OpenArena Message Boards

OpenArena Contributions => Development => Topic started by: sago007 on April 05, 2008, 02:39:03 PM



Title: New binaries (ioquake SVN 1288) (updated 04-19), please test
Post by: sago007 on April 05, 2008, 02:39:03 PM
Ok, here are the new binaries based on ioquake3 SVN 1288 + bloom + oachanges + wav2ogg patch (like the 0.7.0 binaries) + security fixes + vm + sago's input fix

Binaries: http://ostenfeld.dk/~poul19/public_files/oabin-test-2008-04-19.zip
And the modified source (to comply with the GPL): http://ostenfeld.dk/~poul19/public_files/ioquake3svn1288plus3.tar.bz2

To test:
Simply copy the new binaries to your openarena-0.7.0 folder (they are fulle compatible with the 0.7.0 version although some of the new features will be missing)

Report back here if you think the new binaries are worse than the old ones.

About the new binaries
I would especially like to hear if any Windows users have problems with the new binaries as the Windows version is now based on SDL.
openAL is now loaded dynamically under Linux and will be disabled if not available (instead of crashing like the current version)

Stable (svn 1127)
Binaries: http://ostenfeld.dk/~poul19/public_files/ioquake1.34_svn1127_oa_test.zip
Source: http://ostenfeld.dk/~poul19/public_files/ioquake3stable.tar.bz2

VM files
http://ostenfeld.dk/~poul19/public_files/pak8_testvm3.pk3

About VM files
These removes the last references to UI references to CD key and punkbuster. They add new gametypes, optional instantgib and some other changes.
To install them simple place the pak8_testX.pk3 in the baseoa directory.

Edit 2008-04-19 3:37
Added new experimental binaries and source.


Title: Re: New binaries (ioquake SVN 1288), please test
Post by: pikaunforgiven on April 05, 2008, 08:13:17 PM
ive tried the new binaries and i have to say im quite impressed so far. the levels in the game look a LOT better imo, unless my eyes are decieving me. the only problem i seem to have is the game feels like its stuttering sometimes, despite a steady 125fps. ill be fighting a few bots offline, and occasionally when i move the mouse it feels as if its "jerky" or something like a low/fluctuating framerate (i have a saitek gaming mouse, optical fyi).  its not enough to hurt my gameplay too much, as im just an average player anyway. ive tested with max settings @ 640x480 (my normal setup) and the lowest settings @ 640x480 with the same results. this doesnt happen with the release version.  i didnt try online yet.

im running in winxp pro sp2, probably going to test in kubuntu within a day or so when i get to booting into it. here are my relevant system specs in case it matters:

3.0Ghz intel pentium 4 (single core prescott)
1GB DDR400 ram
nvidia geforce 6200OC video card

other than the slightly downgraded performance, i'd have to say im happy enough to use them from now on. if you need to know more info (like variables and such) feel free to ask. if you know of any settings i should test/try feel free.

edit: after playing a few more offline rounds i managed to get it to crash. a few times while playing the client would seem to freeze every now and then for roughly a second then resume play as if nothing happened. its possible this could be due to the custom xfire configuration i made so xfire ingame works for open arena but im not sure. i dont have any debugging tools installed or anything so unfortunately thats the only data i have from the crash. im going to disable that now and report back if it happens again.


Title: Re: New binaries (ioquake SVN 1288), please test
Post by: aantipop on April 06, 2008, 02:20:52 AM
i replaced the openarena file in /usr/bin with the openarena.i386. if i run it i get this error:

[aantipop@blackbox bin]$ ./openarena.i386
ioq3+oa 1.35 linux-i386 Apr  5 2008
----- FS_Startup -----
Current search path:
/home/aantipop/.openarena/baseoa/zl3tourney1.pk3 (112 files)
SNIP: list of all pk3 files
/home/aantipop/.openarena/baseoa/0ups_beta2a.pk3 (65 files)
/home/aantipop/.openarena/baseoa
./baseoa

----------------------
19052 files in pk3 files
Couldn't load default.cfg



EDIT
i i copied the pak0 to the baseoa in my home directory and got it workin, will test it now

EDIT2:
i tested it in windows xp and linux, both work fine for me.
however i get a red background and a strange terminal background, with a missing shader or texture:



Title: Re: New binaries (ioquake SVN 1288), please test
Post by: sago007 on April 06, 2008, 08:29:16 AM
however i get a red background and a strange terminal background, with a missing shader or texture:
Looks strange. Did you only copy the pak0?

About the configurations that I'm interested in is mainly com_hunkmegs and r_bloom. I know that some maps will get a "memory allocation error" if bloom is enabled with the default hunkmegs. I'll try playing a little in Windows to see if I can spot some hangs.

pikaunforgiven's information about hangs despite a steady framerate does sound like the same problem I have whit Nexus. Unfortunately that can be SDL related...


Title: Re: New binaries (ioquake SVN 1288), please test
Post by: Cacatoes on April 06, 2008, 09:16:25 AM
Using Debian here, on a notebook centrino 1.6ghz with geforce 5200.

I just copied the binaries in OA-0.7, my config files were reset but it happens sometimes and I don't mind.
I experience some freezes of a few seconds while playing, I only played online, It happens once every 5 minutes or so.
I think this is annoying, but I'm already annoyed by the temporary poor connection on which I play these days.
Maybe I didn't pay much attention, but I didn't see some improvements compared to the usual ioquake3 which comes with OA 0.7.


Title: Re: New binaries (ioquake SVN 1288), please test
Post by: pikaunforgiven on April 06, 2008, 09:20:25 AM
i usually use a com_hunkmegs of 128, but i can try with the default settings. i noticed r_bloom is set to 0 by default, are the only options 0 and 1 or are there others?

i tested with the linux client just now and i dont experience any of the problems i had with the windows one, so it could indeed be SDL related. then again, for me the linux client always did seem to perform a bit better despite being on the same hardware as my windows xp install.

ive also disabled my xfire ingame hack, and so far havent crashed in windows both offline and online. most likely the crash was partially due to xfire ingame not expecting SDL since i was more or less using the same things it specifies for retail quake 3.

Cacatoes, what video settings are you using?  i only really notice if i have all of the texture settings turned all the way up with trilinear filtering. the freezes you experience sound similar to my freezes in windows, although admittingly i didnt really run the linux client for an extended period to see if they would develop.

edit: just tried setting r_bloom to 1 in windows, and it causes the client to be very slow even on the main menu.


Title: Re: New binaries (ioquake SVN 1288), please test
Post by: iLeft.bye on April 06, 2008, 10:14:07 AM
I deny to use anything using sdl on windows


Title: Re: New binaries (ioquake SVN 1288), please test
Post by: Cacatoes on April 06, 2008, 12:07:09 PM
That wasn't extended tests, and not with my usual config as it has been scratched.
The freezes can be shorter than what I said.
I don't use specific settings, default ones.
I lowered much graphical details to speed things up, but had a bad framerate (did not reach a constant 125 as I usually do). I'll switch back to the old ioquake3 to see if I experience the same.


