Hi Emmanuel, Emmanuel Bourg wrote:
> Le 23/06/2016 à 14:33, Jochen Wiedmann a écrit : > >> The important part, to me, is to find something on which we can agree. >> That doesn't mean that everyone is happy with the outcome, but that >> everyone's got the feeling "I can live with that". In particular, >> there must not be any serious opposition later on. If you'd like to >> oppose: Please do so here, and now. > > Thank you for proposing a plan Jochen. I'm certainly in the "I can live > with that" category, but I'm a bit skeptical about the outcome. Neither > the incubator nor a TLP sound viable to me at this point, because the > community around commons-math hasn't recovered from the loss of its > historical contributors. I'd would rather incubate the Math TLP within > Commons until more contributors like Eric join the party. > > Because the best way to attract developers is to release something > useful, I'd suggest releasing as soon as possible something derived from > the current Commons Math 4 base: > - commons-random: with the random number generators > - commons-math4: with the core math utilities (fractions, complex, > fastmath, stats, FFT, etc) and leaving out the orphaned parts judged too > specialized like the genetic algorithms. > > The core commons-math4 may contain "unsupported" parts as long as they > are properly covered by unit tests and not too specialized. Hmm. Here I got lost. Do you now try to keep the "unsupported" parts in CM4 or leave them out as proposed two lines above? As long as we keep the name "CM" we make promises to existing users and it is difficult to drop something. One benefit of Gilles smaller components is their specialized target. We as PMC failed to realize that CM had been a dumping ground in the last 10 years (see Dimitri's mail) and that aside the real common math utilities we also collected very specialized stuff that is hardly used by anyone and that should have never been added to *Commons* Math (AFAICS the latest released jar is the threefold size compared to any other Commons component). If we keep "CM" we might end up in the same situation again. - Jörg --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
