SH-1990 Fix for redundantly applying some color scaling to sky and water (also add some snazzy specular bloom from the sun)

SH-2031 Disable usage of glMapBuffer (again).  Despite using MapBufferRange, this is still a source of frame stalls.

SH-2031 Followup to curl threading work -- don't start and stop the thread on every request, use a signal (cuts time spent in Pump IO down from 1-2 ms to 0.1ms)

Reviewed by Kelly

Reviewed by Leslie

EXP-997 FIX -- Significant FPS degradation in 2.8.0 Beta on Mac equipped ATI video card

* Disabled VBO's on mac to avoid stalls in renderer
* Placed valid data into the padding between triangles in the vertex buffer to remove uninitialized craziness
* Removed invalid rendering checks causing GL errors in debug mode

Reviewed by davep

Added a function for parsing a date string of specific format.
Added strings defining the date format in Group Profile Land/Assets that should be localized.

Instead of low-level open(O_CREAT | O_EXCL) loop on all platforms, use
GetTempFileName() on Windows and mkstemp() elsewhere.
Don't append a final newline to NamedTempFile: use caller's data literally.
Tweak a couple comments.

SH-2031 Fix for shadow render targets using inappropriate shader for occlusion culling resulting in objects popping in and out of the shadow map.

SH-2031 High risk changeset, but potentially high reward.  Addresses frame stalls in renderer by never using the fixed function pipeline if shaders are available.

Consider this pathname for llsdserialize_test.cpp:
C:\nats\indra\llcommon\tests\llsdserialize_test.cpp
Embed that in a Python string literal:
'C:\nats\indra\llcommon\tests\llsdserialize_test.cpp'
and you get a string containing:
C:
ats\indra\llcommon	ests\llsdserialize_test.cpp
where the \n became a newline and the \t became a tab character.
Hopefully Python raw-string syntax r'C:\etc\etc' works better.

In this case, the Python code in question is being written from a C++ string
literal to a temp script file in a platform-dependent temp directory -- so the
Python __file__ value tells you nothing about the location of the repository
checkout. Embedding __FILE__ from the containing C++ source file works better.

After review and testing, realized that 'WatchdogEnabled' is actually true to its name, and is a boolean, not a timeout value. 'MainloopTimeoutDefault' is the actual timeout value.

Updated descriptions and values accordingly to set Watchdog timeout to 60 seconds.

And at that point, the Python logic needed to bring in the llsd module is big
enough to warrant capturing it in a separate string variable common to
multiple tests.