adfb499dd1
Bug #282103 --> minor dynamic overlay improvements.
...
Add IDL documentation stating that it's ok to pass in a null observer to LoadOverlay
Make sure .Init() succeeds on our hashtables, otherwise return an out of memory error code.
r/sr=bz
2005-08-30 21:57:01 +00:00
f426cc15ee
Stop storing PRIntervalTime values as PRInt64s, and make sure to convert PRIntervalTime to milliseconds before resuming a timeout. Bug 305167, r+sr=jst.
2005-08-30 20:07:58 +00:00
17f1097d7e
bug 306164: The global scope polluter is not reinstalled when document.write is used. r+sr=jst
2005-08-29 22:04:39 +00:00
1226844692
Bug 306160 Fix nsIDOMXULMenuListElement::GetInputField to return a type that matches the element that the XBL is trying to return b=306160 r=mats.palmgren sr=bz
2005-08-28 22:04:43 +00:00
f98e12557e
282103 - better document the dynamic overlay load method, and announce that it will change for 2.0 r=bryner a=cbeard
2005-08-26 19:08:00 +00:00
ec5069804e
Bug 282579 - Implement <svg:textPath>. As a side effect, this removes
...
the gdi+ text highlighting code. r=afri
2005-08-26 02:49:52 +00:00
59e42c6cd2
Fixing bug 302103. Implement SVG events. r=tor@acm.org, sr=jst@mozilla.org
2005-08-25 21:31:09 +00:00
6383a6ee5b
Move resuming of timeouts to after we've restored the docshell hierarchy, so that they're restored correctly for subframes. Bug 305156, r=cbiesinger, sr=jst.
2005-08-25 07:06:59 +00:00
154d3f15fa
Bug 249136. Cannot scroll in pages loaded into foreground tabs. Builds on top of previous patch by attempting to fix perf regressions that caused. r=mats, sr=bryner
2005-08-25 00:55:49 +00:00
197c551676
Bug 249136. Focus lost in newly loading foreground tabs. r=mats, sr=bryner
2005-08-24 00:31:24 +00:00
6377fe1d40
Bug 296036. Remove the very poorly specified (and redundant) field 'point' of nsEvent. r+sr=roc, patch by Eli Friedman
2005-08-23 03:57:07 +00:00
7cb9d2fece
bug 304882: Store the location object on the outer window instead of the inner window so that security checks do the Right Thing. Also change some if statements into assertions so we catch bad states. r+sr=jst
2005-08-23 01:42:49 +00:00
f6f8514c0c
bug 305236: Don't store the current document in the bfcache if we're going to reuse the current inner window for another page. In this case, the current document never was making it into session history anyway. r=bryner sr=jst
2005-08-23 00:24:57 +00:00
0d6a610963
Make nsIPresShell::GetPrimaryFrameFor return nsIFrame* instead of using an out
...
param. Bug 303779, patch by Bastiaan Jacques <b.jacques@planet.nl>, r+sr=bzbarsky
2005-08-22 22:24:29 +00:00
1a8f9d6e76
Bug 302729 attachment 193230, Convert [tag][/tag] in netError.dtd strings to <tag></tag>, patch by Jeff Walden <jwalden+bmo@mit.edu>, r=cbiesinger, sr=neil, a=cbeard
2005-08-22 03:25:20 +00:00
d028e7fd2c
Fixing bug 303765. Don't ever reuse a listener manager on a new inner window. r=sfraser_bugs@smfr.org, sr=brendan@mozilla.org
2005-08-19 22:56:20 +00:00
edac52e953
Fixing bugs 303981 and 304459. Make sure we don't leave window.document pointing to the old document when reusing an inner window, and make sure we don't let Object.prototype leak into a new inner window. r=mrbkap@gmail.com, sr=brendan@mozilla.org
2005-08-19 20:09:38 +00:00
0b8f97b416
bug 304896: Null check mNavigator when we set it, since it isn't guaranteed to be non-null. r+sr=jst
2005-08-19 17:11:30 +00:00
2273edc3a7
Turn off some more classinfo helpers for XPCNativeWrapper. Bug 304886, r+sr=jst
2005-08-19 15:12:36 +00:00
c36cb31367
Bug 304918 don't use cx to mean something other than JSContext cx in DOM
...
also fixing "handles"
r=jst sr=jst
2005-08-19 09:46:25 +00:00
af539a8b76
Backing myself out to see if it fixes the Seamonkey bustage (original checkin was for bug 304882).
2005-08-19 00:52:59 +00:00
ede53a3cce
Fixing bug 304249. Make sure to push cx onto the context stack in nsJSContext::BindCompiledEventHandler() to ensure any security checks etc that happen within it see the right context. r=mrbkap@gmail.com, sr=brendan@mozilla.org
2005-08-18 19:41:44 +00:00
d9703b4bfd
bug 304882: Store the location on the outer window and parent it to the outer window's global JS object so that security checks work as expected. This allows us to let the location object always have a valid docshell. r+sr=jst
2005-08-18 19:03:10 +00:00
1eb00e476b
bug 304896: Reset the docshell on the navigator object when pulling it out of the bfcache to ensure that it gets the right global object parent. r+sr=jst
2005-08-18 18:43:33 +00:00
4be57eb8e0
Fix dumb typo.
2005-08-17 18:17:39 +00:00
9488bc2a6d
Bug 304338 nsWindowSH doesn't allow xpcom components to replace its properties
...
r=jst sr=jst
2005-08-17 06:01:42 +00:00
5abc902ee0
Make sure to propagate out error returns when initing classes. Bug 294432, r+sr=jst.