Title: Re: New binaries (ioquake SVN 1288), please test
Post by: aantipop on April 06, 2008, 03:23:11 PM
Looks strange. Did you only copy the pak0?

nope i copied all pak's, then the "default.cfg" error was gone and i could run the game.
my config got reset too,  but i have a backup :) i will test with a clean openarena windows install soon, i just copied all my linux files over.



Title: Re: New binaries (ioquake SVN 1288), please test
Post by: sago007 on April 06, 2008, 03:58:45 PM
Stable files

Well to compare, here are the stable ioquake version including OA changes. It is missing the resolution detection, but should be at least as good as the current oa binary. If not it must be my who are bad at compiling (how much can go wrong?)

Binaries: http://ostenfeld.dk/~poul19/public_files/ioquake1.34_svn1127_oa_test.zip
Source: http://ostenfeld.dk/~poul19/public_files/ioquake3stable.tar.bz2


Game logic will follow soon...


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: sago007 on April 06, 2008, 04:11:15 PM
I think I discovered why there are no background in the console...

fix pending

EDIT:
or so I thought...


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: pikaunforgiven on April 06, 2008, 07:03:17 PM
with the stable one r_bloom still makes the windows client very slow (around 10fps compared to 125 with it off) and with the linux client the framerate is the same as it always has been (a steady 125fps) with no noticable graphics change.
both clients are using the same exact config.

mind enlightening us as to why the console graphics were broke? i didnt have the problem myself but it would be interesting to know how/why it happened to some.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: fromhell on April 06, 2008, 11:17:08 PM
Menu and console isn't broken here.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: Cacatoes on April 07, 2008, 02:32:08 AM
I confirm the freezing problem.
It does NOT happen with the stable binaries you just gave, and it DOES happen with the binaries you first gave (I switched back to these ones after having try the stable ones).
I didnt see a change in performance, maybe because my graphical details were quite low with almost no effect. It runs smoothly in both cases.
I had no issue with the menu either.
About the resolution detection, maybe I missed a point, but I'm not sure but the recent ioq3 helps me with that. I have usual choices, and some would be incorrect with my screen/card.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: sago007 on April 07, 2008, 04:23:04 AM
About the resolution detection, maybe I missed a point, but I'm not sure but the recent ioq3 helps me with that. I have usual choices, and some would be incorrect with my screen/card.

These also require new VM files, that I hope to be able to provide today.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: Lemonzest on April 07, 2008, 04:59:36 AM
Both versions run fine here, getting around 300-400fps with r_bloom 1

Using a ATi Radeon X800XT with Catalyst version 8.3 on Windows Server 2003 R2 SP1

When you said you need to test new VM's does that mean the one with the server list colour sorting and mouse scrolling etc? that whould be great.  :cool:


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: sago007 on April 07, 2008, 06:41:26 AM
When you said you need to test new VM's does that mean the one with the server list colour sorting and mouse scrolling etc? that whould be great.  :cool:

There are no mouse scrolling, but you can now use the numpad to type numbers.

Here are the new vm files (first test): http://ostenfeld.dk/~poul19/public_files/pak8_testvm1.pk3
Based on EliminationB17 but with GUI cleanups and all gameplay affecting CVARS disabled by default.

EDIT:
The arenas might not be displayed properly if you have pak7 in ~/.openarena


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: pikaunforgiven on April 07, 2008, 10:51:43 AM
anything we should look out for with the new vm's?  personally i dont really notice a difference ui or otherwise unless i did something wrong.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: sago007 on April 07, 2008, 11:16:48 AM
anything we should look out for with the new vm's?  personally i dont really notice a difference ui or otherwise unless i did something wrong.

The changes are kept simple, so that they do not ruin the gameplay. It is options like this:
(http://ostenfeld.dk/~poul19/public_files/gameserver.jpg)


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: pikaunforgiven on April 07, 2008, 12:03:57 PM
ohh i see, thats why i didnt notice it. i just kinda tried a bit of offline story mode (i guess thats what you would call it) and a bit of online.
there any plans on implementing other options like sv_allowdownload, sv_rate and sv_fps for the server setup menu?  its sorta a pain to set them up manually at lan parties without a dedicated server config.  i mean, i dont want it to be as complicated as something like defrag, but more options people actually use would be quite nice.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 07, 2008, 03:47:26 PM
ok, should i build some for os x?
are these similar to the ones for 0.7.6?
they come from source/code of svn?


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: Cacatoes on April 07, 2008, 06:33:46 PM
I added the VM files you gave (and checked in the UI they were used as seen on your screenshot).
I didn't experience any freeze at all while I played this evening. Quite strange ... I dunno if it's related to the addition of these files which would have fixed something.
I should do further tests ;)


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: aantipop on April 08, 2008, 06:50:32 AM
ok i fetched the openarena.stable.i386 and the pak8_testvm1.pk3.
without pak0 in my ~/.openarena/baseoa/ i couldnt run the game. (my binary is in /usr/bin/)
my pak0 - pak9 are in /usr/share/openarena/baseoa.

so i copied the pak0 to my home baseoa. when i tried to connect to a server it tried to download all the other paks, so i copied them too. when i launch the openarena.stable.i386 without my old config file i still get the red background and strange terminal. here is the output of my linux terminal:

[aantipop@blackbox bin]$ openarena.stable.i386
ioQ3 1.34-rc4+oa linux-i386 Apr  6 2008
----- FS_Startup -----
Current search path:
/home/aantipop/.openarena/baseoa/zl3tourney1.pk3 (112 files)
snip
/home/aantipop/.openarena/baseoa/0ups_beta2a.pk3 (65 files)
/home/aantipop/.openarena/baseoa
/usr/bin/baseoa

----------------------
22079 files in pk3 files
execing default.cfg
couldn't exec q3config.cfg
couldn't exec autoexec.cfg
Hunk_Clear: reset the hunk ok
----- Client Initialization -----
----- Initializing Renderer ----
-------------------------------
QKEY found.
----- Client Initialization Complete -----
----- R_Init -----

------- Input Initialization -------
Joystick is not active.
------------------------------------
...loading libGL.so.1:
Calling SDL_Init(SDL_INIT_VIDEO)...
SDL_Init(SDL_INIT_VIDEO) passed.
Initializing OpenGL display
...setting mode 3: 640 480
Using 8/8/8 Color bits, 24 depth, 0 stencil display.
GL_RENDERER: GeForce4 Ti 4400/AGP/SSE/3DNOW!
Initializing OpenGL extensions
...ignoring GL_S3_s3tc
...ignoring GL_EXT_texture_env_add
...using GL_ARB_multitexture
...using GL_EXT_compiled_vertex_array
...ignoring GL_EXT_texture_filter_anisotropic

