Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Chris Kees
Thanks Volker, Aron, and Ondrej. I think I understand most of Volker's questions about HashDist a lot better now. One thing I wanted to clarify is that HashDist uses the abstract idea of functional package management from Nix, but HashDist is not built on the Nix package manager or a rewrite/refa

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Ondřej Čertík
On Mon, Jun 16, 2014 at 1:59 PM, Volker Braun wrote: > On Monday, June 16, 2014 8:40:34 PM UTC+1, Ondřej Čertík wrote: >> >> Volker, my understanding is that this would be useful for developing a >> package, to be able to quickly >> run a build, without committing. But for the end user, you always

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Volker Braun
On Monday, June 16, 2014 8:40:34 PM UTC+1, Ondřej Čertík wrote: > > Volker, my understanding is that this would be useful for developing a > package, to be able to quickly > run a build, without committing. But for the end user, you always want > to build from some commit. Yes, I agree. But o

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Ondřej Čertík
On Mon, Jun 16, 2014 at 1:09 PM, Aron Ahmadia wrote: > > On Mon, Jun 16, 2014 at 3:07 PM, Volker Braun wrote: >> >> In particular its not possible to build from the already-existing git >> repo? I don't want to have to specify the version of sage-the-library, I >> just want to build it out of the

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Aron Ahmadia
On Mon, Jun 16, 2014 at 3:07 PM, Volker Braun wrote: > In particular its not possible to build from the already-existing git > repo? I don't want to have to specify the version of sage-the-library, I > just want to build it out of the source tree. The version is the git tree > sha1 if the tree is

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Volker Braun
On Monday, June 16, 2014 7:32:06 PM UTC+1, Aron Ahmadia wrote: > > At this point, you'd need to depend on a commit that contained your git > tree. The full commit would be unpacked in the build, and you could do a > sub-install from there. > In particular its not possible to build from the alr

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Aron Ahmadia
Hi Volker, We have implemented mirror support that does what you want: https://github.com/hashdist/hashdist/blob/master/hashdist/core/source_cache.py#L740 Mirrors are specified in the hashdist config file, not profiles, but I don't think it would be hard to add support for mirrors at the profile-

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Volker Braun
On Monday, June 16, 2014 7:27:54 PM UTC+1, Ondřej Čertík wrote: > > Ah, from a git repository? That's easy, you just use the git hash, see > e.g.: > sources: > - url: /nh/nest/u/ondrej/repos/ginac > key: git:edfa67d26bac695b5ef9911f3cda3ff50232e35a > I don't want to use the SHA1 of the enti

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Ondřej Čertík
On Mon, Jun 16, 2014 at 12:20 PM, Volker Braun wrote: > On Monday, June 16, 2014 7:04:51 PM UTC+1, Ondřej Čertík wrote: >> >> Yes. You just modify the url field to your own mirror. > > > No. I know that I want to download xyz.tar.gz, and I have a list of sage > mirrors ranked by speed, and I want

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Volker Braun
On Monday, June 16, 2014 7:04:51 PM UTC+1, Ondřej Čertík wrote: > > Yes. You just modify the url field to your own mirror. No. I know that I want to download xyz.tar.gz, and I have a list of sage mirrors ranked by speed, and I want to download from the fastest one. > What if I want to install

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Ondřej Čertík
Hi Volker, I was also pointed to this thread. Aron answered pretty much everything, so just a few comments: On Mon, Jun 16, 2014 at 11:49 AM, Volker Braun wrote: > For the record, Sage build a lot slower to build if you build packages one > after the other. > > So hashdist packages can sort of c

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Aron Ahmadia
Hi Volker, Inline-repies this time: On Mon, Jun 16, 2014 at 1:49 PM, Volker Braun wrote: > For the record, Sage build a lot slower to build if you build packages one > after the other. > That's good to know. We're doing our best to get hashdist in shape for generating relocatable binary build

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Volker Braun
For the record, Sage build a lot slower to build if you build packages one after the other. So hashdist packages can sort of change how they are built by inheriting ("extends:"). But can we override / extend: * source download (to use the Sage mirror network) * the hash computation for the sourc

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Aron Ahmadia
On Monday, June 16, 2014 8:44:43 AM UTC-4, Volker Braun wrote: > > Some questions from playing around with hashdist a bit: > > * How do I build packages in parallel? > * I can't build Python on Fedora 20? What are the actual hashdist > dependencies? > > On Monday, June 16, 2014 5:53:36 AM UTC+1,

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Volker Braun
On Monday, June 16, 2014 1:44:43 PM UTC+1, Volker Braun wrote: > > * How do I build packages in parallel? > Just to clarify, the question is how can I build different packages at the same time. Not: run make -j 10 for each package. -- You received this message because you are subscribed to the

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Volker Braun
Some questions from playing around with hashdist a bit: * How do I build packages in parallel? * I can't build Python on Fedora 20? What are the actual hashdist dependencies? On Monday, June 16, 2014 5:53:36 AM UTC+1, Chris Kees wrote: > > * Modular, allowing for easy experimentation with per-p

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-16 Thread Volker Braun
On Monday, June 16, 2014 5:53:36 AM UTC+1, Chris Kees wrote: > > Nix package manager > I'm aware of that. Going the Nix way means a lot of modifications to how packages are built, much of it at odds with how normal distributions package software. I wasn't the only one who thought that would be

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-15 Thread Chris Kees
HI Volker, If you haven't already looked at it, you might be interested in the hashdist project: https://github.com/hashdist/hashdist. I've included a few more comments below. I will be at sage days this week. On Sun, Jun 15, 2014 at 3:59 PM, Volker Braun wrote: > This is a RFC for new pack

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-15 Thread Volker Braun
On Sunday, June 15, 2014 10:42:55 PM UTC+1, François wrote: > So I guess you don't record files installed. Is it at least feasible? > Its definitely something that I want. But in the interest of keeping manageable milestones I'm not going to change the spgk-install scripts for now. Recording t

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-15 Thread François Bissey
On Sun, 15 Jun 2014 14:40:13 Volker Braun wrote: > On Sunday, June 15, 2014 10:14:59 PM UTC+1, Snark wrote: > > (1) what is the relation between your work and Felix Salfelder's > > None. > No surprise if a bit sad. > > which still hasn't been fully included as far as I know? > > Because it doe

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-15 Thread Volker Braun
On Sunday, June 15, 2014 10:14:59 PM UTC+1, Snark wrote: > > (1) what is the relation between your work and Felix Salfelder's None. > which still hasn't been fully included as far as I know? > Because it doesn't work. There is always trac if you want to know the details. (2) does it allow

Re: [sage-devel] RFC: New Build/Packaging System

2014-06-15 Thread Julien Puydt
Hi, Le 15/06/2014 22:59, Volker Braun a écrit : This is a RFC for new packaging system for "sage-the-distribution". I've already talked about this with a few of you at the last sage days, but finally it managed to do something about it. The goal is to be: * Git-aware: use SHA1 hashes instead of