5a8866260b
Additional tweak to make session history a little safer. Bug 305181, r=bryner, sr=darin
2006-02-23 05:30:25 +00:00
efc254dc2d
Move the load type constants into nsDocShellLoadTypes.h, for easier reuse outside of docshell. Bug 326706, r+sr=bzbarsky.
2006-02-11 00:17:36 +00:00
2dbdd7da0d
Move the "force into tab" code out of Gecko and allow embeddors to do the same
...
sort of thing. Bug 323810, r=bsmedberg, sr=jst
2006-02-07 20:46:39 +00:00
c235c2652b
Bug 326107 r=darin, sr=bzbarsky, branch-1.8.1=darin:
...
Send referrer to GlobalHistory even when it isn't being sent over the network
2006-02-07 02:02:56 +00:00
b3e8edec7e
Bug 317843 r=darin, sr=bzbarsky: Better redirect tracking for global history.
2006-02-06 20:52:18 +00:00
0c00f57e07
Add a global limit to the number of cached content viewers that scales with the amount of physical memory. Patch by Marria Nazif <marria@gmail.com>. Bug 292965, r=biesi, sr=me.
2005-09-23 18:16:40 +00:00
59513aa496
observe the "browser.xul.error_pages.enabled" pref so that it takes effect
...
immediately
bug 282050 patch by Vidar Braut Haarr <vhaarr+bmo@gmail.com> r=biesi sr=jst
2005-09-22 20:38:56 +00:00
33955684cd
Backing out 292965 until Tp regression is understood.
2005-09-14 07:28:05 +00:00
300a390b7b
Add a global limit to the number of cached content viewers that scales with the amount of physical memory. Patch by Marria Nazif <marria@gmail.com>. Bug 292965, r=biesi, sr=me.
2005-09-14 03:38:40 +00:00
ebac95c422
If we created a transient about:blank content viewer, make sure we're locating the root view sibling for that viewer and not the viewer it's replacing. Bug 300533, r=cbiesinger.
2005-08-24 17:27:41 +00:00
6fdb2d611a
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
a692c025e1
fixes bug 301358 "HTTP 'refresh' doesn't load page with bfcache enabled" r=biesi sr=bryner a=bsmedberg
2005-08-11 20:14:00 +00:00
8b08bb3c07
When navigating a subframe, make sure that all docshells have navigated to the correct cloned session history entry. Bug 301397, r=darin, a=bsmedberg
2005-08-03 22:32:10 +00:00
eaa40fe7ac
When restoring a presentation, move everything except adding the request to the loadgroup into the async plevent callback. This fixes the problem with restoring window properties when history.back() is used (with untrusted JS on the stack). Bug 301516, r=darin, a=asa
2005-08-02 18:35:07 +00:00
bd84f51f5d
Bug 286811 don't add ("xul") error pages to global history
...
r=bz sr=darin a=bsmedberg
2005-07-27 12:38:38 +00:00
5e7db39b5d
Fix stuck designMode when using fastback by making sure the DOM window still contains the old document when the STATE_START notification is sent. Bug 300411, r+sr=darin, a=asa.
2005-07-14 16:24:32 +00:00
f4fce88c63
bug 298255: prevent browser screwiness caused by lazy extension authors loading their chrome extension in a content window instead of adding the -chrome commandline option. r=mconnor, sr=jst, a=chofmann
2005-07-07 09:45:47 +00:00
b10035aabc
Fix pending-loads check for fastback so that we detect loads initiated either before or after Stop() is called. Fix assert on reload. Fix problem where the wrong load type was being checked in CanSavePresentation(). Fix problem where we cached a content viewer on the current history entry when it was equal to the new history entry. Bugs 292954 and 292950. r+sr=darin, a=asa.
2005-06-24 20:53:43 +00:00
592dbdc176
Add support for caching content viewers in session history to speed up back/forward (bug 274784). This initial landing has the feature disabled by default; set browser.sessionhistory.max_viewers to the maximum number of pages to cache to enable the feature. r=bzbarsky, sr/a=brendan.
2005-05-04 20:22:32 +00:00
f2c4953613
Move processing of various headers from the content sink into the document so
...
it'll happen for all of our document types. Bug 288921, r+sr=jst
2005-04-05 03:28:54 +00:00
66b19fc60b
Don't fire onLocationChange for initial loads in subframes; reverts old
...
behavior. Bug 285738, patch by William A. Gianopoulos
<wgianopoulos@yahoo.com>, r=jst, sr=bzbarsky
2005-03-30 04:22:00 +00:00
eb65bc6267
Fixing bug 284993. Make sure the new content viewer is properly initialized before fireing location changed notifications. r=bzbarsky@mit.edu, sr=darin@meer.net
2005-03-09 14:31:02 +00:00
ec1ef4b44d
Docshell doesn't need to hold a pointer to the scriptcontext; the window does
...
it. Bug 282807, r+sr=jst
2005-02-20 17:02:02 +00:00
874340e9fb
Make sure that we have an mParentContentListener at all times in docshell. Bug
...