GL_VENDOR: NVIDIA Corporation
GL_RENDERER: GeForce4 Ti 4400/AGP/SSE/3DNOW!
GL_VERSION: 1.5.8 NVIDIA 96.43.05
GL_EXTENSIONS: GL_ARB_depth_texture GL_ARB_imaging GL_ARB_multisample GL_ARB_multitexture GL_ARB_occlusion_query GL_ARB_pixel_buffer_object GL_ARB_point_parameters GL_ARB_point_sprite GL_ARB_shadow GL_ARB_shader_objects GL_ARB_shading_language_100 GL_ARB_texture_border_clamp GL_ARB_texture_compression GL_ARB_texture_cube_map GL_ARB_texture_env_add GL_ARB_texture_env_combine GL_ARB_texture_env_dot3 GL_ARB_texture_mirrored_repeat GL_ARB_texture_rectangle GL_ARB_transpose_matrix GL_ARB_vertex_buffer_object GL_ARB_vertex_program GL_ARB_vertex_shader GL_ARB_window_pos GL_S3_s3tc GL_EXT_texture_env_add GL_EXT_abgr GL_EXT_bgra GL_EXT_blend_color GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_compiled_vertex_array GL_EXT_Cg_shader GL_EXT_draw_range_elements GL_EXT_fog_coord GL_EXT_gpu_program_parameters GL_EXT_multi_draw_arrays GL_EXT_packed_pixels GL_EXT_paletted_texture GL_EXT_pixel_buffer_object GL_EXT_point_parameters GL_EXT_rescale_normal GL_EXT_secondary_color GL_EXT_separate_specular_color GL_EXT_shadow_funcs GL_EXT_shared_texture_palette GL_EXT_stencil_wrap GL_EXT_texture3D GL_EXT_texture_compression_s3tc GL_EXT_texture_cube_map GL_EXT_texture_edge_clamp GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_EXT_texture_filter_anisotropic GL_EXT_texture_lod GL_EXT_texture_lod_bias GL_EXT_texture_object GL_EXT_timer_query GL_EXT_vertex_array GL_HP_occlusion_test GL_IBM_rasterpos_clip GL_IBM_texture_mirrored_repeat GL_KTX_buffer_region GL_NV_blend_square GL_NV_copy_depth_to_color GL_NV_depth_clamp GL_NV_fence GL_NV_fog_distance GL_NV_light_max_exponent GL_NV_multisample_filter_hint GL_NV_occlusion_query GL_NV_packed_depth_stencil GL_NV_pixel_data_range GL_NV_point_sprite GL_NV_register_combiners GL_NV_register_combiners2 GL_NV_texgen_reflection GL_NV_texture_compression_vtc GL_NV_texture_env_combine4 GL_NV_texture_rectangle GL_NV_texture_shader GL_NV_texture_shader2 GL_NV_texture_shader3 GL_NV_vertex_array_range GL_NV_vertex_array_range2 GL_NV_vertex_program GL_NV_vertex_program1_1 GL_SGIS_generate_mipmap GL_SGIS_multitexture GL_SGIS_texture_lod GL_SGIX_depth_texture GL_SGIX_shadow GL_SUN_slice_accum
GL_MAX_TEXTURE_SIZE: 4096
GL_MAX_ACTIVE_TEXTURES_ARB: 4

PIXELFORMAT: color(24-bits) Z(24-bit) stencil(0-bits)
MODE: 3, 640 x 480 fullscreen hz:N/A
GAMMA: hardware w/ 0 overbright bits
CPU:
rendering primitives: single glDrawElements
texturemode: GL_LINEAR_MIPMAP_NEAREST
picmip: 1
texture bits: 0
multitexture: enabled
compiled vertex arrays: enabled
texenv add: disabled
compressed textures: disabled
Initializing Shaders
trying gfx/misc/sun.TGA...
^3WARNING: R_FindImageFile could not find 'gfx/misc/sun.tga' in shader 'sun'
^3Shader sun has a stage with no image
----- finished R_Init -----
trying gfx/misc/console01.TGA...
^3WARNING: R_FindImageFile could not find 'gfx/misc/console01.tga' in shader 'console'
^3Shader console has a stage with no image
------ Initializing Sound ------
Initializing SDL audio driver...
SDL audio driver is "dsp".
SDL_AudioSpec:
  Format:   AUDIO_S16LSB
  Freq:     22050
  Samples:  512
  Channels: 2
Starting SDL audio callback...
SDL audio initialized.
----- Sound Info -----
    1 stereo
16384 samples
   16 samplebits
    1 submission_chunk
22050 speed
0x8ba93e8 dma buffer
No background file.
----------------------
Sound initialization successful.
--------------------------------
Sound memory manager started
Loading vm file vm/ui.qvm...
...which has vmMagic VM_MAGIC_VER2
Loading 1229 jump table targets
VM file ui compiled to 629625 bytes of code
ui loaded in 1390048 bytes on the hunk
96 arenas parsed
40 bots parsed
--- Common Initialization Complete ---
Opening IP socket: localhost:27960
Hostname: blackbox.domain.org
Alias: localhost.localdomain
Alias: localhost
Alias: blackbox
IP: 127.0.0.1
Started tty console (use +set ttycon 0 to disable)


i hope this is not caused by any obscure pk3's in my folder ?


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: sago007 on April 08, 2008, 08:19:39 AM
ok i fetched the openarena.stable.i386 and the pak8_testvm1.pk3.
without pak0 in my ~/.openarena/baseoa/ i couldnt run the game. (my binary is in /usr/bin/)
my pak0 - pak9 are in /usr/share/openarena/baseoa.

My binaries does not use /usr/share for data (that is Debian specific). The game prioritizes pk3 files in .openarena over everything else, and priorities files that come alphabetically last is preferred (they override the files that are alphabetically earlier). You can cheat and place rename the pack to 'zzz_pak8.pk3' like many modifications do.

However the best way is to, place the binaries in there own folder, create a baseoa folder, place pak0-6 from 0.7 and the pak8 in that folder. And finally start the binary with the command ./ioquake3.arch +set fs_homepath /home/USERNAME/.openarena2 so that the game will use a clean config that are separated from the old one.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: aantipop on April 08, 2008, 10:29:48 AM
this is working! no more missing terminal background and such.

i have a lot of this error lines in terminal:
Com_sprintf: overflow of 63 in 62


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 10, 2008, 05:15:25 PM
i'm trying to build on mac...

this is the unstable build, all builds fine on leopard and tiger but the game icon was somewhat removed from the sources, so the bundling script fails.
i'm including a new one, it has to be placed in ./misc and is Leopard ready (the old one, done as a tiger and earlier systems one, was 128x128px, the new one, that will anyway look the same on the older systems and in the Finder, is 512x512px).
The build was done on tiger 10.4.10 and should support only tiger and leopard, if i'll have enough time i'll add support for panther (os 10.3.x) and jaguar (10.2.x) too in the release.

i tested (for a couple of minutes of network play) the ppc version on both tiger and leopard, but the intel one is totally untested.

i used the libogg 1.1.3 and libvorbis 1.1.2 that i built for 0.7.0 release and i will not yet update libvorbis to 1.2.0 since i don't have an intel mac at home and will not have access to an os x capable x86 machine in time for 15th. i still failed to crossbuild libvorbisfile, so i can't update to libvorbis 1.2 using only my ppc macs.

you can download the build here (http://www.jazzmobile.it/jackoverfull/games/openarena/oabin-test-2008-04-05.zip). just place it in the same folder of baseoa.

a question to anyone who knows: why are cgamexxx.dylib, qagamexxx.dylib and uiixxx.dylib (where xxx is the platform) included in the bundle? it is a ioq3 thing and at the time of the 0.7.0 release they where built but not included...
and what are those files anyway? i remember to have seen similar files (with .so extensions on linux and .dll on windows) in all quake-based games...




i'm also trying to build the "stable" one, but i had some problems. i think that they're related to gcc4, but i'm still not sure, so i'm performing furter tests. obviusly, it will not build on leopard, but this will no longer be a problem for 0.7.6 release, since i've prepared a tiger enviroment...


the leopard-ready icon (http://www.jazzmobile.it/jackoverfull/games/openarena/quake3.icns). please, put it alongside the others, in ./misc, when you'll commit to the svn...


good testing, mac users!


