I'm exploring various Scheme implementations and at the moment am liking Chicken. I don't know the community very well, but it sounds like it's one of the large and active Scheme communities. For the casual bystander: the Lisp community is fractured into Scheme and Common Lisp, and each of these is further fractured into many different "implementations." Maybe the biggest Scheme community is PLT Scheme? I don't know. But Chicken is one of the large ones. Chicken Scheme has a package system. They call their packages "eggs" (they did this before Python did) and there's a lot of them available, at <http://chicken.wiki.br/chicken-projects/egg-index-4.html>. For my own use, I converted all of these I could build (217 of them, I think) to Arch packages. Of course I automated the process, but there was a lot of refining of the automation and further tweaking. Now I've got all of these chicken-* packages I could make available on AUR. I think anyone else using Arch and Chicken would find them very convenient. Chicken has its own package management system, but if you use Arch I think it's nicer to do things through pacman. (In the same way that we like to manage perl modules through pacman.) However, I don't know how much I'm going to be able to commit to maintaining these packages. Some of the packages I certainly won't be using myself. I'm not even sure how long, or how seriously, I'll be using Chicken Scheme. It won't be my main programming environment. But I do like it so far and I want to find a Scheme implementation I can stick with and have handy. Maybe this is it. Still, I don't have much time and I'm not confident I'll be able to give these packages much attention down the road. What should I do? Post them to AUR and maintain them minimally until someone else comes along who can commit more seriously to them? Or just leave this note out on the list in the off chance such another person might come across it? The packages I'd submit all check fine with namcap (and as I've said, I've built and installed them all, though I haven't run the tests on them. I didn't have the time to automate or tweak the building of the test dependencies.) The only exceptions are: there are often some redundant dependencies, for instance if package A depends on B and C, and B also depends on C, namcap wants you just to list B in A's depends. But the way Chicken calculates dependencies made it much easier to automate the PKGBUILD-generation with A depending on both B and C. The other exception is the licenses. A few unclear cases aside, I was always able to list the right license in the PKGBUILD. But there are a lot of BSD and MIT licenses in the pack. What I'm supposed to do then is say license=(BSD) and install a copy of the license in /usr/share/licenses/$pkgname. But the downloads from the Chicken repository don't provide separate license files. I'd have to extract the license from comments in the source code, and I haven't done that for these 217 packages. Other than that I think the packages are all fine. My automatic script generates all but 10 or so of them, those last 10 it's easiest to tweak by hand. -- Jim Pryor profjim@jimpryor.net