| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
we never did that here in meta7.
This causes a problem, since LL based viewers react differently to those first 5 estates. In particular
there are dire warnings and greyed out widgets in the estate access tools. In meta7 there are now 2 active
estates with numbers in the 1 to 5 range that are just ordinary estates, but their owners can't unban residents.
The server developers never gave me an answer about wether just renumbering the estates would cause any problems.
So I'll fix this in the meta7 viewer.
These sorts of things should be server side enforced policy anyway. This may break LL TPV, but meta7 does not care
about some other grids policies.
|
|
|
|
|
|
|
|
| |
location.
This is an unimplimented feature for Imprudence, as they update the max Z position for prims.
Since we don't actually provide these sorts of things from the sims, I'm just gonna hard code it for now.
|
| |
|
|
|
|
|
|
|
|
| |
grid to meta7.
Prevent setting the default grid to anything but meta7 (except by editing the settings file).
By Tom Meta from his fork. See the TomsOldWork branch.
|
|
|
|
|
|
| |
toggled by CTRL+SHIFT+G.
By Tom Meta from his fork. See the TomsOldWork branch.
|
|
|
|
|
|
|
|
|
| |
out that important system textures such as the
water normal map are delivered in the SKINS folder and if they were in the cache directory they'd be deleted when the cache
was reset. Also fix a bug with resetting the cache location.
By Tom Meta from his fork. See the TomsOldWork branch.
|
|
|
|
| |
large number. Only affects linux builders. I'll fix it properly later, after I learn python.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
Fixes missing definition for LLWindowMacOSX::getNativeKeyData().
|
| |
|
| |
|
|
|
|
| |
GCC 4.0 on Mac OS X 10.5 can't handle SSE4, GCC 4.2 fails.
|
|
|
|
|
|
|
|
| |
followingrepercussions:
* CTRL-A while a prim is selected switches to the align tool
* CTRL-SHIFT-A while a prim is selected switches to the pack tool
* Area Object Search shortcut has been changed from CTRL-SHIFT-A to ALT-A
|
|
|
|
| |
object, by Beeks
|
| |
|
| |
|
|
|
|
| |
register/password urls and science sim name
|
| |
|
| |
|
|
|
|
| |
Decosta
|
|
|
|
|
|
| |
hopefully lower the crash rate a little" as it's a poor fix, and many photographers find the higher sizes useful
This reverts commit 2819c9ae3652e913ea2d2da1c1ed8ef0d2175e00.
|
| |
|
| |
|
|
|
|
| |
(fly) up animation
|
|
|
|
| |
a menu item check rather than a call
|
| |
|
|
|
|
| |
ruler after applying Qarl's align patch
|
| |
|
| |
|
|\ |
|
| |\ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
To use ld.gold configure with:
-DCMAKE_EXE_LINKER_FLAGS:STRING="-Wl,-use-gold".
ld.gold links the viewer on my machine in 8 seconds, as
opposed to 19 seconds with ld.bfd. Moreover, it uses a
LOT less memory during linking (about 750 MB instead of
2.5 GB!). Thanks to Siana Gearz for pointing out that
ld.gold is so much faster.
|
| |\ \ |
|
|\ \ \ \ |
|
| |/ / / |
|
| |\ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Conflicts:
linden/indra/newview/viewer_manifest.py
Trivial proximity collision. Resolved.
|
| |\ \ \ \ |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
stage it in viewer_manifest.py
|
| | | | | | |
|
| |/ / / / |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Changes:
* Added changes from snowglobe 1.5 to indra/llcommon/llstring.h
(compile errors as a result of a missing include, but I copied
everything else too).
* Added #include "linden_common.h" to emeraldboobutils.cpp.
Really it's one the header files that needed that, but that's
how this header works: every source file should include it as
first header anyway, then there is no need for other headers
to do that.
* Renamed LLPanelRegionOpenSettingsInfo::sendUpdate(void*) to
LLPanelRegionOpenSettingsInfo::onClickOrs because it was
hiding a virtual function (BOOL sendUpdate(void)).
* Made cutToClipboard more equal to copyToClipboard (was also
hiding a virtual function).
* Install libllcommon.so in lib64 on Linux_x86_64, instead of lib.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
See http://jira.secondlife.com/browse/SNOW-713
This patch makes llcommon shared. It contains the following snowglobe
(SVN) changesets: 3478, 3479, 3480, 3481, 3482, 3485, 3496, 3498, 3500,
3519 and 3531, plus the fix for all rejects. Note that changes
to scripts/automated_build_scripts/opensrc-build.sh (changesets 3500
and 3625) were ignored as we don't have scripts/automated_build_scripts.
|
|\ \ \ \ \ |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Use gFrameIntervalSeconds instead of gFrameTimeSeconds - which is what
the original author tried to do. Also, clean up some dead code.
*However*, there is still the issue of texture fetch workers staying
alive after finishing partial fetches, which falsly manifests here as
pending work.
|
|\ \ \ \ \ \
| |/ / / / /
| | | | | /
| |_|_|_|/
|/| | | | |
|