endian-neutrality guideline

[demonburrito]

Hi, everybody... First time posting to your forum. Super cool project.

I've dug into the code a couple of times and put in nasty hacks to get it to run on my ppc. I am wondering if you guys have a plan or policy on how to abstract these problems away from the people hacking on little-endian machines. And if someone has the time and wants to school me on when/where/how the game uses the filesystem, I'd love to know more about it.

I fear that I may be the only person with a big-endian machine and a working build environment that you have (hopefully not)... I really dig this project, and I'm having a great time seeing the code for a 3d game for the first time. I'm too green with this kind of programming to offer much in the way of solid fixes, but I'll be building nightlies regularly.

[EvilGuru]

Our OS X builds are universal binaries, of which the PPC fork is big-endian. This receives a reasonable amount of attention.

The netcode, is, as of 2.1 endian-safe as should be the save/load game code. To the best of my knowledge there are no endian issues with the current SVN head revision. (Or if there are, they have not been reported.)

If you have encountered any specific issues with the latest SVN revisions, please do report them.

Regards, Freddie.

[demonburrito]

Our OS X builds are universal binaries, of which the PPC fork is big-endian. This receives a reasonable amount of attention.

Umm... Thanks for the reply, but not exactly what I was getting at. Maybe I wasn't clear... I am a programmer, and I wanted to know about where the game interacts with the filesystem, and whether you guys have a strategy for dealing with endianness in the future (so that little-endian people don't accidentally break anything). I've seen progress towards this in the svn, like muggenhor's work on game.c. Doesn't really have anything to do with the universal binaries.

The netcode, is, as of 2.1 endian-safe as should be the save/load game code. To the best of my knowledge there are no endian issues with the current SVN head revision. (Or if there are, they have not been reported.)

Errrr... No. Unfortunately, your trac is broken, but at the top of the active tickets is this: http://developer.wz2100.net/ticket/127. The reason why I wrote this post was buginator's comments. The ticket info is kind of screwed up, but the issue continued past rc1. I can't verify that muggenhor's work a few days ago fixed it as I'm waiting for something that Per is working on (but it looks like it will fix it). Also, I can't search for it now, but I remember there being a wontfix on at least one issue.

If you have encountered any specific issues with the latest SVN revisions, please do report them.

http://developer.wz2100.net/ticket/147

I was just offering to help... One of your contributors was lamenting your lack of ppc developers. So...