(Reduced the Cc) > Now to the cons (the prior was the pros): I donnu about Guile Scheme, > heard from former Cygnus ppl (including Ian Taylor IIRC) it was a > mistake.
A language change is not going to happen. I should say that one of the main reasons why I took interest in Guix was Guile. As an Emacs user and Scheme fan I’m obviously biased towards '(list guile elisp racket), but I really think that using Guile makes Guix really attractive: package recipes are just Scheme objects that also contain quoted code to be run at build time. This approach of just using one flexible language makes it very easy to use Guix as a library, as shown by Guix Web, a web interface for Guix, or the Emacs user interface. I’m very glad Guix isn’t built on Python or JavaScript... > Is this why it doesn’t make any buzz or is it because you don’t try > hard enough to contact the papers??? Would be nice if someone signed > up to advocate Guix for distress etc… Guix is a pretty young project. I wouldn’t worry so much about adoption at this point. That said, Guix is becoming a serious contender for enabling reproducible scientific environments in bioinformatics. At the bioinformatics institute where I have my system administrator hat on we’re using Guix on two clusters. It helps that Guix now has packages for a great number of common scientific applications, libraries, and language environments. For bioinformatics people there is a separate mailing list on which Guix is discussed for use in bioinformatics environments: http://lists.open-bio.org/mailman/listinfo/bio-packaging ~~ Ricardo
