0.3

[Z-Man]

Looks like we need a new 64-bit builder. Our sysadmin decided that the new Ubuntu should be a 32 bit version. In the long term, there may be a workaround (cross compilation?), but right now, I can't build for x86.

[Luke-Jr]

aabeta's server is 64-bit, so feel free to build there...

[k]

The situation seems under control now, we've handled it on IRC. I'm currently using revision 4872 to do an alpha build.

I figured something was going on on IRC. I glanced over wrtlprnft's chat log afterwards. In case anyone is interested, here's a batch file for generating the documentation on Windows (based on the 0.2.8.2 branch. minor mods needed for the trunk). It goes in the same directoy as makedist.bat and requires a few Windows ports of Unix utils.

Code: Select all

@echo off
SETLOCAL
REM ##########################################################
REM #
REM # Generate the Armagetron Advanced documentation
REM #
REM # This script is written for Windows XP / 2000 and
REM # relies on several Unix utilities ported to Windows.
REM # Make sure these executables are in the PATH or the
REM # same directory as this batch file.
REM #
REM # CAT.EXE  - http://unxutils.sourceforge.net/UnxUpdates.zip
REM # GAWK.EXE - http://gnuwin32.sourceforge.net/packages/gawk.htm
REM # GREP.EXE - http://gnuwin32.sourceforge.net/packages/grep.htm
REM # M4.EXE   - http://gnuwin32.sourceforge.net/packages/m4.htm
REM # SED.EXE  - http://gnuwin32.sourceforge.net/packages/sed.htm
REM #
REM ##########################################################

SET PROG_TITLE=Armagetron Advanced
SET PROG_NAME=armagetronad
SET AA_DIR=..\armagetronad
SET AA_BUILD_DIR=..
SET SERVER_EXE=armagetronad_dedicated.exe

SET DIST_DIR_BASE=dist
SET DEBUG_DIR_BASE=debug
SET PROFILE_DIR_BASE=profile
SET DIST_DIR=%AA_BUILD_DIR%\%DIST_DIR_BASE%
SET DEBUG_DIR=%AA_BUILD_DIR%\%DEBUG_DIR_BASE%
SET PROFILE_DIR=%AA_BUILD_DIR%\%PROFILE_DIR_BASE%

SET AWK_EXE=gawk.exe
SET CAT_EXE=cat.exe
SET GREP_EXE=grep.exe
SET M4_EXE=m4.exe
SET SED_EXE=sed.exe

SET AWKSUB1=%AWK_EXE% "{ sub(/@progtitle@/, "%PROG_TITLE%"); sub(/@progname@/, "%PROG_NAME%"); print }"
SET AWKSUB2=%AWK_EXE% "{ sub(/CHANGEDATE/, strftime("%%b %%d, %%Y", systime())); print }"

IF EXIST %DIST_DIR% (
	CALL :DO_FILES %DIST_DIR% %DIST_DIR_BASE%
)

IF EXIST %DEBUG_DIR% (
	call :DO_FILES %DEBUG_DIR% %DEBUG_DIR_BASE%
)

IF EXIST %PROFILE_DIR% (
	call :DO_FILES %PROFILE_DIR% %PROFILE_DIR_BASE%
)

DEL %AA_DIR%\src\doc\commands.txt 2<&1
GOTO :END

:DO_FILES
REM Create the doc directories if they do not exist
mkdir %1\doc > nul 2<&1
mkdir %1\doc\net > nul 2<&1

REM Generate commands.txt from the dedicated server executable
DEL %AA_DIR%\src\doc\commands.txt > nul 2<&1
IF EXIST %1\%SERVER_EXE% %1\%SERVER_EXE% --doc | %SED_EXE% -e "s,^\[0\] ,," | %GREP_EXE% -v "_help$$" | %GREP_EXE% -v "^changed" | %GREP_EXE% -v "^Executable path: " | %GREP_EXE% -v "^TestPath" | %GREP_EXE% -v "^Testing data path: " | %GREP_EXE% -v "^Testing configuration path: " | %GREP_EXE% -v "^Closing socket" | %GREP_EXE% -v "^Locale item" > %AA_DIR%\src\doc\commands.txt
IF NOT EXIST %1\%SERVER_EXE% ECHO Compile %1\%SERVER_EXE% and run this batch file again
IF NOT EXIST %AA_DIR%/src/doc/commands.txt ECHO Sorry, command help generation failed. > %AA_DIR%\src\doc\commands.txt

