Activity for Ozkan Sezer
-
Ozkan Sezer committed [12ff5e] on libxmp
update version number and copyright year in README files
-
update copyright years of changed sources
-
mdataio.h, mix_paula.c, win32.c: add missing libxmp license header.
-
lzw.c, path.c: remove unused assert.h includes.
-
updated documents for 0.97.0 changes so far.
-
image.c: Accept type 3 (uncompressed grayscale) TGA files.
-
Bump MAX_MODELS/MAX_MOD_KNOWN to 4096
-
add note about the XMP_MAX_FRAMESIZE constant change to changelog.
-
bump version to 4.7.0
-
Fix xmp_start_player unmuting regression from 4.5.0.
-
Host_Map_f: replace Q_strrchr and Q_strcmp with libc calls
-
ice_unpack: make variadic macro debug() work with old gcc versions
-
Removed con_warnings of not-loop LIST WAV chunks
-
NEWS: update for changes after 3.3.13, so far
-
cmake: add vita driver after commits 9f26ffd and 82d914c
-
examples: provide MikMod_Sleep() for dos when using watcom compiler
-
player: a few OpenWatcom/DOS bits
-
Try to load texture from bsp if external wad load fails
-
ci: add bsd and haiku workflows
-
read_event_med(): fix shadowed variable warning after PR/#916
-
Anyhow, I imported the most recent patch. Is current revision 5532 fine for you? Looks good to me
-
Discard my big patch above. The following revised version of the Marteen's patch should do the trick: diff --git a/ports/cmake/src/CMakeLists.txt b/ports/cmake/src/CMakeLists.txt index 7861e35..f6a77b9 100644 --- a/ports/cmake/src/CMakeLists.txt +++ b/ports/cmake/src/CMakeLists.txt @@ -74,9 +74,9 @@ check_function_exists(execvp HAVE_EXECVP) check_function_exists(ctermid HAVE_CTERMID) check_function_exists(clock_gettime HAVE_CLOCK_GETTIME) -check_type_size(off_t SIZEOF_OFF_T) +check_type_size(off_t...
-
test-dev/test_path.c: fix MSVC C4130 warnings
-
avoid -Wmissing-prototypes for mkstemp():
-
fix some -Wzero-as-null-pointer-constant warnings.
-
ci: bump cross-platform-actions.
-
test_api_get_format_list.c: fix mistake from PR/898
-
fix some -Wzero-as-null-pointer-constant warnings.
-
configure.ac: fail if ar or ranlib isn't found for static builds
-
CMake (lite): fix installation of .cmake scripts on MinGW (c.f.: #888)
-
Fix link failures in builds C++ with compilers.
-
common.h: fix a very stupid typo introduced in PR/#821.
-
cmake: set AC_APPLE_UNIVERSAL_BUILD when necessary
-
cmake: use POSITION_INDEPENDENT_CODE property instead of manually adding -fPIC to CFLAGS
-
cmake: go back to 3.2 as minimum required cmake version
-
fix some -Wzero-as-null-pointer-constant warnings.
-
h2_stdinc.h: revised bool stuff.
-
h2_stdinc.h: revised bool stuff.
-
lodepng.c: fix -Wzero-as-null-pointer-constant warnings
-
miniz.c: fixed -Wzero-as-null-pointer-constant warnings
-
q_stdinc.h: revised bool stuff
-
q_stdinc.h: revised bool stuff
-
fix some -Wzero-as-null-pointer-constant warnings.
-
minor clean-up for win32 directx-using sources.
-
fix some -Wzero-as-null-pointer-constant warnings.
-
fix some -Wzero-as-null-pointer-constant warnings.
-
fix some -Wzero-as-null-pointer-constant warnings.
-
dos, sndpci (libau): bump optimization level to -O2 and rebuild.
-
dos, sndpci (libau): Intel HDA PCI ID updates from Mpxplay-1.68.
-
cmake, android: support Google Play 16 kiB page size requirement
-
fixed a -Wunused-parameter warning.
-
ci: bump bsd versions
-
cmake: set maximum policy to 3.10
-
update license files to latest FSF versions.
-
oslibs: rebuilt libvorbisidec for djgpp again ,
-
rebuilt libvorbisidec from lowmem branch with comments loading disabled
-
add OpenWatcom makefiles targeting Linux
-
Revert "SV_Physics(): fix entity cap check in the for loop" for now
-
minor doc update
-
hexen2, R_RunQuakeEffect(): fix q_sincosrad typo from commit c3609eab1
-
hexen2, R_RunQuakeEffect(): fix q_sincosrad typo from commit c3609eab1
-
updated 1.5.10 release date.
