Fixes a "buffer overflow detected" abort when compiled with _FORTIFY_SOURCE=1.
The realpath(3) function in glibc checks if the destination buffer is large
enough to hold up to PATH_MAX characters and aborts if that is not the case.
PATH_MAX doesn't have to be defined so assume that it's equal to 4096.
We should really be using pathconf(_PC_PATH_MAX) instead of a hard-coded value
but that means we can no longer use static buffers to hold paths.
and dirty workaround to problems in libs/picomodel/pm_obj.c which
passes files that aren't Wavefront ASCII.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@491 8a3a26a2-13c4-0310-b231-cf6edde360e5
ever since certain updates in that distribution. This fixes the q3map2 crash,
but no more optimizations.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@486 8a3a26a2-13c4-0310-b231-cf6edde360e5
copied from "-scale" and was incorrect. Thanks to Ensiform from ioquake3 IRC
room for reporting this bug.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@482 8a3a26a2-13c4-0310-b231-cf6edde360e5
- Added plugins/imagehl to list of modules built (both Windows and Linux).
Not really sure what this does, but it's the only module (as defined by
stuff in plugins/ or contrib/) that exists and isn't currently part of the
build.
- The code to disable desktop composition in Windows has been moved from start
of main() to an option in Preferences. The pref reads as follows:
"NVIDIA/Aero bug - disable Windows composition". It is enabled by default.
- Color themes controlled by "Misc" -> "Colors" -> "Themes" options are now
saved as soon as you select them. The color theme is no longer clobbered
when opening Preferences.
- The ATI broken driver (bug #802) pref has been reworded since it applies to
Intel cards as well. Before:
"ATI cards with broken drivers - bug #802". After:
"ATI and Intel cards w/ buggy drivers (disappearing polygons)".
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@479 8a3a26a2-13c4-0310-b231-cf6edde360e5
Problem was that the compile flags were overly aggressive and caused segfaults.
They included: -O3 -Winline -ffast-math -fno-unsafe-math-optimizations
-fno-strict-aliasing
Changed to: -O2 -fno-strict-aliasing
From my experience, -O3 is usually a bad idea.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@476 8a3a26a2-13c4-0310-b231-cf6edde360e5
Fixing plugins on both Linux and Windows. It's consistent now.
Fixing gl font on Windows.
Mostly, added lots of plugins.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@474 8a3a26a2-13c4-0310-b231-cf6edde360e5
A couple of things are broken right now:
- OpenGL font in Windows not appearing (expected to be broken, will fix)
- Linux build broken because it loads some .vcproj files that are removed
- 3rd party libs are downloaded from porky.nerius.com, not id Software
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@473 8a3a26a2-13c4-0310-b231-cf6edde360e5
I didn't realize they were so broken.
The fix is safe and should not break plugins or anything else.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@437 8a3a26a2-13c4-0310-b231-cf6edde360e5
"Default texture scale:". (Preferences are game specific by the way.)
The default value for this preference is specific to a game, but most
of the time it's 0.5. This field allows the user to override that default.
This is a feature that was specifically requrested by someone who wants to
try using ZeroRadiant (he wants to set it to 0.25 for his "premium" maps).
It's already a feature in NetRadiant.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@436 8a3a26a2-13c4-0310-b231-cf6edde360e5
to reduce the size of Windows Radiant builds by a factor of 2. JAPack
can still be used, but must either be installed manually after a build
is made, or it must be explicitly downloaded before the build.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@435 8a3a26a2-13c4-0310-b231-cf6edde360e5
This is more convenient than using custinfoparms.txt. TTI (from the
Reaction team) game these to me.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@434 8a3a26a2-13c4-0310-b231-cf6edde360e5
Otherwise, "hall of mirrors" effect takes place in the OpenGL viewports.
This fix is a real hack. It's actually borrowed from other Radiants such
as NetRadiant. I'd like to fix the underlying issue someday so that Aero
can be used with Radiant.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@433 8a3a26a2-13c4-0310-b231-cf6edde360e5
only adding path if it's not already added (remove duplicates).
It's cleaner this way.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@432 8a3a26a2-13c4-0310-b231-cf6edde360e5
we use ~/.Reaction instead of ~/.q3a . Oh well. This is for Reaction Quake III.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@431 8a3a26a2-13c4-0310-b231-cf6edde360e5
Introducing Q3MAP2_EXPERIMENTAL_MODEL_CLIPPING_FIX to control the new code,
current value is 1 (enable the fix).
This code fixes the q3map2 regression test model_clipping_45_degrees.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@429 8a3a26a2-13c4-0310-b231-cf6edde360e5
version of code in VectorNormalize() is used. Yes, I put the old code back
in there, and it's active if MATHLIB_VECTOR_NORMALIZE_PRECISION_FIX is 0.
Right now it's 1, so the fixed code is active. I need this quick way to
test regression tests.
git-svn-id: svn://svn.icculus.org/gtkradiant/GtkRadiant/trunk@424 8a3a26a2-13c4-0310-b231-cf6edde360e5
