Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
floater_tools build menu
|
|
"enumeration value 'MEDIA_EVENT_CLICK_LINK_NOFOLLOW' not handled in switch"
|
|
|
|
|
|
Added code to support SLAPPS for Media on a Prim
Reviewed by Monroe and Callum via Code Collab (Review 143).
|
|
into build floater
Simple fix to use dynamic cast - the previous unsafe static cast was treating folders as if they were items and then querying permissions, which don't exist for folders.
This code actually existed back in 1.23 but by some stroke of luck worked.
|
|
type 'ne' or 'na' or some other letters combinations
reviewed by Leyla
|
|
|
|
|
|
|
|
reviewed by Palmer
|
|
|
|
|
|
|
|
|
|
Certain tests of allowModifyBy(), allowCopyBy() and allowMoveBy() were
computing a bit-shifted S32 value and testing that the method return value
matched that specific bit. Whoever originally coded the test probably started
with allowModifyBy() == TRUE, but observed that it didn't work. Instead of
just testing for nonzero, in the spirit of BOOL, he then determined the
specific BOOL value (typedef for 'int') returned by each test so he could
still compare for equality. In other words, the tests were far too
knowledgeable about the method implementation. At some point these methods
were changed to return C++ standard 'bool' instead, so the internal
bit-masking operations got mapped to 'true' and 'false' anyway, making it
impossible for the original equality comparison to succeed. Fix tests
accordingly.
Use ensure_equals("msg", expr1, expr2) instead of ensure("msg", expr1 == expr2).
On failure, the former will report the actual mismatched values.
Remove other extraneous Microsoft-style "BOOL" usage (e.g. expr == TRUE).
|
|
|
|
Make the fix work for alpha masks ('fast alpha') too - ironically this hasn't ever worked either AFAICT.
|
|
actually "solid"
and its many many dupes.
In summary this does the following:
* When laying down alpha, uses the new separated color-and-alpha blend func support to suppress existing glow in proportion to the opacity of the incoming fragments
* Moves glowing-alpha support from the glow pool to the alpha pool, ensures that the glow pass for a vertex buffer happens after the alpha pass for the vertex buffer so that it doesn't suppress its *own* glow.
Tested with all shader modes, FBO on/off, and deferred rendering. Tested with every permutation of alpha/non-alpha/glow/non-glow/fullbright/waterglow/particle interaction I could contrive so far. :)
|
|
|
|
|
|
view " << mName << " during UI draw() phase" caused in lltextureview.cpp (from the texture console).
|
|
|
|
This fix should be able to improve the speed of loading texturs from cache noticeably (30% or more). More fix to come.
|
|
I want it to fix glow-through-alpha, but this checkin doesn't includes any of those changes.
|
|
|
|
water has depth. Using 4th component of the normal map to encode a position displacement (along the normal) in the class2 shadow lookup.
|
|
|
|
|
|
thing we did to this vector was normalize it anyway.
|
|
from class1 except for some minor const tweaks (I've migrated class3's tweaks into class1 on the assumption that this was what davep sees all the time)
|
|
|
|
|
|
|
|
|
|
recently.
list.erase(list.find()) is not safe, because erasing list.end() is undefined. Check to make sure the find succeeded before doing the erase.
Reviewed by Richard at http://codereview.lindenlab.com/341001
|
|
|
|
|
|
|