimage.php?img=" + Seite;
Version: unspecified → Trunk
Looks like to be the same problem like in Bug 278916, see bug 278916 comment 4.
Well, I can reproduce this bug, but NOT bug 278916 on Linux. So they're likely different problems. For this bug, it looks like the frameset may be relevant -- if I just load that subframe as a toplevel page the bug doesn't appear.
Created attachment 172585 [details] [diff] [review] Patch This basically backs out the "fix" for bug 135811. Now I can't figure out how finding targets outside our window for subframes _ever_ worked after the fix for that bug.... jst, any idea? In any case, what happens is that the subframe asks its parent to look, the parent doesn't find the target amongst its kids, but has the same tree owner as aRequestor (the kid), so doesn't ask the treeowner either. So the named target is never found. Before the patch for bug 103638, we had basically the same code here, which is why I don't understand how it worked... :( I'd really like to, if someone can explain it to me. In any case, this patch also adds code to our treeowners to ensure the invariant that the treeowner is the requestor only for calls to the root docshell of type. This should make sure bug 135811 does not reappear. I didn't change nsDocShellTreeOwner.cpp because I'm still trying to figure out how that one works, exactly. It'll probably be the subject of bug 278916.
13 years ago
Comment on attachment 172585 [details] [diff] [review] Patch Ugh, I can't explain how this worked... not w/o digging through old code I don't have handy. sr=jst nonetheless.
Attachment #172585 - Flags: superreview?(jst) → superreview+
I _did_ dig through the code before our patch landed... and I still don't see how it worked. :(
Component: General → Embedding: Docshell
Product: Mozilla Application Suite → Core
QA Contact: general → adamlock
Checked in for 1.8b.
Status: ASSIGNED → RESOLVED
Last Resolved: 13 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla1.8beta
For reference: Disable Targets For Downloads eliminates a common error incurred by many webmasters, which ultimately causes blank windows to be opened when you try to download a binary file. http://www.cusser.net/extensions/
13 years ago
Comment on attachment 172585 [details] [diff] [review] Patch a=caillon (on behalf of drivers) for checkin to 1.0.1 and 1.7.6
Fixed on 1.0.1 and 1.7 branches.
Verified Fixed everywhere with the recent builds (FF101, FFTrunk, M176, Trunk).
Status: RESOLVED → VERIFIED
10 years ago
Test for this got added in bug 408052.
You need to log in before you can comment on or make changes to this bug.