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: sago007 on April 10, 2008, 05:53:16 PM
a question to anyone who knows: why are cgamexxx.dylib, qagamexxx.dylib and uiixxx.dylib (where xxx is the platform) included in the bundle? it is a ioq3 thing and at the time of the 0.7.0 release they where built but not included...
and what are those files anyway? i remember to have seen similar files (with .so extensions on linux and .dll on windows) in all quake-based games...

They contain a platform specific version of the gamecode and can be used instead of the Virtual Machine. However they platform specific rules for things like rounding making them a bad choice especially for network play. There has also been several problems on special platforms (including 64 bit linux).


I don't know much about Macs, but there is a script in the source for generating an "universal build".


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: pikaunforgiven on April 10, 2008, 06:20:10 PM
universal build = binary that works on both the powerpc and newer x86 macintosh platforms.

ah yes, the dll's.... i remember having to replace them with "speed dll's" for q3 because id's vm code didnt run well on amd processors. apparently their 3dnow code was broken or something, causing slow performance on amd processors but leaving intel processors just fine. really the only case where using the dll's was an advantage over using the vm code. not sure if ioquake fixed that though, never really checked.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 09:04:53 AM


They contain a platform specific version of the gamecode and can be used instead of the Virtual Machine. However they platform specific rules for things like rounding making them a bad choice especially for network play. There has also been several problems on special platforms (including 64 bit linux).


thanks. so it's better to not include them in the release, right?

Quote
I don't know much about Macs, but there is a script in the source for generating an "universal build".
yes, it' a simple shell script that automates a lot of things. very useful.
the universal binaries are, as pikaunforgiven wrote, binaries that runs on both intel and ppc macs (really two different binaries merged together, usually using a program called "lipo", a method that comes directly from the NeXT lecacy of os x).
applications and libraries can be usually cross-compiled for both archs on a single mac, but sometimes there are problems. for oa i was never able to crossbuild libvorbisfile, wich is needed for the music, so i built this lib on a pc wich was running os x, lent by a friend at the time of 0.7.0 release.
the game itself usually builds fine on ppc macs also for x86 ones, but i'm having some problems with the "stable" sources.
now i'm sure that they builds fine with gcc 3, i'm trying with gcc 4 (wich is required for crossbuild".


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: pikaunforgiven on April 11, 2008, 10:08:50 AM
thanks. so it's better to not include them in the release, right?

naa, some people still use the libraries for one reason or another, since for some people the dll's are faster. if the amd bug was fixed in ioquake, then it probably wouldnt matter so much anymore though.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 11:30:25 AM
just wondering why in ioq3 builds now  they are included while wasn't some months ago…

regarding the build of the "stable" sources, i can now confirm that it always fails on os x using gcc 4 (doesn't matter if cross-compiling or not), still don't know exacly why. it seems that it can't find the ogg and vorbis libraries, but they are installed (and recognized correctly by oa 0.7.0, the "unstable" sources and latest ioq3, always using gcc 4).
i'm including the log, i'll try to find a solution (this evening or tomorrow), but any hint will be surely appreciated…
Code:
jackoverfull:~/oa/ioquake3stable/1.34 jack$ CC=gcc-4.0 BUILD_MACOSX_UB=pcc make
QVM tools not built when cross-compiling

Building ioquake3 in build/release-darwin-ppc:
  CC: gcc-4.0

  CFLAGS:
    -Wall
    -Wimplicit
    -Wstrict-prototypes
    -fno-strict-aliasing
    -DMACOS_X
    -fno-common
    -pipe
    -gfull
    -DUSE_OPENAL=1
    -DUSE_OPENAL_DLOPEN=1
    -DUSE_CURL=1
    -DUSE_CURL_DLOPEN=1
    -DUSE_CODEC_VORBIS=1
    -DUSE_SDL_VIDEO=1
    -DUSE_SDL_SOUND=1
    -D_THREAD_SAFE=1
    -Icode/SDL12/include
    -DUSE_LOCAL_HEADERS=1
    -MMD
    -DNDEBUG
    -faltivec
    -O3
    -ffast-math
    -falign-loops=16

  Output:
    build/release-darwin-ppc/ioq3ded.ppc
    build/release-darwin-ppc/ioquake3.ppc

