summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2014-12-19Automated merge with ssh://bitbucket.org/lindenlab/viewer-tools-updateNat Goodspeed
2014-12-19Sort include(...) lines to make it easier for a person to verify.Nat Goodspeed
2014-12-19Force DEBUG_PREBUILT on in Prebuilt.cmake to debug failures on TC.Nat Goodspeed
2014-12-18Automated merge with ssh://bitbucket.org/lindenlab/viewer-tools-updateNat Goodspeed
2014-12-17Update autobuild to point to new version of jsoncpp and change lib names to ↵callum_linden
match those in package
2014-12-17Update autobuild to point to new (VS2013) Havok files and fix up name mismatchcallum_linden
2014-12-17Remove reference to Windows version of GTK/ATK/Pango in autobuild - not usedcallum_linden
2014-12-17Move the cmake package tracking files under the packages directoryOz Linden
2014-12-17Delete misleading comment about testing llcapabilitylistener.Nat Goodspeed
There was a comment in CMakeLists.txt left over from 2009 (!) about commenting out the unit test for llcapabilitylistener. Fortunately we've since reinstated the test as an integration test, but the comment was never removed -- until now.
2014-12-17merge up to latest changes from callum and natOz Linden
2014-12-17many package updates (mostly linux)Oz Linden
2014-12-17adjust cmake not to attempt loads of packages on the wrong platformsOz Linden
2014-12-17merge latest updates from nat and callumOz Linden
2014-12-17Produce error message, not traceback, when test program doesn't build.Nat Goodspeed
A traceback from a Python script always makes people think there's a bug in your script. Even when a test program fails to build, CMake often (always?) tries to run it anyway, via our run_build_test.py script. For that case, produce a straightforward error message -- rather than an OSError traceback that doesn't even mention the program name!
2014-12-16For one use case, boost::bind() better than boost::phoenix::bind().Nat Goodspeed
We use boost::phoenix::placeholders::arg1 to imply a whole lambda expression, replacing boost::lambda. But to bind a plain function in a more straightforward way, seems classic boost::bind() works while boost::phoenix::bind() does not.
2014-12-16Remove Visual Studio workaround for lack of std::fpclassify().Nat Goodspeed
Visual Studio 2013 evidently does provide std::fpclassify(), so we no longer need the funky local alias.
2014-12-16Fix another empty #elif to plain #else.Nat Goodspeed
2014-12-16Isolate #pragma clang to __clang__ compilations.Nat Goodspeed
It's very irritating that Visual Studio produces a warning for unrecognized #pragmas, which we then merrily turn into an error -- #pragma is inherently compiler-specific!
2014-12-15fix for some std::pair/STL syntaxcallum_linden
2014-12-15Merge with headcallum_linden
2014-12-15Add feature (enabled) to turn off media pluginscallum_linden
2014-12-15include quicktime package only on windowsOz Linden
2014-12-15include libuuid and fontconfig only on linuxOz Linden
2014-12-15don't include glext on Mac where it is not usedOz Linden
2014-12-15remove redundant include of Audio.cmakeOz Linden
2014-12-15merge changes for 3.7.22-releaseOz Linden
2014-12-12Remove unneeded media plugin header filecallum_linden
2014-12-12Implicit conversion from ostream to string needs help in VS2013callum_linden
2014-12-12merge latest fix from natOz Linden
2014-12-12Make JsonCpp.cmake reference Windows lib names we actually use.Nat Goodspeed
2014-12-12merge dictionary changeOz Linden
2014-12-12fix dos line endingsOz Linden
2014-12-12Eliminate use of boost::lambda with boost::function (Trac #10864).Nat Goodspeed
https://svn.boost.org/trac/boost/ticket/10864 I've used boost::lambda with boost::function in a number of creative ways over the years. But the clang 6 shipped with Xcode 6 seems to have somehow broken lambda + function in Boost 1.57. boost::phoenix is a partial workaround. Sadly, lambda's comma-operator overload doesn't seem to be supported, necessitating a couple ugly workarounds. With real lambdas now supported by current compilers, I'm sure the Boost community has little incentive to repair the lambda + function problem. Presumably we'll be able to use such features ourselves Real Soon Now...
2014-12-12Make LINE_NUMBER_HERE workaround conditional on LL_WINDOWS.Nat Goodspeed
The referencing code is all within LL_WINDOWS conditionals, so the dummy variable must be enclosed that way too -- otherwise we get unreferenced-variable warnings-as-errors on other platforms.
2014-12-12Automated merge with ssh://bitbucket.org/lindenlab/viewer-tools-updateNat Goodspeed
2014-12-12Wrap call to autobuild in some diagnostic code.Nat Goodspeed
Hopefully this will make certain errors clearer, notably failure to launch autobuild due to "cannot find the file specified".
2014-12-12Fix up remaining llround() -> ll_round() issuescallum_linden
2014-12-12Merge with headcallum_linden
2014-12-12Update VSTool that is used to set configuration in MSVC solution file for ↵callum_linden
MSVC 2013 (v12.x)
2014-12-12Another round of ll_round()Nat Goodspeed
2014-12-12Automated merge with http://bitbucket.org/lindenlab/viewer-tools-updateNat Goodspeed
2014-12-12Update viewer to consume googlemock build 297460.Nat Goodspeed
2014-12-11Automated merge with ssh://bitbucket.org/lindenlab/viewer-tools-updateNat Goodspeed
2014-12-11Update viewer to consume colladadom build 297450Nat Goodspeed
2014-12-11Re-add missing variable declaration (not that it's very useful and no idea ↵callum_linden
how it was removed)
2014-12-11Add bespoke ctor/dtor that align on 16byte boundaries when creating things ↵callum_linden
on the heap
2014-12-11Update viewer to consume boost build 297445 (no fchmodat())Nat Goodspeed
2014-12-11Fix a couple more ll_round() callsNat Goodspeed
2014-12-11Disambiguate a few ensure() calls.Nat Goodspeed
These were of the form ensure(std::string, something convertible to bool). Not sure what the ambiguity was, but ensure(std::string, bool(something)) works better.
2014-12-11Add /SAFESEH:NO /NODEFAULTLIB:LIBCMT to viewer linker switches.Nat Goodspeed
zlib cannot be linked with SAFESEH, which is the VS 2013 default, so every consumer must set /SAFESEH:NO. /NODEFAULTLIB:LIBCMT seems to eliminate many duplicate-symbol link errors.