2005-08-16 22:54:59 +00:00
e9fbf5f581
Fixing bug 287008. Don't call a method that just asserts and returns NS_OK when we can just return NS_OK. Cut n' paste caused this one... r+sr=peterv@propagandism.org
2005-08-16 21:52:52 +00:00
2ceeb6b6ec
Fixing regression from bug 296639. When accessing frame objects by index, wrap the outer frame, not the inner. r=mrbkap@gmail.com, sr=shaver@mozilla.org, a=asa@mozilla.org
2005-08-16 06:38:43 +00:00
77e6b8d4ac
Make DEBUG_PAGE_CACHE compile. NPOTB.
2005-08-15 23:35:14 +00:00
f0d2e05a7e
Bug 299992 (re-landing) - Move pldhash etc. into the xpcom glue, and centralize various instances of nsINIParser in the xpcom glue. Some files were cvscopied from xpcom/ds to xpcom/glue r=brendan/darin
2005-08-15 18:29:55 +00:00
2b7c7bca7a
bug 303267: Store the entire inner window in the window state holder instead of copying properties and restoring them. This fixes going back and forth with the bfcache on when the page gone to has javascript in it. r=jst sr=bryner
2005-08-15 18:16:42 +00:00
c7191e818a
Make the one-arg version of scrollIntoView scriptable in IDL. Bug 296965,
...
r+sr=jst.
2005-08-15 02:37:14 +00:00
20edc5ab39
Missed file from checkin for bug 271802: whether to extend print dialog should not be part of localization. r=bsmedberg
2005-08-13 19:42:37 +00:00
c5688dfef2
Backing out changes for bug 303267 until the perf hits on the Seamonkey tinderbox can be explained.
2005-08-13 03:54:29 +00:00
59fbced3fe
bug 303267: Store the entire inner window object on the history entry instead of copying/restoring JS properties. r=bryner sr=jst. This patch also fixes bugs 304284 and 304078.
2005-08-13 00:53:39 +00:00
e23ea875f3
Followup changes to bzabarsky's review for bug 296639. Reviews pending, a=drivers@mozilla.org
2005-08-12 04:11:00 +00:00
8992fb4c7d
Backout bug 299992 - too much odd platform-specific bustage
2005-08-11 22:07:08 +00:00
65d2ab642f
Additional bustage fix, bug 299992
2005-08-11 21:02:28 +00:00
65891781d9
Release timeouts when a cached page is evicted from session history. Bug 302290, r=darin sr=jst a=brendan
2005-08-10 20:10:18 +00:00
112e7e0cd6
Fixing bug 303420. Making the global scope polluter work again, it was failing due to security checks failing when performed before the context and global objects were fully initialized. r=mrbkap@gmail.com, sr=bryner@brianryner.com, a=cbeard@mozilla.org
2005-08-10 00:15:51 +00:00
8ce85b83a1
Bug 304003 Bfcache does not work anymore
...
patch by biesi r=bryner sr=smfr a=chase
2005-08-09 21:29:23 +00:00
9a35b089ca
Fix gc crash if we ever failed to copy a window's properties. Bug 298459, r=darin, a=brendan.
2005-08-09 01:46:01 +00:00
cb4ed7c2b3
Fix bug 302962. When someone calls Activate() or Deactivate() on an nsWebBrowser which doesn't have a content viewer yet (and hence no pres shell), don't just bail because nsGlobalWindow will complain. Instead, change nsDocShell to make the content viewer on demand, and go ahead and pass the activate/deactivate onto the dom window. This fixes a serious focus bug in Camino.
...
r/sr=bryner, a=bsmdedberg.
2005-08-06 20:28:42 +00:00
006c256aff
Use the target window's outer window when we're checking for the focused window for fastback. Bug 298077, r=dbaron a=bsmedberg
2005-08-06 03:17:28 +00:00
b396c88c41
Fixing bug 302889, and more tinderbox orange (balsa). Make window.close() actually close the window. Events that were incorrectly targetted at an inner window caused one of our DOMWindowClose handlers to prevent the close. See the bug for more details. r+sr=dbaron@mozilla.org, a=chase@mozilla.org
2005-08-02 01:23:40 +00:00
91e96f10ed
Fixing crash in nsFocusController::GetWindowFromDocument. Patch by allan@beaufour.dk, r+sr=jst@mozilla.org, a=benjamin@smedbergs.us
2005-08-02 01:14:00 +00:00
f89acc7b89
Switch order of first two base classes to work around btek orange.
2005-07-31 19:43:27 +00:00
9096d968f8
Fixing orange on tinderbox. The problem was that when we were tearing down a window and releasing its document we didn't remember the document principals, so any security checks that happened after a window was torn down failed. Partial backout of the fix for bug 296639. r=dbaron@mozilla.org
2005-07-31 16:44:28 +00:00
b83b23cb6b
Fixing win32 build bustage.
2005-07-30 22:06:42 +00:00
scott%scott-macgregor.org
bryner%brianryner.com
mrbkap%gmail.com
neil%parkwaycc.co.uk
ben%bengoodger.com
tor%cs.brown.edu
jwatt%jwatt.org
aaronleventhal%moonset.net
roc+%cs.cmu.edu
bzbarsky%mit.edu
gavin%gavinsharp.com
jst%mozilla.jstenback.com
timeless%mozdev.org
bsmedberg%covad.net
dbaron%dbaron.org
smfr%smfr.org