I'm not a lawyer, but my general take on images would be that we should not use images we know are not in the public domain or under GPL. The fact a game may no longer be published or the author may not be commercially published doesn't really change the copyright issues - someone still holds the copyright. The fact that we don't make money from crossfire may reduce the incentive for people to take legal action. OTOH, at the same point, we give the impression that all the images are under GPL, and if someone else has done the image, that is not a right we really have to do. That said, taking images that not easily recognized descendents is OK - at some level, when you have small images, things will generally look similar (eg, all the humanoid ones, etc). As far as footprint of unipart images: A lot of this depends on the image. Some multipart images (which is more than just monsters of course), have depth as well as width - to me, the stores fall into this case. They aren't 2 high because the image itself is high - they are 2 high because they are deep. So they should have a 2x2 arch. Other things have height. the demon lord is a good example - it is 3x5 I think. However, a lot of that 5 spaces is height. So it should probably have a footprint that is 3x2 (or maybe 3x3 - representing that for many monsters, the footprint should probably be square if the monster seems tall). Dragons are a bit trickier - they seem to have both depth and height, so the footprint for them would be a bit harder. However, I would say that this fact remains - big monsters should always have a width that matches the image width. It is the arch (and a multipart object) which determines from the server perspective where the monster actually is, which means which spaces spells will effect it, which spaces (if you try to move into it), you will attack it, etc. If you take the ent, and make it so that it is now just a 1x1 image, you'd get odd effect when attacking from the side - it'd seem that you are inside the ent (or it is on top of you) - for height, this can make some sense (you are really behind it), but such drawing does not make sense in terms of the side to side stuff. However, you can still use the single image for this - which means it then saves the effort of splitting or re-combining. Schedule to convert to unipart images? Personal thought is no reason not to start - if you make a new image, might as well just have it be a single image. There is no reason to start converting all the ones currently in the arch directory - it'd probably be a good idea to start slow and see what things seem to break (convert one of the shops for example - see what issues people observe). There is some advantages to single part images - mostly from the artist perspective, but it does mean there are few images in the game, which has some cost savings. I personally wouldn't worry about changing the footprint of monsters breaking maps - if we get into that, then you can almost stop and say lets not do any changes, because it may break a map. If it turns out a map gets broken, someone will notify us, and we can fix it then. But my personal thought is that changing the footprint of monsters to be _smaller_ is not likely to break any maps. It may make some much tougher, as monsters can now go places they couldn't before, but I can't really see that breaking anything - arguably, maps that are defined such that big monsters are trapped and can't go to certain places are already broken if the players can go to those places and kill the monsters. As for window compilation - yes, the server has previously been compiled on windows - I'm not sure last time that was tried. I'm also not sure if all the necessary build bits for anyone but P. Stolarc to compile the gtk client is avaialable (may be on the website - haven't checked) - that should certainly be put into CVS. It'd be nice to have someone volunteer to be the windows release engineer, eg, regularly test compile of the server/client on windows, and also make pre-compiled releases of a windows gtk client. _______________________________________________ crossfire-devel mailing list crossfire-devel at lists.real-time.com https://mailman.real-time.com/mailman/listinfo/crossfire-devel