Thu, 29 Mar 2007 05:39:16 +0000 Don't put comments in BAIL_* macros where we want an empty argument...
Ryan C. Gordon <icculus@icculus.org> [Thu, 29 Mar 2007 05:39:16 +0000] rev 867
Don't put comments in BAIL_* macros where we want an empty argument... apparently it aggravates some compilers. Thanks to Chris Taylor for this one.
Wed, 28 Mar 2007 17:29:52 +0000 Logic bug in MVL/HOG/GRP archivers: these archives never contain subdirs...but they
Ryan C. Gordon <icculus@icculus.org> [Wed, 28 Mar 2007 17:29:52 +0000] rev 866
Logic bug in MVL/HOG/GRP archivers: these archives never contain subdirs...but they only enumerated their files when looking in a directory other than the root, instead of enumerating only for the root. Thanks to Chris Taylor for the catch.
Mon, 26 Mar 2007 07:53:43 +0000 Bumped version numbers.
Ryan C. Gordon <icculus@icculus.org> [Mon, 26 Mar 2007 07:53:43 +0000] rev 865
Bumped version numbers.
Mon, 26 Mar 2007 07:51:57 +0000 Man, I knew I would need getOSInfo() as soon as I deleted it!
Ryan C. Gordon <icculus@icculus.org> [Mon, 26 Mar 2007 07:51:57 +0000] rev 864
Man, I knew I would need getOSInfo() as soon as I deleted it! Apparently Win9x has the Unicode entry points, they just don't work, so only do a real lookup on NT-based platforms, and just force the fallbacks on 95/98/ME systems.
Mon, 26 Mar 2007 06:46:03 +0000 Ignore Visual C++ 6.0 build files.
Ryan C. Gordon <icculus@icculus.org> [Mon, 26 Mar 2007 06:46:03 +0000] rev 863
Ignore Visual C++ 6.0 build files.
Mon, 26 Mar 2007 06:38:36 +0000 Whoops, forgot to uncomment this from my testing...
Ryan C. Gordon <icculus@icculus.org> [Mon, 26 Mar 2007 06:38:36 +0000] rev 862
Whoops, forgot to uncomment this from my testing...
Mon, 26 Mar 2007 06:36:39 +0000 Tons of Unicode work in windows.c ... should now use UCS-2 on NT/XP/Vista/etc
Ryan C. Gordon <icculus@icculus.org> [Mon, 26 Mar 2007 06:36:39 +0000] rev 861
Tons of Unicode work in windows.c ... should now use UCS-2 on NT/XP/Vista/etc versions of the OS, and fallback to "ANSI" versions for 95/98/ME, tapdancing around the system codepage if it has to. Other minor Windows cleanups and corrections.
Mon, 26 Mar 2007 06:34:55 +0000 Whoops, this should probably be chopping the string instead of turning stable-1.0
Ryan C. Gordon <icculus@icculus.org> [Mon, 26 Mar 2007 06:34:55 +0000] rev 860
Whoops, this should probably be chopping the string instead of turning the first newline it sees into a space character...
Sun, 25 Mar 2007 06:26:53 +0000 Reworked getExePath() to remove some FIXMEs and respect Unicode.
Ryan C. Gordon <icculus@icculus.org> [Sun, 25 Mar 2007 06:26:53 +0000] rev 859
Reworked getExePath() to remove some FIXMEs and respect Unicode.
Sun, 25 Mar 2007 05:17:08 +0000 Improved dynamic loader and initial Unicode work in windows.c ...
Ryan C. Gordon <icculus@icculus.org> [Sun, 25 Mar 2007 05:17:08 +0000] rev 858
Improved dynamic loader and initial Unicode work in windows.c ...
Sun, 25 Mar 2007 04:59:11 +0000 Patched to compile on MSVC.
Ryan C. Gordon <icculus@icculus.org> [Sun, 25 Mar 2007 04:59:11 +0000] rev 857
Patched to compile on MSVC.
Sun, 25 Mar 2007 00:38:14 +0000 Pushed a bunch of FIXMEs into the Windows platform layer.
Ryan C. Gordon <icculus@icculus.org> [Sun, 25 Mar 2007 00:38:14 +0000] rev 856
Pushed a bunch of FIXMEs into the Windows platform layer.
Sat, 24 Mar 2007 05:42:29 +0000 Updated.
Ryan C. Gordon <icculus@icculus.org> [Sat, 24 Mar 2007 05:42:29 +0000] rev 855
Updated.
Sat, 24 Mar 2007 05:42:22 +0000 Replaced BeOS mutex implementation. Now all platforms have recursive mutexes.
Ryan C. Gordon <icculus@icculus.org> [Sat, 24 Mar 2007 05:42:22 +0000] rev 854
Replaced BeOS mutex implementation. Now all platforms have recursive mutexes.
Sat, 24 Mar 2007 05:13:54 +0000 Added some FIXME comments.
Ryan C. Gordon <icculus@icculus.org> [Sat, 24 Mar 2007 05:13:54 +0000] rev 853
Added some FIXME comments.
