Hi Luis, I have been pushing for a very long time to split down PDL, which is accepted in principle (see this message from 2015: https://sourceforge.net/p/pdl/mailman/message/34158032/), but literally any actions to pursue such have each individually not been accepted, due to an understandable desire to minimise risk. I call this a logjam. If one looks at the GitHub "master" branch commits, the commit for the last release is at the bottom of the second page. You can draw your own conclusions from that.
PDLA was conceived as a way to break this logjam. Therefore, while I of course agree with your point, and not coincidentally with myself, PDLA is the only way forward. It's a misfortune that I haven't updated it since the first effort, but I am going to rectify that. Hence this conversation! Best regards, Ed -----Original Message----- From: Luis Mochan Sent: Sunday, April 14, 2019 7:09 PM To: [email protected] Subject: Re: [Pdl-general] Fw: [Pdl-devel] Whither PDL? On Sun, Apr 14, 2019 at 04:30:42PM +0000, Ed . wrote: > Dear pdl-general list, (had to send again, after subscribing - whoops) > > Luis has quite correctly reminded me of this list. I have very recently > revived the question of "whither PDL", and am very interested to get the > thoughts of the wider community. The relevant posting is: > > https://sourceforge.net/p/pdl/mailman/message/36638395/ - some explanation > of detail on this crazy "PDLA" thing I recall that some parts of PDL have been factored out in the past. Would it be better to merge into PDL the kind of changes proposed for PDLA instead of replacing in the future PDL by PDLA? PDLA could become some kind of development branch of PDL. I would be glad to experiment with it. I do think it is a good idea to break-up PDL into smaller parts that are easier to install and that can be installed as they are needed. Currently, the automatic installation of PDL (say running 'cpanm PDL') can succeed though leaving many parts uninstalled due to missing dependencies. My students have suffered believing they had succeded and then trying to use some parts of PDL that were not operational. For similar reasons, I guess, the binary distributions of PDL (at least in debian, apt install pdl) only install a subset of the functionality. I guess this might be alleviated with finer-grained packages that force installation of their dependencies. > > (I've checked, and PDLA does in fact still call its executable "pdl" - as > I've said on the thread, that's a bug and I will fix it) Ok. > > Basically, the question is: do you, dear scientist/PDL user, still use > PDL? > Would the increased reproducibility that is easily within reach add value? As I answered in the other list, in our group we do use regularly. Best regards, Luis > (I emphasise reproducibility because a quick scan of the -general list > showed people wanting to patch PDL "proper", which means anyone trying to > reproduce their results would need to also use the patched version, which > is > a disaster in my opinion) > > Best regards, > Ed > -- o W. Luis Mochán, | tel:(52)(777)329-1734 /<(*) Instituto de Ciencias Físicas, UNAM | fax:(52)(777)317-5388 `>/ /\ Apdo. Postal 48-3, 62251 | (*)/\/ \ Cuernavaca, Morelos, México | [email protected] /\_/\__/ GPG: 791EB9EB, C949 3F81 6D9B 1191 9A16 C2DF 5F0A C52B 791E B9EB _______________________________________________ pdl-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/pdl-general _______________________________________________ pdl-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/pdl-general