REM Generate the html files
CALL :DO_FILE %1 bugs
CALL :DO_FILE %1 changelog
CALL :DO_FILE %1 commands
CALL :DO_FILE %1 compile
CALL :DO_FILE %1 config
CALL :DO_FILE %1 faq
CALL :DO_FILE %1 index
CALL :DO_FILE %1 install_linux
CALL :DO_FILE %1 install_macosx
CALL :DO_FILE %1 install_result
CALL :DO_FILE %1 install_windows
CALL :DO_FILE %1 network
CALL :DO_FILE %1 readme_macosx
CALL :DO_FILE %1 todo
CALL :DO_FILE %1 versions
CALL :DO_FILE %1 net\index
CALL :DO_FILE %1 net\lower
CALL :DO_FILE %1 net\middle
CALL :DO_FILE %1 net\upper
GOTO :RETURN

:DO_FILE
%CAT_EXE% %AA_DIR%\html.m4.in %AA_DIR%\src\doc\%2.html.m4 | %AWKSUB1% | %M4_EXE% -I %AA_DIR%\src\doc | %AWKSUB2% > %1\doc\%2.html
GOTO :RETURN

:END
ECHO Done!

:RETURN

[radian]

in detail settings...the help text on dithering only reads........if the color depth of your current.....................and thats it

[wrtlprnft]

Hmm, it reads just fine for me... do you have a screenshot?

[Lucifer]

Ok, I give. I created the 0.3.0 branch from revision 4887, which includes some fixes based on the alpha released.

I think we'll need one more alpha, at least, if I were on the QA, I'd flunk this one over the sound issues. If they haven't been fixed before then, I'll clear up my computer problems after my calc test and get it taken care of so we can send another alpha to QA.

Sound issues, for anybody who's noticed, are causing a crash at pretty much arbitrary times when sdl_mixer fails to initialize audio somehow. I think I'd like to do the proper fix, which is to disable sound if sdl_mixer fails. So take advantage of the chance to pick some low-hanging fruit on the 0.3.0 branch if you'd like. And feel free to tear up the trunk, if you've been waiting to do so for 0.3.0, the release is out of your way now.

[Luke-Jr]

You know, it could be a RC now... =p

[Z-Man]

We don't do RCs without alphas or betas before that.

[Luke-Jr]

We don't do RCs without alphas or betas before that.

Shrug, a beta then... not sure if beta applies to experimental releases, though, since a beta would generally mean feature-complete.
Think of it another way: alphas and betas apply to 0.3, 0.4, 1.0, etc-- only RCs apply to 0.2.8.0, 0.3.0, etc...

[Z-Man]

You're right, no betas. Betas are for longer, in-depth testing. That's not what we're doing here, the release is experimental and highly explosive anyway. But actually, that is not for us to decide, Lucifer grabbed the release manger hat for this one If you think you'll do it that much better, not too many people would object to you doing 0.3.1 later.

[Z-Man]

Bump. I just made it so that the m_active variable of eSoundMixer is only set to true after the whole initialization succeeds; at least for me and my laptop with dodgy sound, that fixed a crash. Sorry, it probably was my fault that it happened in the first place, I missed the variable's significance when rearranging the ::Init function.

Shall I built another alpha this weekend?

[Lucifer]

Well, I'd really like to test it myself, but I can't promise that'll happen. If you want, sure, but there's one more little fix I'd like, which is to see it properly start the music when you enter a game. It doesn't right now, you have to tell it to.

Also, does m_active work for when the user disables sound? It's supposed to... (I don't know that it ever did, but it was part of my intended fix)

[Z-Man]

I think eSoundMixer::Init ignores the user's wish to disable sound There's a switch on the sound quality, and there's no case for SOUND_OFF. Adding that, and exiting in that case, should do the trick. I'll try that, but I have no idea how I could make the game music start.

Another thing I could do is take the by-default-unused music out of the tarball and thus out of the binary distributions. Should I? Nobody complained about the download size, and I don't mind the longer upload time terribly.

[Lucifer]

Well, people discovered the default music plays if you start it by hand. So we should leave it there.

I'm still restoring files, it could be tonight or tomorrow before I can beat on it. So if you want to build it, that's fine (as long as it respects the user's wish to turn off sound). If you want to wait for me to finish, that's fine too. Up to you. It's not a showstopper if the default music doesn't start itself, and since it's still available, it should be left in the distribution.