Re: [Caml-list] GODI is shutting down

[Keyan <ml@pulsschlag.net>]

Just to through in my 2 cents. I didn't understand what GODI did, when I first tried it. Therefore, I sticked with my OS package manage for a long time (home-brew on Mac OS X). The reason I now use OPAM, it worked straight out of the box, and I know exactly what is happening. The biggest benefit: there is a clean and easy way to uninstall everything (rm -rf ~/.opam), which is the killer argument for me.

This is not intended to heat up the discussion, but to explain, from a regular user perspective, why I chose opam over godi.

Cheers,
Keyan


On 23 Jul 2013, at 08:49, Gabriel Scherer <gabriel.scherer@gmail.com> wrote:

> On Tue, Jul 23, 2013 at 12:32 AM, Anil Madhavapeddy <anil@recoil.org> wrote:
>> Instead of doing such a heroic solo job (which I'm not being funny about --
>> it really is a lot of work), why not spend the time to open up GODI as well?
>> 
>> Most of the OPAM package descriptions come from *other people*. We just review
>> them, give feedback if necessary, click the merge button and fix errors that
>> creep through.  And even that takes up a surprising amount of time...
> 
> To be fair, I never felt GODI to be a closed system. There are in fact
> a reasonable number that were packaged by someone else than Gerd (see
> http://godi.camlcity.org/godi/packages.html , the "released by" field
> on packages; several were packaged by Virgile Prevosto, for example).
> It's more that the packaging for GODI is a bit more difficult, and
> probably just as importantly as there never was a strong community
> dynamics to contribute to GODI packaging. It is now the norm that
> everyone should write a META file for ocamlfind, and it is becoming
> the norm to package for OPAM, but GODI probably came too early at a
> time where developers expected distributions to do the packaging work
> (which Debian and Fedora still do superbly), and there never was much
> of a community dynamics for GODI packaging.
> 
> And implementation-wise, I have had the pleasure of interacting with
> Gerd a couple of time to discuss possible changes to ocamlfind, and I
> did not get the impression that the project was "closed"; I'm quite
> sure anyone could contribute (or have contributed) to GODI with a
> well-justified, good-looking patch.
> 
> On Tue, Jul 23, 2013 at 12:32 AM, Anil Madhavapeddy <anil@recoil.org> wrote:
>> On 22 Jul 2013, at 22:21, "Gerd Stolpmann" <info@gerd-stolpmann.de> wrote:
>> 
>>> There is also another point that was driving me mad in the past weeks,
>>> namely missing respect from the OPAM guys. Given the fact that OPAM is
>>> only a thin layer around ocamlfind (and guess who wrote it), and given the
>>> fact that GODI was pioneering in many fields, I was expecting nicer
>>> wordings, and less dumb campaigning ("we have 400 packages, and you only
>>> 170"). OPAM is only harvesting what I seeded many years ago.
>> 
>> So when you describe the virtues of GODI, you call it 'advocating', and when
>> Fabrice describes the virtues of OPAM, it's 'dumb campaigning'. This is
>> splendidly reminiscent of US politics!
>> 
>> I don't find any of this source-based package management particularly novel.
>> I've worked in the OpenBSD ports tree since the late 90s, and it's way ahead
>> of either OPAM or GODI. I think where OPAM contributes most over any other
>> source manager are the three points on the front page that I iterated earlier
>> in this thread:
>> 
>> - multiple simultaneous compiler installations.
>> - flexible package constraints.
>> - a Git-friendly development workflow.
>> 
>> The first is very useful, but PATH munging is certainly not novel. The Git
>> workflow is heavily inspired by Homebrew.
>> 
>> The constraint solver has seen a tremendous amount of work though:
>> there's one version built into OPAM (tweaked relentlessly by Thomas with
>> heuristics), and the external CUDF solvers from the Mancoosi project which
>> are magically accurate (but not yet integrated by default until they are
>> packaged on different repositories).
>> 
>> Anyway, I'm not a GODI user, but I think you're doing a disservice to your
>> loyal GODI users if you shut it down in a huff. There are features (notably
>> Windows support) where it'll take OPAM some time to catch up.  I think that
>> the real reason that you're shutting it is nothing to do with "respect", but
>> is good old-fashioned burnt-out:
>> 
>>> Although there are still a lot of GODI users, it is unavoidable to shut GODI down due to lack of supporters, especially package developers. I was more or less alone in the past months, and my time contingent will not allow it to do the upgrade to OCaml 4.01 alone (when it is released).
>> 
>> Instead of doing such a heroic solo job (which I'm not being funny about --
>> it really is a lot of work), why not spend the time to open up GODI as well?
>> 
>> Most of the OPAM package descriptions come from *other people*. We just review
>> them, give feedback if necessary, click the merge button and fix errors that
>> creep through.  And even that takes up a surprising amount of time...
>> 
>> -anil
>> --
>> Caml-list mailing list.  Subscription management and archives:
>> https://sympa.inria.fr/sympa/arc/caml-list
>> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
>> Bug reports: http://caml.inria.fr/bin/caml-bugs
> 
> -- 
> Caml-list mailing list.  Subscription management and archives:
> https://sympa.inria.fr/sympa/arc/caml-list
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> Bug reports: http://caml.inria.fr/bin/caml-bugs