Sat, 24 Mar 2007 03:54:58 +0000 Replaced some Malloc and all the alloca() calls with __PHYSFS_smallAlloc(),
Ryan C. Gordon <icculus@icculus.org> [Sat, 24 Mar 2007 03:54:58 +0000] rev 852
Replaced some Malloc and all the alloca() calls with __PHYSFS_smallAlloc(), which will stack allocate small (128 or less bytes) blocks and Malloc the rest...naturally these now have to be paired with __PHYSFS_smallFree() calls, so you can't be as lazy as a basic alloca() would let you be. The benefit is both less malloc pressure for those temporary allocations and better stack overflow safety (so if some jerk tries to push a 78 megabyte string through the library as a filename, we won't try to strcpy it to the stack).
Wed, 21 Mar 2007 20:19:20 +0000 Apparently this is how you do 64-bit literals in MSVC...untested.
Ryan C. Gordon <icculus@icculus.org> [Wed, 21 Mar 2007 20:19:20 +0000] rev 851
Apparently this is how you do 64-bit literals in MSVC...untested.
Wed, 21 Mar 2007 20:14:46 +0000 Replaced LONGLONGLITERAL with __PHYSFS_UI64/__PHYSFS_SI64.
Ryan C. Gordon <icculus@icculus.org> [Wed, 21 Mar 2007 20:14:46 +0000] rev 850
Replaced LONGLONGLITERAL with __PHYSFS_UI64/__PHYSFS_SI64.
Wed, 21 Mar 2007 20:06:29 +0000 Updated.
Ryan C. Gordon <icculus@icculus.org> [Wed, 21 Mar 2007 20:06:29 +0000] rev 849
Updated.
Wed, 21 Mar 2007 05:22:48 +0000 Cleanups in calculateBaseDir.
Ryan C. Gordon <icculus@icculus.org> [Wed, 21 Mar 2007 05:22:48 +0000] rev 848
Cleanups in calculateBaseDir.
Wed, 21 Mar 2007 05:03:17 +0000 Split out Mac OS X code from unix.c and added some Carbon-specific code...
Ryan C. Gordon <icculus@icculus.org> [Wed, 21 Mar 2007 05:03:17 +0000] rev 847
Split out Mac OS X code from unix.c and added some Carbon-specific code... moved some stuff around and implemented a few things in BeOS and POSIX bits to accomodate this.
Wed, 21 Mar 2007 04:31:13 +0000 Replaced manual management of recursive pthread mutexes with
Ryan C. Gordon <icculus@icculus.org> [Wed, 21 Mar 2007 04:31:13 +0000] rev 846
Replaced manual management of recursive pthread mutexes with PTHREAD_MUTEX_RECURSIVE.
Tue, 20 Mar 2007 18:33:56 +0000 Since all the platform layers were using the same cut-and-paste of the
Ryan C. Gordon <icculus@icculus.org> [Tue, 20 Mar 2007 18:33:56 +0000] rev 845
Since all the platform layers were using the same cut-and-paste of the malloc/realloc/free allocator, I moved it into physfs.c as a default, which is used if the platform layer doesn't offer a platform-specific default allocator, which none do at this point, but will soon.
Tue, 20 Mar 2007 17:57:45 +0000 Removed __PHYSFS_platformTimeslice()...we don't actually use this anymore.
Ryan C. Gordon <icculus@icculus.org> [Tue, 20 Mar 2007 17:57:45 +0000] rev 844
Removed __PHYSFS_platformTimeslice()...we don't actually use this anymore.
Tue, 20 Mar 2007 17:53:42 +0000 Fixed a FIXME in the documentation comments.
Ryan C. Gordon <icculus@icculus.org> [Tue, 20 Mar 2007 17:53:42 +0000] rev 843
Fixed a FIXME in the documentation comments.
Tue, 20 Mar 2007 11:11:36 +0000 Removed skeleton.c from the repository...it's out of date.
Ryan C. Gordon <icculus@icculus.org> [Tue, 20 Mar 2007 11:11:36 +0000] rev 842
Removed skeleton.c from the repository...it's out of date.
Mon, 19 Mar 2007 20:15:16 +0000 Updated.
Ryan C. Gordon <icculus@icculus.org> [Mon, 19 Mar 2007 20:15:16 +0000] rev 841
Updated.
Mon, 19 Mar 2007 20:15:09 +0000 Try to use /proc/self/exe to determine the base dir. This lets systems with a
Ryan C. Gordon <icculus@icculus.org> [Mon, 19 Mar 2007 20:15:09 +0000] rev 840
Try to use /proc/self/exe to determine the base dir. This lets systems with a Linux-like /proc filesystem work with PHYSFS_init(NULL). Other cleanups in Unix basedir calculation.
Mon, 19 Mar 2007 20:13:37 +0000 PHYSFS_init() should fail if argv0 is NULL and we can't do without it.
Ryan C. Gordon <icculus@icculus.org> [Mon, 19 Mar 2007 20:13:37 +0000] rev 839
PHYSFS_init() should fail if argv0 is NULL and we can't do without it.
Mon, 19 Mar 2007 20:12:54 +0000 Added a FIXME.
Ryan C. Gordon <icculus@icculus.org> [Mon, 19 Mar 2007 20:12:54 +0000] rev 838
Added a FIXME.
(0) -300 -100 -50 -30 +30 +50 +100 +300 tip