DED_CC code/server/sv_bot.c
DED_CC code/server/sv_client.c
DED_CC code/server/sv_ccmds.c
DED_CC code/server/sv_game.c
DED_CC code/server/sv_init.c
DED_CC code/server/sv_main.c
DED_CC code/server/sv_net_chan.c
DED_CC code/server/sv_snapshot.c
DED_CC code/server/sv_world.c
DED_CC code/qcommon/cm_load.c
DED_CC code/qcommon/cm_patch.c
DED_CC code/qcommon/cm_polylib.c
DED_CC code/qcommon/cm_test.c
DED_CC code/qcommon/cm_trace.c
DED_CC code/qcommon/cmd.c
DED_CC code/qcommon/common.c
DED_CC code/qcommon/cvar.c
DED_CC code/qcommon/files.c
DED_CC code/qcommon/md4.c
DED_CC code/qcommon/msg.c
DED_CC code/qcommon/net_chan.c
DED_CC code/qcommon/huffman.c
DED_CC code/qcommon/q_math.c
DED_CC code/qcommon/q_shared.c
DED_CC code/qcommon/unzip.c
DED_CC code/qcommon/vm.c
DED_CC code/qcommon/vm_interpreted.c
BOT_CC code/botlib/be_aas_bspq3.c
BOT_CC code/botlib/be_aas_cluster.c
BOT_CC code/botlib/be_aas_debug.c
BOT_CC code/botlib/be_aas_entity.c
BOT_CC code/botlib/be_aas_file.c
BOT_CC code/botlib/be_aas_main.c
BOT_CC code/botlib/be_aas_move.c
BOT_CC code/botlib/be_aas_optimize.c
BOT_CC code/botlib/be_aas_reach.c
BOT_CC code/botlib/be_aas_route.c
BOT_CC code/botlib/be_aas_routealt.c
BOT_CC code/botlib/be_aas_sample.c
BOT_CC code/botlib/be_ai_char.c
BOT_CC code/botlib/be_ai_chat.c
BOT_CC code/botlib/be_ai_gen.c
BOT_CC code/botlib/be_ai_goal.c
BOT_CC code/botlib/be_ai_move.c
BOT_CC code/botlib/be_ai_weap.c
BOT_CC code/botlib/be_ai_weight.c
BOT_CC code/botlib/be_ea.c
BOT_CC code/botlib/be_interface.c
BOT_CC code/botlib/l_crc.c
BOT_CC code/botlib/l_libvar.c
BOT_CC code/botlib/l_log.c
BOT_CC code/botlib/l_memory.c
BOT_CC code/botlib/l_precomp.c
BOT_CC code/botlib/l_script.c
BOT_CC code/botlib/l_struct.c
DED_CC code/unix/linux_signals.c
DED_CC code/unix/unix_main.c
code/unix/unix_main.c:1363:2: warning: #warning Sys_BinName not implemented
DED_CC code/unix/unix_net.c
DED_CC code/unix/unix_shared.c
DED_CC code/null/null_client.c
DED_CC code/null/null_input.c
DED_CC code/null/null_snddma.c
DED_CC code/qcommon/vm_ppc.c
LD build/release-darwin-ppc/ioq3ded.ppc
CC code/client/cl_cgame.c
CC code/client/cl_cin.c
CC code/client/cl_console.c
CC code/client/cl_input.c
CC code/client/cl_keys.c
CC code/client/cl_main.c
CC code/client/cl_net_chan.c
CC code/client/cl_parse.c
CC code/client/cl_scrn.c
CC code/client/cl_ui.c
CC code/client/cl_avi.c
CC code/qcommon/cm_load.c
CC code/qcommon/cm_patch.c
CC code/qcommon/cm_polylib.c
CC code/qcommon/cm_test.c
CC code/qcommon/cm_trace.c
CC code/qcommon/cmd.c
CC code/qcommon/common.c
CC code/qcommon/cvar.c
CC code/qcommon/files.c
CC code/qcommon/md4.c
CC code/qcommon/md5.c
CC code/qcommon/msg.c
CC code/qcommon/net_chan.c
CC code/qcommon/huffman.c
CC code/client/snd_adpcm.c
CC code/client/snd_dma.c
CC code/client/snd_mem.c
CC code/client/snd_mix.c
CC code/client/snd_wavelet.c
CC code/client/snd_main.c
CC code/client/snd_codec.c
CC code/client/snd_codec_wav.c
CC code/client/snd_codec_ogg.c
code/client/snd_codec_ogg.c:34:31: error: vorbis/vorbisfile.h: No such file or directory
code/client/snd_codec_ogg.c:107: error: parse error before 'ogg_int64_t'
code/client/snd_codec_ogg.c:108: warning: function declaration isn't a prototype
code/client/snd_codec_ogg.c: In function 'S_OGG_Callback_seek':
code/client/snd_codec_ogg.c:113: error: 'datasource' undeclared (first use in this function)
code/client/snd_codec_ogg.c:113: error: (Each undeclared identifier is reported only once
code/client/snd_codec_ogg.c:113: error: for each function it appears in.)
code/client/snd_codec_ogg.c:123: error: 'whence' undeclared (first use in this function)
code/client/snd_codec_ogg.c:128: error: 'offset' undeclared (first use in this function)
code/client/snd_codec_ogg.c: At top level:
code/client/snd_codec_ogg.c:213: error: parse error before 'S_OGG_Callbacks'
code/client/snd_codec_ogg.c:213: warning: type defaults to 'int' in declaration of 'S_OGG_Callbacks'
code/client/snd_codec_ogg.c:215: warning: initialization makes integer from pointer without a cast
code/client/snd_codec_ogg.c:216: warning: excess elements in scalar initializer
code/client/snd_codec_ogg.c:216: warning: (near initialization for 'S_OGG_Callbacks')
code/client/snd_codec_ogg.c:217: warning: excess elements in scalar initializer
code/client/snd_codec_ogg.c:217: warning: (near initialization for 'S_OGG_Callbacks')
code/client/snd_codec_ogg.c:219: warning: excess elements in scalar initializer
code/client/snd_codec_ogg.c:219: warning: (near initialization for 'S_OGG_Callbacks')
code/client/snd_codec_ogg.c:219: warning: data definition has no type or storage class
code/client/snd_codec_ogg.c: In function 'S_OGG_CodecOpenStream':
code/client/snd_codec_ogg.c:231: error: 'OggVorbis_File' undeclared (first use in this function)
code/client/snd_codec_ogg.c:231: error: 'vf' undeclared (first use in this function)
code/client/snd_codec_ogg.c:234: error: 'vorbis_info' undeclared (first use in this function)
code/client/snd_codec_ogg.c:234: error: 'OGGInfo' undeclared (first use in this function)
code/client/snd_codec_ogg.c:235: error: 'ogg_int64_t' undeclared (first use in this function)
code/client/snd_codec_ogg.c:235: error: parse error before 'numSamples'
code/client/snd_codec_ogg.c:260: warning: implicit declaration of function 'ov_open_callbacks'
code/client/snd_codec_ogg.c:270: warning: implicit declaration of function 'ov_seekable'
code/client/snd_codec_ogg.c:272: warning: implicit declaration of function 'ov_clear'
code/client/snd_codec_ogg.c:282: warning: implicit declaration of function 'ov_streams'
code/client/snd_codec_ogg.c:294: warning: implicit declaration of function 'ov_info'
code/client/snd_codec_ogg.c:307: error: 'numSamples' undeclared (first use in this function)
code/client/snd_codec_ogg.c:307: warning: implicit declaration of function 'ov_pcm_total'
code/client/snd_codec_ogg.c: In function 'S_OGG_CodecCloseStream':
code/client/snd_codec_ogg.c:340: error: 'OggVorbis_File' undeclared (first use in this function)
code/client/snd_codec_ogg.c:340: error: parse error before ')' token
code/client/snd_codec_ogg.c: In function 'S_OGG_CodecReadStream':
code/client/snd_codec_ogg.c:389: warning: implicit declaration of function 'ov_read'
code/client/snd_codec_ogg.c:389: error: 'OggVorbis_File' undeclared (first use in this function)
code/client/snd_codec_ogg.c:389: error: parse error before ')' token
code/client/snd_codec_ogg.c:361: warning: unused variable 'BS'
make[2]: *** [build/release-darwin-ppc/client/snd_codec_ogg.o] Error 1
make[1]: *** [targets] Error 2
make: *** [release] Error 2




Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: tw3k on April 11, 2008, 01:33:49 PM
I just edit the Makefile like so:
Quote
BUILD_CLIENT     = 1
BUILD_CLIENT_SMP = 0
BUILD_SERVER     = 0
BUILD_GAME_SO    = 1
BUILD_GAME_QVM   = 0

And run make.

That is with GNU Make 3.80 and gcc 4.0.1 on Tiger.

It doesn't build a nice clickable Mac app but I like running it from the Terminal anyways.

Never bothered installing vorbis.

maybe add:
USE_CODEC_VORBIS = 0

ans see if you can make it without vorbis.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 02:58:49 PM
i'll try in a couple of minutes, but as far i can see this shouldn't affect the compilation problem i'm having…

Quote
That is with GNU Make 3.80 and gcc 4.0.1 on Tiger.
are you on an intel mac?

regarding the qvm, my question was only if it is better to include them in the release or not…
Quote
Never bothered installing vorbis.

maybe add:
USE_CODEC_VORBIS = 0

ans see if you can make it without vorbis.
tried already (forgot to wrote): same problem even with vorbis support disabled.

not sure on what "BUILD_GAME_SO" should do…

thanks for your suggestions, anyway.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 03:01:58 PM
tried. no changes.

investigating right now…


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: sago007 on April 11, 2008, 03:14:15 PM
not sure on what "BUILD_GAME_SO" should do…

It is the dll, so, dyn*** files.

