Malcolm> What's the preferred way of doing this for things like Malcolm> enabling the use of system malloc for a given Malcolm> architecture? Put it in a case statement in configure.{in,ac}. Currently this is not done systematically, I think we should work in the direction of having standards (at least informal ones) for this. I'm really impressed by how well your keyword and complex option macros came out. I think we could do more in that direction to make configure more manageable. Malcolm> XEmacs on PPC64 Linux crashes without system malloc and Malcolm> I'd like to enable it permanently. Is burying this info Malcolm> in a sea of case statements within configure.[in|ac] any Malcolm> better that using s&m files? Yes. We simply can't do without configure, and it's best to have it all in one place done in one style, I think. Since the s&m files are disposable, let's dispose of them. Also, that stuff gets seen regularly by people who work on configure, much of it is now standardized by autoconf (which is nice when it works), and shell scripting is a lot more flexible than C preprocessor for this kind of thing. IMO YMMV, but I will say that Martin Buchholz strongly advocated the autoconf path for years. -- Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your business can "do for" free software.