This isn't occurring for me with a build I built this morning. Could you retry the next time you get a build?
Unfortunately I could not verify this with the build from 7 September - it crashes at starting. Shall try with a newer build.
It fails for me now. Thanks for retrying.
I'll reassign this to you while you work on the document part of the fix.
Checked in the document fix (document was incorrectly being parented by the global object of the calling context, not the containing context). The other places that need to be similarly fixed are included in the list generated by http://lxr.mozilla.org/seamonkey/ident?i=GetGlobalObject.
We can fix the other places after beta.
Vidur fixed nsHTMLDocument::GetScriptObject
Fixed the use of GetGlobalObjet for the location obejct.
See comments on bug 14916, looks like this bug may be fixed?
The test case from firstname.lastname@example.org no longer fails. The bug is still open pending the review of GetGlobalObject calls. I'm changing the title to reflect that.
Push security review tasks off until M16.
Bulk moving all Browser Security bugs to new Security: General component. The previous Security component for Browser will be deleted.
Changing Qa contact to myself.
Bulk reassigning most of norris's bugs to mstoltz.
Security reviews and denial-of-service attacks. These will be addressed in the post-beta2 timeframe (unless someone's interested in tackling them earlier?)
Assigning QA to czhang
Mass changing QA to ckritzer.
> Checked in the document fix (document was incorrectly being parented by the > global object of the calling context, not the containing context). The other > places that need to be similarly fixed are included in > the list generated by http://lxr.mozilla.org/seamonkey/ident?i=GetGlobalObject. Old bug. My guess is that this is no longer a problem, but I wanted to run it by you guys. cc'ing vidur and jst. Are we still using the wrong parent object anywhere?
I didn't look through all the callers of GetGlobalObject, but if there's a problem somewhere I would think that people would be complaining, and I don't see people complaining so I say close this one, unless someone is interested in reviwing all the callers...
I think we're probably OK.
Verified on build: 2001-09-13-0.9.4 platform: Win NT Following tests passed: 1. http://www.nat.bg/~joro/mozilla/oversec.html For above link, 2 link buttons appears. Link1 opens the about:blank window. This is the correct behavior. Link2 is for opening the about:cache window, which should not be allowed. And the window is not opened as expected. Also, an error is written in the JS console: "The link to about:cache was blocked by the security manager. Remote content may not link to local content." 2. The test script attached also works as expected. An alert dialog is opened as expected.
16 years ago