There have been some security updates to ioquake, so I might soon upload a new 1288 with backports from newer versions (I wont use the new versions directly as they introduce some rapidly changing code)


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: tw3k on April 11, 2008, 03:16:04 PM

not sure on what "BUILD_GAME_SO" should do…



builds the shared object, dylib, rather than the qvm.

run with:
 +set vm_cgame 0 +set vm_game 0 +set vm_ui 0

nope, not on an intel mac, old ppc7400 that I don't use no mo.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 03:21:20 PM


builds the shared object, dylib, rather than the qvm.

ok, thanks.

Quote
nope, not on an intel mac, old ppc7400 that I don't use no mo.
and are you sure that you tried to build the "stable" sources posted at the top of this topic (not the one in svn neither the 1288 one, which builds fine for mee too) using gcc 4?


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 03:29:10 PM
trying to force ogg and vorbis recognition...
for now it's building and it's already far beyond that point.

crossing fingers... :-D

but if that works...why it doesn't work in the normal way? :-/


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 03:32:31 PM
it built.

but the linker cannot locate libvorbisfile.dylib, i'll have to find a way to do it manually or to pass it's location directly.

i'll do that later, now i've to go.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: tw3k on April 11, 2008, 03:37:10 PM


Quote
nope, not on an intel mac, old ppc7400 that I don't use no mo.
and are you sure that you tried to build the "stable" sources posted at the top of this topic (not the one in svn neither the 1288 one, which builds fine for mee too) using gcc 4?

oops, nope, regular ioq3 build.

building Stable (svn 1127) now.

Looks like the vorbis var is in Makefile.local


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 03:51:51 PM

oops, nope, regular ioq3 build.

building Stable (svn 1127) now.

ah, ok.
that's building fine with vorbis for me too.


Quote
Looks like the vorbis var is in Makefile.local
tnx, looking...


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 03:57:15 PM
Code:
USE_CODEC_VORBIS=1
interesting...

that should explain why it's building ogg vorbis even if it's disabled in the main makefile. i'm not sure that it explains also why it builds it anyway when it's specified from the command line to not build it, but i'll take that as true...

but still i don't have any idea on why it doesn't find the vorbisfile header with gcc4.

trying to build with that option setted to 0, just to see if it works, in the meanwhile i'll try to solve the problem.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 04:01:36 PM
negative: it tried to build with vorbis support even with USE_CODEC_VORBIS=0 in makefile.local! :-/

trying another way...


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: tw3k on April 11, 2008, 04:06:36 PM
Stable (svn 1127) error'd out on OpenAL.
Quote
make[2]: `build/release-darwin-ppc/ioq3ded.ppc' is up to date.
LD build/release-darwin-ppc/ioquake3.ppc

[big snip]

/usr/bin/ld: multiple definitions of symbol _alcProcessContext
build/release-darwin-ppc/client/qal.o definition of _alcProcessContext in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alcProcessContext in section (__DATA,__common)
/usr/bin/ld: multiple definitions of symbol _alcSuspendContext
build/release-darwin-ppc/client/qal.o definition of _alcSuspendContext in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alcSuspendContext in section (__DATA,__common)
collect2: ld returned 1 exit status
make[2]: *** [build/release-darwin-ppc/ioquake3.ppc] Error 1
make[1]: *** [targets] Error 2
make: *** [release] Error 2
Quote

I'll try again with my old Makefle.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 04:13:18 PM
yeah, fromhell told me in chat that the svn source is outdated and we should use the ones provided in this topic. it doesn't build for me too.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 04:26:42 PM
ok, this doesn't make any sense to me, but i had that idea and i had to try...

so, my ogg and vorbis libraries are all built in /Developer/SDK/MacOSX10.4u.sdk since i'm using that sdk to build the game (at least for now). remember that everithing works perfectly with older oa release, with newer sources, with vanilla ioq3 and even with the "stable" sources while using gcc 3.

well, for some obscure reason the "stable" sources with gcc 4 ignores the "include" folder included in the sdk. if i link the ogg and vorbis folders in the main include folder it builds. :-/

since i don't want to track down and link every file manually i'll build libogg and libvorbis again, this time using /usr/local as the prefix, then try to build again. i can't build libvorbisfile universally, but it should be enough to relocate it with install_name_tool, as usual.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 11, 2008, 04:47:14 PM
it builds but it fails to complete (linking?). if i remember correctly (looking also at tw3k post) it is the same problem of the "experimental" sources in svn.