-
re-diff'ed some of the optional patches so that they apply cleanly
-
remove m68k-amigaos stuff from documentation.
-
Revert "r_misc.c: don't use Cvar_Set in R_SetupFrame in multiplayer"
-
revert Cmd_MoveToFront and Cvar_MoveToFront micro-optimizations.
-
Revert "amiga: fixed the SV_SpawnServer race condition on meso2 and romeric5"
-
remove rest of m68k-amigaos3 support code from sources.
-
remove m68k-amigaos support from opengl code.
-
remove m68k-amigaos assembly stuff.
-
remove m68k-amigaos target-specific source and libraries
-
remove m68k-amigaos target from makefiles
-
@bszili: I'll be dropping support for m68k-amigaos for 1.5.10 today: will send you an email for it. P.S.: Created a m68k-amigaos branch in git if someone wants to mess with it later..
-
Great that you found the culprit. It's troublesome that one file system from a former version being incompatible ...
-
bump version to 0.97.0 and update documents for changes
-
Well, then there won't be m68k-amigaos binaries in 1.5.10 release
-
@bszili: Can you produce working binaries so that I can replace those in the 1.5.10 lha above?
-
It should be set to no OK, I just disabled Bebbo toolchain in all makefiles. (See latest git.)
-
Makefiles: disable Bebbo's GCC6 toolchain by default for m68k-amigaos3
-
Strange: When copying the whole game dir to RAM and starting it from here, saving works. Copying the savegame back to my internal drive loads the game, but I can't save again. I have no issues with other games... Quake 1, 2, AmiDuke and so on with saving. As I understand, directory creation is failing for who knows why.. @bszili: Can you have peek at that?
-
It should be set to no, it's no longer publicly available for the time being. OK, will do The 1.5.10 version you linked doesn't work for me. It produces a lot of illegal memory accesses and it has broken in-game grahpics. Hmph. I have only one toolchain, that means I can't build a working binary for m68k-amigaos3. Can you create a working binary yourself? P.S.: IIRC, the build defaults to clib2. I wnder if anything changes if built for libnix...
-
I never had this issue, OK then, to be sure: Does the new build I linked above behave properly? If it's good, I'd like to use that in 1.5.10 release (I hope I do the release before a million years more..) but compiling the game with the new toolchain is probably not a good idea. There are some optimizer issues causing random crashes I never managed to track down. Ah, I remember that you liked the new bebbo toolchain.. Should we change the BEBBO_TOOLCHAINoption to noin the Makefiles, then?
-
clean-ups after the escripten/wasm port patch
-
I assume that you are using the 1.5.9 release. If that is the case, does this new build help at all? https://uhexen2.sourceforge.net/tmp/hexen2-1.5.10-amigaos-m68k.lha I built that from current (today's) sources in git. However my toolchain is old and it might have bugs. If my new build still has the issue, then @bszili uses a newer toolchain: I'd like to see results from his builds..
-
I assume that you are using the 1.5.9 release. If that is the case, does this new build help at all? https://uhexen2.sourceforge.net/tmp/hexen2-1.5.10-amigaos-m68k.lha II built that from current (today's) sources in git. However my toolchain is old and it might have bugs. If my new build still has the issue, then @BSzili uses a newer toolchain: I'd like to see results from his builds..
-
@bszili ?
-
small cmake fixes.
-
snd_mpg123.c: support for MPG123_API_VERSION 49 (mpg123 >= 1.33.0)
-
snd_mpg123.c: support for MPG123_API_VERSION 49 (mpg123 >= 1.33.0)
-
tempfile.c: more precise preprocessor checks for io.h inclusion.
-
updated windows SDL2 to latest official 2.32.8 version.
-
SV_Physics(): fix entity cap check in the for loop.
-
host_cmd.c (ExtraMaps_List): fix build in C89 mode.
-
playmidi.c (find_samples): timidity-0.2i behavior for gus patches:
-
playmidi.c (find_samples): timidity-0.2i behavior for gus patches:
-
playmidi.c (find_samples): add special case for 1-sample instruments ,
-
don't fail if a gus percussion instrument has more than one samples
-
sndfont.c: replaced use of strcpy() with memcpy().
-
readsbk.c: replaced use of strncpy() with timi_strxcpy()
-
initial minimal error checking to load_sbk().
1 >
