| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
version used in Viewer 2
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
bit package.
The dbusglib package contains headers from both, libdbus-1 as well as
libdbus-glib-1, but only the shared object file of the latter.
This isn't entirely consistent with how we do this for other packages
that we want to use the system provided ones for: then we link against
prebuilts, but just don't install them. However, if that is working
then it's very likely that this method will work too for a C library,
so I guess it's ok like this.
This package now contains the shared library from
libdbus-glib-1-2_0.76-1_i386.deb (plus symlinks) and the headers from
libdbus-glib-1-dev_0.76-1_i386.deb AND libdbus-1-dev_1.2.1-5+lenny1_i386.deb.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch has the minimum changes needed to be compatible with
LLQTWEBKIT_API_VERSION 2. From llqtwebkit.h:
Version 2:
indra/media_plugins/webkit/media_plugin_webkit.cpp:
* Changed the usage of the event parameters in onClickLinkHref and
onClickLinkNoFollow events slightly. The clicked URI for both should
now be retrieved with getEventUri() instead of getStringValue().
The "target" string in onClickLinkHref is now retrieved with
getStringValue() instead of getStringValue2().
* The contents of getStringValue2() in the onClickLinkHref event is
now a unique ID for the window proxy the click targets.
all changed files:
* Removed the "link target type" concept, since it doesn't really belong
here.
Note that not all changes have been taken into account, since we need
a LOT of code changes for MoaP before that gets relevant; most notably:
"target" and "uuid" aren't used anywhere.
The new LLEmbeddedBrowserWindowObserver::onWindowGeometryChangeRequested
and LLEmbeddedBrowserWindowObserver::onRequestFilePicker are still not
implemented.
Version 1 introduced LLEmbeddedBrowserWindowObserver::onWindowCloseRequested
which is also still not implemented.
|
| |
|
| |
|
| |
|
| |
|
| |
|
|\ |
|
| |\ |
|
| | | |
|
| | | |
|
| | | |
|
| | |\ |
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Not sure this helps with non-baking, but it looked wrong.
Also, only bind local texture before they are covered by a bake.
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Port of SNOW-802 from SG1. Originally ported by Robin cornelius from
SG 2.1.
Also added a paramter can_use_http to the LLTextureFetchWorker
constructor, to avoid messing with locks.
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Ugly work around for http://opensimulator.org/mantis/view.php?id=5081
by adding 1 to the http request range when on affected versions, and
always requesting ranges that start with 0 (viewer2 style).
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Specifically allow for status 416 (unsatisfiable range) and scuessful
responses without status 216 (i.e. whole file at once).
This *should* allow this to work on different servers, as long as they
comply with RFC 2616 WRT range requests.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Treat http texture requests that returned less data than requsted
as indication that we have the whole file, even if we did not
specifically requested the whole file by using a very large range
yet.
IMPORTANT: This will fail to load past 599 bytes in OpenSim <= 0.7.0.2,
as it has buggy handling of ranged requests. But OpenSim http texture
breakage is not new. http://opensimulator.org/mantis/view.php?id=5081
|
| | | | |
|
| | | |
| | | |
| | | |
| | | | |
one prim is selected with Edit linked parts enabled
|
| | | |
| | | |
| | | |
| | | | |
and type 'BOOL' in operation)
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Qarl has given permission to use this code under the terms
of the GPL v2 + FLOSS exception and/or the LGPL v2.1.
|
| | |\ \ |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
CacheWillClear, CacheWillBeMoved, ChangeConnectionPort, ChangeSkin,
and ChangeFont.
There are many other alerts that should be made non-modal in the future,
but this should prevent a focus fight with the Vivox policy window.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
If this causes blurry fonts for a user, they should enable
"Force integer font sizes" in Preferences > Fonts.
|
| | | | | |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Gives more control over font sizes, addresses font being smaller
than some users desire.
|
| | | | | |
|
| | |\ \ \ |
|
| | |\ \ \ \ |
|
| | | | | | | |
|
| | |\ \ \ \ \ |
|
| | | | | | | |
| | | | | | | |
| | | | | | | |
| | | | | | | | |
last weekly.
|
| | | |/ / / /
| | | | | | |
| | | | | | |
| | | | | | | |
default, gcc-4.2 is used and 10.5 SDK is auto selected. Build from commandline in Release mode now work perfectly (./develop.py -t Release build) without the need for Xcode at all.
|
| | | |_|/ /
| | |/| | |
| | | | | |
| | | | | | |
window
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
OpenSim when a user's personal info isn't sent)
|
| | |/ / /
| | | | |
| | | | |
| | | | | |
startup
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
I searched the darwin and windows llqtwebkit tarballs and neither
qInitResources_WebCore nor qInitResources_WebCore_QT_NAMESPACE exist in
there... No idea why, but then we can't call it to initialize it now can
we?
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Also, darwin doesn't use quicktime, so disabled compiling that plugin
for darwin.
|
|\ \ \ \ \ |
|
| | | | | | |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
the version in the skins directory correctly
|