282644, r=biesi, sr=jst
2005-02-18 18:20:42 +00:00
5d749a9a14
157004 don't fire onLocationChange for error pages, and don't change the
...
docshell's currentURI either. be sure to add the failed URL to session history.
r=bz sr=darin
2005-01-31 11:26:31 +00:00
7817f06738
When doing window access security checks, make sure to treat wyciwyg URIs as
...
equivalent to their creators. Bug 270414, r+sr=jst
2005-01-21 16:58:45 +00:00
27d21949cf
Add a webnavigation flag to allow callers to stop content, not just stop
...
network activity. Make nsLocation use this so it doesn't stop content
unnecessarily. Bug 277224, r=darin, sr=jst
2005-01-20 16:57:04 +00:00
92f5ec0f1b
Fixing bug 277564. Make sure to update the lock icon state correctly when dealing with wyciwyg: URLs. r=bugmail@sicking.cc, dveditz@cruzio.com, darin@meer.net, sr=brendan@mozilla.org, a=asa@mozilla.org
2005-01-12 00:00:53 +00:00
1a8e54667c
Fixing bug 103638 (and bug 273699). Targets with same name in different windows open in the wrong window. Patch by bzbarsky@mit.edu and myself, r=danm.moz@gmail.com, bzbarsky@mit.edu, sr=dveditz@cruzio.com, a=asa@mozilla.org
2005-01-11 19:36:40 +00:00
636189a39e
bug 266554 need real referrer internally for various uses. patch by jst, r/sr=dveditz,darin (ported from aviary)
2005-01-05 06:04:46 +00:00
194e16cd98
Make docshell inherit from docloader; clean up related code, move toward
...
eliminating the nsIWebShell, nsIWebShellContainer, nsIDocumentLoader
interfaces and clarifying the other docshell-related interfaces.
r=darin, biesi, sr=jst
2004-12-08 17:32:06 +00:00
c45eef7d26
single window mode aviary branch merge. bug 172962 r=jst,peterv
2004-12-07 16:08:57 +00:00
29050563b3
bug 265135: put internal referrers on channels to fix xpinstall whitelisting the right way. r=darin,sr=jst
2004-10-28 22:52:45 +00:00
f67b37e6c1
Bug 234257 nsIDocumentLoader/nsDocLoader misc cleanup
...
r=bz sr=darin
2004-09-28 20:48:58 +00:00
a2447a60f5
Bug 128398 store referrer in history patch by James Andrewartha
...
<trs80@tartarus.uwa.edu.au> r=biesi sr=neil
2004-09-19 20:50:32 +00:00
c762989259
Bug 258106: nsDocShell should use nsIPrefService APIs instead of nsIPref.
...
Patch by Mikael Parknert (mikael@parknert.se ), r=bzbarsky, sr=roc.
2004-09-06 19:31:22 +00:00
613d104fb5
Land bug 72747 in pieces: Simplify the mechanism used to prevent framesets from having scrollbars. b=72747 r+sr=roc
2004-09-04 19:51:21 +00:00
e4e0192e47
Back out previous round of bug 72747 changes due to Tp regression.
2004-09-04 08:34:55 +00:00
46a0c16189
Land bug 72747 in pieces: Simplify the mechanisms used to prevent framesets from having scrollbars and used to implement the scrolling attribute on frames and iframes. Implement scrolling='yes'. b=72747 r+sr=roc
2004-09-04 06:53:36 +00:00
f353e1e3a4
b=205947, proxy: mailnews does not display proxy auth dialog, r=darin, sr=shaver
2004-09-02 05:13:25 +00:00
67e0ee1634
Back out the rest of bug 72747 due to Tp regression.
2004-08-27 01:26:55 +00:00
0430fe1b67
Implement 'overflow-x' and 'overflow-y' CSS properties, long implemented by IE for Windows and proposed for CSS3. Implement scrolling='yes' and equivalents on frame and iframe. Various pieces of related cleanup (see comment 22 in bug). b=72747 r+sr=roc
2004-08-26 17:11:20 +00:00
cc807b07b1
Change nsIPresContext to nsPresContext globally, follow-up to bug 253470. rs=roc
2004-07-31 23:15:21 +00:00
693923a31f
Landing fix for bug 239202, patch by trev@gtchat.de. Making nsIScriptGlobalObjectOwner not scriptable. r+sr=jst@mozilla.org
2004-07-16 17:03:10 +00:00
4b7cdb04f4
bug 249107 directory listings coming from session history (cache?) displayed in raw form
...
r=jst sr=darin
2004-07-01 17:04:31 +00:00
34eb60a8d1
Fixing regression bug 246923. Bring back to life the fix for bug 13871, and improve on the fix for bug 246448. r=dveditz@cruzio.com, sr=darin@meer.net
2004-06-23 17:42:31 +00:00
08fa94ab3c
Fixing bug 246448. Prevent loading of content into a frame from a origin different than the one owning the frame. r=dveditz@cruzio.com, sr=brendan@mozilla.org
2004-06-15 08:28:52 +00:00
94fc777387
Fixing bug 243213. Don't waste time looking up the nsIURIFixup service for every docshell, and move some members around to make things align up and pack better on some 64-bit platforms. r=bzbarsky@mit.edu, sr=darin@meer.net
2004-05-11 00:12:09 +00:00
692411203a
Bug 236613: change to MPL/LGPL/GPL tri-license.
2004-04-17 21:52:36 +00:00
e5724471fc
When creating a frame after a page has loaded and loading a new document into
...
it, don't try to treat it as a subframe history traversal, since there is no
original history entry involved so the cloning operation fails. Just append
the session history entry for the newly loaded frame to the history entry for
the page as we would have if the load were done during pageload. Bug 237717,
r=adamlock, sr=darin
2004-04-13 01:42:40 +00:00
bzbarsky%mit.edu
bryner%brianryner.com
brettw%gmail.com
cbiesinger%web.de
mrbkap%gmail.com
darin%meer.net
dveditz%cruzio.com
jst%mozilla.jstenback.com
danm-moz%comcast.net
kjh-5727%comcast.net
dbaron%dbaron.org
vladimir%pobox.com
gerv%gerv.net