Am Dienstag, dem 03.05.2022 um 22:59 +0200 schrieb Maxime Devos: > Liliana Marie Prikler schreef op di 03-05-2022 om 22:04 [+0200]: > > > > but the fact that they need to code up their own shell wrappers > > > > to manage multiple profiles is not good optics imo. > > > They don't have to code up shell wrappers or split profiles. > > > They can just use a single profile. > > This sentence signals to me that you didn't think about this mail > > at all. Consider that people may want to split their packages > > across small thematic profiles, but the current implementation of > > Guix hinders them in doing so. > > I did think about this mail. In particular, the parts about faster > union building. So my response was about making profiles faster > without having to figure out how to split profiles -- neither > manually (see suggestion about automatic profile splitting) nor > automatically (see O(n²) vs. O(n lg n) union-build). Note that I only spoke about costs and benefits in terms of runtime, because I assumed that to be what Andrew meant by costs and benefits. Managing multiple profiles declaratively OTOH is itself a benefit that ought not be ignored! Also kn log kn is still bigger than k(n log n).
> Until the previous mail, I have not seen anything about thematic > profiles, so I did not have thematic profiles in mind in my response. > Even then, I'm not sure what these thematic profiles are supposed to > solve that is not working around some underlying problem (e.g. slow > profile building times). Pierre's "Guix Profiles in Practice" is a 2.5 years old blog post. If you can't think of any uses for multiple profiles, you're not the target audience at this point. Cheers