Code:
ranlib build/release-darwin-ppc/libSDLmain.a
LD build/release-darwin-ppc/ioquake3.ppc
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alBuffer3f
build/release-darwin-ppc/client/qal.o definition of _alBuffer3f in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alBuffer3f in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alBuffer3i
build/release-darwin-ppc/client/qal.o definition of _alBuffer3i in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alBuffer3i in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alBufferData
build/release-darwin-ppc/client/qal.o definition of _alBufferData in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alBufferData in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alBufferf
build/release-darwin-ppc/client/qal.o definition of _alBufferf in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alBufferf in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alBufferfv
build/release-darwin-ppc/client/qal.o definition of _alBufferfv in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alBufferfv in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alBufferi
build/release-darwin-ppc/client/qal.o definition of _alBufferi in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alBufferi in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alBufferiv
build/release-darwin-ppc/client/qal.o definition of _alBufferiv in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alBufferiv in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alDeleteBuffers
build/release-darwin-ppc/client/qal.o definition of _alDeleteBuffers in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alDeleteBuffers in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alDeleteSources
build/release-darwin-ppc/client/qal.o definition of _alDeleteSources in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alDeleteSources in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alDisable
build/release-darwin-ppc/client/qal.o definition of _alDisable in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alDisable in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alDistanceModel
build/release-darwin-ppc/client/qal.o definition of _alDistanceModel in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alDistanceModel in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alDopplerFactor
build/release-darwin-ppc/client/qal.o definition of _alDopplerFactor in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alDopplerFactor in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alDopplerVelocity
build/release-darwin-ppc/client/qal.o definition of _alDopplerVelocity in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alDopplerVelocity in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alEnable
build/release-darwin-ppc/client/qal.o definition of _alEnable in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alEnable in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGenBuffers
build/release-darwin-ppc/client/qal.o definition of _alGenBuffers in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGenBuffers in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGenSources
build/release-darwin-ppc/client/qal.o definition of _alGenSources in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGenSources in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetBoolean
build/release-darwin-ppc/client/qal.o definition of _alGetBoolean in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetBoolean in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetBooleanv
build/release-darwin-ppc/client/qal.o definition of _alGetBooleanv in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetBooleanv in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetBuffer3f
build/release-darwin-ppc/client/qal.o definition of _alGetBuffer3f in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetBuffer3f in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetBuffer3i
build/release-darwin-ppc/client/qal.o definition of _alGetBuffer3i in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetBuffer3i in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetBufferf
build/release-darwin-ppc/client/qal.o definition of _alGetBufferf in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetBufferf in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetBufferfv
build/release-darwin-ppc/client/qal.o definition of _alGetBufferfv in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetBufferfv in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetBufferi
build/release-darwin-ppc/client/qal.o definition of _alGetBufferi in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetBufferi in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetBufferiv
build/release-darwin-ppc/client/qal.o definition of _alGetBufferiv in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetBufferiv in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetDouble
build/release-darwin-ppc/client/qal.o definition of _alGetDouble in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetDouble in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetDoublev
build/release-darwin-ppc/client/qal.o definition of _alGetDoublev in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetDoublev in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetEnumValue
build/release-darwin-ppc/client/qal.o definition of _alGetEnumValue in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetEnumValue in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetError
build/release-darwin-ppc/client/qal.o definition of _alGetError in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetError in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetFloat
build/release-darwin-ppc/client/qal.o definition of _alGetFloat in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetFloat in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetFloatv
build/release-darwin-ppc/client/qal.o definition of _alGetFloatv in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetFloatv in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetInteger
build/release-darwin-ppc/client/qal.o definition of _alGetInteger in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetInteger in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetIntegerv
build/release-darwin-ppc/client/qal.o definition of _alGetIntegerv in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetIntegerv in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetListener3f
build/release-darwin-ppc/client/qal.o definition of _alGetListener3f in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetListener3f in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetListener3i
build/release-darwin-ppc/client/qal.o definition of _alGetListener3i in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetListener3i in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetListenerf
build/release-darwin-ppc/client/qal.o definition of _alGetListenerf in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetListenerf in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetListenerfv
build/release-darwin-ppc/client/qal.o definition of _alGetListenerfv in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetListenerfv in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetListeneri
build/release-darwin-ppc/client/qal.o definition of _alGetListeneri in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetListeneri in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetListeneriv
build/release-darwin-ppc/client/qal.o definition of _alGetListeneriv in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetListeneriv in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetProcAddress
build/release-darwin-ppc/client/qal.o definition of _alGetProcAddress in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetProcAddress in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetSource3f
build/release-darwin-ppc/client/qal.o definition of _alGetSource3f in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetSource3f in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetSource3i
build/release-darwin-ppc/client/qal.o definition of _alGetSource3i in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetSource3i in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetSourcef
build/release-darwin-ppc/client/qal.o definition of _alGetSourcef in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alGetSourcef in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alGetSourcefv
[...]
build/release-darwin-ppc/client/snd_openal.o definition of _alcGetString in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alcIsExtensionPresent
build/release-darwin-ppc/client/qal.o definition of _alcIsExtensionPresent in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alcIsExtensionPresent in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alcMakeContextCurrent
build/release-darwin-ppc/client/qal.o definition of _alcMakeContextCurrent in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alcMakeContextCurrent in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alcOpenDevice
build/release-darwin-ppc/client/qal.o definition of _alcOpenDevice in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alcOpenDevice in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alcProcessContext
build/release-darwin-ppc/client/qal.o definition of _alcProcessContext in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alcProcessContext in section (__DATA,__common)
/usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: multiple definitions of symbol _alcSuspendContext
build/release-darwin-ppc/client/qal.o definition of _alcSuspendContext in section (__DATA,__common)
build/release-darwin-ppc/client/snd_openal.o definition of _alcSuspendContext in section (__DATA,__common)
collect2: ld returned 1 exit status
make[2]: *** [build/release-darwin-ppc/ioquake3.ppc] Error 1
make[1]: *** [targets] Error 2
make: *** [release] Error 2


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: tw3k on April 11, 2008, 04:51:36 PM
lets see, it built with the following Makefile.local:

Code:
COPYDIR="/usr/local/games/openarena"

USE_OPENAL=0

#Next we want to bind openAL on runtime.
USE_OPENAL_DLOPEN=0

USE_CODEC_VORBIS=0


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: sago007 on April 11, 2008, 05:02:47 PM
it builds but it fails to complete (linking?). if i remember correctly (looking also at tw3k post) it is the same problem of the "experimental" sources in svn.

Looks like it is openAL it fails to link.

Be the way, I have uploaded new experimental binaries with Mac icon and a few security fixes.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: tw3k on April 11, 2008, 05:38:01 PM

Looks like it is openAL it fails to link.


Yup, I'm giving it another try with the OpenAL and Vorbis frameworks.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: Lemonzest on April 11, 2008, 05:59:54 PM
Getting way better performance with this exe (windows), also has the official q3a logo in the corner and task switch menu, think that needs to go. sometimes i'll also get a micro freeze.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: tw3k on April 11, 2008, 06:34:42 PM
The OpenAL framework linked. Vorbis looks like it is failing with a SDL error. Don't have SDL either.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: sago007 on April 11, 2008, 06:56:42 PM
also has the official q3a logo in the corner and task switch menu, think that needs to go.

It does...

I wonder why its kept separate from the ordinary icon.

I'll try compiling in Visual Studio to work. It could make a huge difference on the Windows performance


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 12, 2008, 12:09:12 AM
it builds but it fails to complete (linking?). if i remember correctly (looking also at tw3k post) it is the same problem of the "experimental" sources in svn.

Looks like it is openAL it fails to link.

what are the cons of building without openal?
i'm remembering something about it and ioq3 from the last time: was it disabled for os x?

Quote
Be the way, I have uploaded new experimental binaries with Mac icon and a few security fixes.

thanks.
i'll build them.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: Lemonzest on April 12, 2008, 04:17:10 AM
is r_bloom much brighter in this version? looks much better imo


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: sago007 on April 12, 2008, 04:20:18 AM
what are the cons of building without openal?

No 5.1 sound. The cons of building without vorbis is bigger as a lot of the new sounds are in vorbis, although nothing important.


Title: Re: New binaries (ioquake SVN 1288) (now also stable ioquake), please test
Post by: jackoverfull on April 12, 2008, 04:33:49 AM


No 5.1 sound.
well...
newer macs support 5.1 sounds, so it should be better to use openal if possibile...

Quote
The cons of building without vorbis is bigger as a lot of the new sounds are in vorbis, although nothing important.
well, essentially the music is in the vorbis format...


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: jackoverfull on April 12, 2008, 06:28:17 AM
tried to build without openal. still doesn't build.

i built the new 1288plus1 sources, will bundle the libs and upload soon.

i see that the cgame and similar files are no longer included. if they were not needed that's probably a good idea.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: jackoverfull on April 12, 2008, 07:30:59 AM
uploaded 1288plus1 binaries.

here to download (http://www.jazzmobile.it/jackoverfull/games/openarena/oabin-test-2008-04-11.zip).


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: jackoverfull on April 12, 2008, 07:34:00 AM
it even works on a vanilla 10.3.9 installation! :-D


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: jackoverfull on April 12, 2008, 07:36:00 AM
i wonder if openal isenabled in that sources: 0.7.0 didn't run on a standard 10.3.9 installation becouse of the lacks of that lib...


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: sago007 on April 12, 2008, 07:59:04 AM
i wonder if openal isenabled in that sources: 0.7.0 didn't run on a standard 10.3.9 installation becouse of the lacks of that lib...

I have sat:
USE_OPENAL_DLOPEN=1

This means that openAL will be used if available.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: sago007 on April 12, 2008, 08:01:29 AM
And be the way.

This file:
http://ostenfeld.dk/~poul19/public_files/sdl_icon.h

need to be placed in code/sdl/ to replace the quake3 logo in the title line.

