yquake2remaster/CONTRIBUTE
2012-08-22 17:30:17 +02:00

149 lines
7 KiB
Text

* ****************************** *
* Yamagi Quake II *
* http://www.yamagi.org/quake2 *
* http://github.com/yquake2 *
* ****************************** *
TODO List
===============================================================================
This is a list of features that are currently not implemented in Yamagi Quake
II, but would be nice to have. If you have some freetime to spare, fun coding or
just want to contribute something, this list has some suggestions. All tasks
need strong C knowledge, the necessary amount of understandig of the Quake II
source code or external ABIs differs between the tasks.
Some hints:
- Sign up for a Github account and fork our yquake2 repository. This allows
the easy integration of upstream changes into your branch and sending of
pull requests. You'll get a wiki and a bugtracker for free.
- To contribute your changes back into the main project send pull requests
via Github. It's much easier to review and merge pull requests than patches.
Please send only pull reqeuests from a distinct branch at not from your
"master" branch!
- Quake II has a very fragile and broken codebase. Even after years of cleanup
it's still a disaster. Therefore:
- Do only one change at a time!
- Test after each change (play at least through base1.bsp)
- Commit early and commit often to create a fine grained history.
This helps "git bisect" to find bugs and errors.
- Do not try to clean up things or even rewrite code that you do not
understand to 110%! Even small behavioral changes can introduce
gameplay changes and trigger new bugs! Especially everything that
depends on map data (e.g. path finding or collision detection) is
very likely to break in interesting ways!
- Do not add new dependencies. If you must add a new one contact the Yamagi
Quake II developers prior to it! Everything that adds dependencies should
be hided behint preprocessor macros.
- If your changes change the gameplay experience, make them optional by
introducing a new cvar.
- Linux is not the only operating system out there. All changes should be
portable to other platform (writing pure ANSI-C or C99 is recommended but
not always applicable).
- x86 ist not the only CPU architecture. All changes should be done in pure
C (e.g. no inline assembler) and in an endianess independed way.
- gcc is not the only compiler. Test your changes with clang.
===============================================================================
1. Port Yamagi Quake II to Mac OS X
Difficulty: Medium
Knowledge: Mac OS X API, good knowledge of src/backends/unix/
A port to Mac OS X is a frequently requested feature but can't be done
by the Yamagi Quake II developers due to the lack of hardware. Since Mac
OS X is just another unixoid platform porting the game should be easy,
the most notable difference to FreeBSD and Linux is the MACH-O binary
format. So at least all calls to dlopen(), dlsym() and dlclose() must be
replaced by the corresponding Mac OS X functions. All other platform
dependend stuff should be hidden by SDL, but there may be some SDL bugs
on OS X.
===============================================================================
2. Port Yamagi Quake II to new unixoid platforms (for example DragonflyBSD,
NetBSD, Solaris, etc.)
Difficulty: Medium
Knowledge: Good knowledge of the target platform
Yamagi Quake II runs fine on Linux and FreeBSD. Due to it's very low hardware
requirements it's an ideal game for platforms without good 3D acceleration.
Ports to new unixoid operating systems should be easy. In most cases only some
#ifdef need to be added and the Makefile integration written.
===============================================================================
3. Source code cleanup
Difficulty: Hard
Knowledge: Good knowledge of the Yamagi Quake II source
While the Yamagi Quake II source code was cleaned a few times there is still
much left to do. Someone need to go through the source, read and audit it. Dead
code should be removed, inefficient functions rewritten and bugs resolved. This
can be done at one module (e.g. client, server, refresher, game, etc) at a time.
===============================================================================
4. Addon cleanup
Difficulty: Medium
Knowledge: How to debug hard to read code
While the baseq2 game modul was cleaned up and made more robust by adding
hundreds of sanity checks, both addons (xatrix and rogue) still require a lot
of work. Go through the code, read it and understand it. Remove bugs, and add
the missing sanity checks. Test (e.g. play) and add implement map quirks.
Especially the coop-support of both addons is still fragile at best.
===============================================================================
5. Finish the port of Zaero
Difficulty: Hard
Knowledge: How to work with broken code
Zaero is an unofficial but popular addon to Quake II. It was release as
freeware. The Yamagi Quake II developers did an inital port, but it's
unfinished and still buggy. Grab the source (take a look at our Github
organization), clean it and debug it. Zaero will need some extensive
testing, have fun while playing. :)
===============================================================================
6. Port Quake II to various consoles (especially the Nintendo Wii)
Difficulty: Hard
Knowledge: How to port games to the console, Knowledge of the platform
dependent parts of the Yamagi Quake II source
Game consoles are popular and having Quake II on them is a often
requested feature. Porting Yamagi Quake II to the XBOX 360 isn't
necessary, since an official port (by id Software / Bethesda) was
included in Quake IV for the XBOX 360. A port to the Playstation 3
would be nice but most likely it's impossible due to the DRM. But
a port to the Wii should be feasible and support for the several
open source or alternative consoles straight forward.
===============================================================================
7. Add head tracking to Quake II
Difficulty: Hard
Knowledge: How to design APIs and integrate new paradigms into an old codebase
Head tracking is becoming more and more popular. After the hype of early,
homegrown solutions based on the Nintendo Wiimote and similar techniques
professional build head tracking helmets start to emerge. One example is
the community funded "Oculus Rift" helmet build by Oculus and supported by
John Carmack, Cliff Bleszinski, Gabe Newell and other famous game industry
staff. The goal of this project is to implement Head Tracking into Yamagi
Quake II. Due to the great number of competing head tracking solutions a
generic frontend needs to be designed and integrated into the client. This
frontend should be supported by a device dependent backend. For testing
purposes at least one backend must be implemented.
===============================================================================