Also: I have succeeded in building openarena in Visual studio 2008 a new binary will appear soon.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: jackoverfull on April 12, 2008, 08:04:16 AM
i wonder if openal isenabled in that sources: 0.7.0 didn't run on a standard 10.3.9 installation becouse of the lacks of that lib...

I have sat:
USE_OPENAL_DLOPEN=1

This means that openAL will be used if available.
at runtime? great!


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: fromhell on April 13, 2008, 10:39:04 PM
Added Windows binaries compiled with Visual Studio 2008.
There goes the Windows9x crowd. They're still targets, btw. VS2008 compiles stuff that forces links against kernel32.dll/shell32.dll functions that don't exist in these operating systems just to try give some xp sales.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: Lemonzest on April 14, 2008, 03:59:14 AM
the msvc2008 builds seem a little slower, not hitting the same fps as the mingw builds are, does someone have a demo we could use for timedemo testing?


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: sago007 on April 14, 2008, 04:29:24 AM
I know that the MSVC lacks Win9x support and I'm not really interested in there performance.

The thing I'm interested in is weather or not the micro freezes still exist.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: Lemonzest on April 14, 2008, 05:22:05 AM
about the micro freezes, they was caused by a program eating the cpu (yahoo beta) its all fine now.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: pikaunforgiven on April 14, 2008, 10:48:59 AM
as for the test vm's, you may wanna coordinate with this guy so you dont have people getting conflicts with two different ones being tested at once  http://openarena.ws/board/index.php?topic=1093.0


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: jackoverfull on April 14, 2008, 10:55:32 AM
now...

the "stable" sources are the ones that will be used in the upcoming release, right?

what should i do for the mac version, since the sources are currently not buildable on mac using gcc4?

the options are:
-we find some way to build them or to fix whatever needs to be fixed;
-no mac release (somewhat not like that...);
-mac release for ppc only macs (since i can build with gcc3 for ppcs)...now lot of macs are intel based...
-mac release with another sources (the unstabe ones?);
-postpone release until the problem is solved;

any other?


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: sago007 on April 14, 2008, 12:17:57 PM
now...

the "stable" sources are the ones that will be used in the upcoming release, right?
Properly. Since I cannot find any reason for (/reproduce) the micro freezes


what should i do for the mac version, since the sources are currently not buildable on mac using gcc4?

the options are:
-we find some way to build them or to fix whatever needs to be fixed;
-no mac release (somewhat not like that...);
-mac release for ppc only macs (since i can build with gcc3 for ppcs)...now lot of macs are intel based...
-mac release with another sources (the unstabe ones?);
-postpone release until the problem is solved;

any other?

Use a different engine for the Mac. All the engines are fully compatible with each other and the QVMs.

as for the test vm's, you may wanna coordinate with this guy so you dont have people getting conflicts with two different ones being tested at once  http://openarena.ws/board/index.php?topic=1093.0

That was the reason I bumped it up. It is a rather old topic but I despite a lot of his work was put into the ioquake release there are missing a few pieces.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: jackoverfull on April 14, 2008, 02:52:02 PM


Use a different engine for the Mac. All the engines are fully compatible with each other and the QVMs.
which one? remain at the 0.7.0 engine release? or include directly the 1288+1?


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: jackoverfull on April 14, 2008, 03:14:15 PM
maybe i should try to build the original revision vanilla engine, then, if it builds, apply the patches...

to do so i should checkout 1127 from

svn.icculus.org/quake3/branches/1.34

right?

and patch with the bloom.diff and ioq3_1106_oggsub.diff included in your sources files...


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: sago007 on April 14, 2008, 03:34:58 PM
maybe i should try to build the original revision vanilla engine, then, if it builds, apply the patches...

to do so i should checkout 1127 from

svn.icculus.org/quake3/branches/1.34

right?

and patch with the bloom.diff and ioq3_1106_oggsub.diff included in your sources files...

This is exactly where the stable source I provided comes from. So if it doesn't compile then this properly wont compile either.

You don't need to checkout a specified revision. 1.34 has not been updated since 1127

which one? remain at the 0.7.0 engine release? or include directly the 1288+1?

If the 1288+1 runs fine on Mac when it should be included... of course it might be hard getting enough testing of it.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: jackoverfull on April 14, 2008, 03:57:31 PM
Quote
This is exactly where the stable source I provided comes from. So if it doesn't compile then this properly wont compile either.
just tried. it compiles fine.

Quote
You don't need to checkout a specified revision. 1.34 has not been updated since 1127
ok...

Quote
If the 1288+1 runs fine on Mac when it should be included... of course it might be hard getting enough testing of it.
right...


anyway, i tried to apply the patches and build.
the ioq3_1106_oggsub.diff applies and builds fine.

the bloom.diff exits with
Code:
ibooky:1127 carlo$ cat /Users/carlo/Applications/Progetti/OpenArena/0.7.6/ioquake3stable/1.34/bloom.diff | patch -p0
(Stripping trailing CRs from patch.)
patching file code/renderer/tr_backend.c
(Stripping trailing CRs from patch.)
patching file code/renderer/tr_bloom.c
(Stripping trailing CRs from patch.)
patching file code/renderer/tr_init.c
(Stripping trailing CRs from patch.)
patching file code/renderer/tr_local.h
(Stripping trailing CRs from patch.)
patching file Makefile
Hunk #3 succeeded at 216 (offset -10 lines).
Hunk #4 succeeded at 853 (offset 29 lines).
Hunk #5 succeeded at 940 (offset 38 lines).
Hunk #6 succeeded at 1008 (offset 38 lines).
Hunk #7 FAILED at 1242.
Hunk #8 FAILED at 1314.
Hunk #9 succeeded at 1161 (offset -153 lines).
Hunk #10 FAILED at 1300.
Hunk #11 succeeded at 1701 with fuzz 1 (offset -154 lines).
3 out of 11 hunks FAILED -- saving rejects to file Makefile.rej
ibooky:1127 carlo$
and, obviusly, doesn't build at all...

shouldn't be in here also something like oa-changes.diff?


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: sago007 on April 14, 2008, 04:16:17 PM
the bloom.diff exits with

and, obviusly, doesn't build at all...

shouldn't be in here also something like oa-changes.diff?

1. Look in Makefile.rej
They started to use a different Makefile layout

2. it is in the code-folder there it expects to be applied.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: jackoverfull on April 14, 2008, 04:55:05 PM
thanks.

trying to apply manually...

there are some differences here, but it shouldn't matter, since i'm not building the dedicated server version for now.

trying to build...


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: sago007 on April 18, 2008, 01:11:33 PM
For Windows users that experience sticky keys with the SDL build can try this binary:
http://brie.ostenfeld.dk/~poul19/public_files/openarena_svn1288_sagoinput.zip

I believe the problem is caused by overflowing the SDL event queue. I have lifted the pressure on the queue in this binary although it can still occur of you press many thousands of keys per second or a very slow computer.


Title: Re: New binaries (ioquake SVN 1288) (updated 04-12), please test
Post by: Lemonzest on April 18, 2008, 04:59:01 PM
Just tried the sagoinput version, and i must say its much better, i can strafe jump much quicker and keys seem more responsive, and the mouse is a little smoother not jittering sometimes.