unofficial mirror of help-guix@gnu.org 
 help / color / mirror / Atom feed
* gnutls guile
@ 2019-03-12  0:20 ng0
  2019-03-13 17:59 ` ng0
  0 siblings, 1 reply; 5+ messages in thread
From: ng0 @ 2019-03-12  0:20 UTC (permalink / raw)
  To: help-guix

Ludovic,

as a follow-up to the short chat on #guix:

if the problem can not be patched in the guile module of gnutls,
it appears that http://git.savannah.gnu.org/cgit/guile.git/tree/module/system/base/target.scm
could be a source of  the problem?

At the moment I'm trying to patch (or rather sed/awk) in gnutls
possible occurences of the tripplet appearing with the GNU_...
in pkgsrc.

Thanks,
ng0

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: gnutls guile
  2019-03-12  0:20 gnutls guile ng0
@ 2019-03-13 17:59 ` ng0
  2019-05-17  6:54   ` Ricardo Wurmus
  0 siblings, 1 reply; 5+ messages in thread
From: ng0 @ 2019-03-13 17:59 UTC (permalink / raw)
  To: help-guix

ng0@n0.is transcribed 402 bytes:
> Ludovic,
> 
> as a follow-up to the short chat on #guix:
> 
> if the problem can not be patched in the guile module of gnutls,
> it appears that http://git.savannah.gnu.org/cgit/guile.git/tree/module/system/base/target.scm
> could be a source of  the problem?
> 
> At the moment I'm trying to patch (or rather sed/awk) in gnutls
> possible occurences of the tripplet appearing with the GNU_...
> in pkgsrc.

Due to the short lifetime of IRC I'll add some more comments:

For a manual build on NetBSD (.99 being the CURRENT), the tripplet:

x86_64-unknown-netbsd8.99
 
is correct. For a build in pkgsrc (which is not exclusive for
NetBSD but I haven't looked at Linux behavior so far) it becomes

x86_64--netbsd8.99

because of code specified in

http://cvsweb.netbsd.org/bsdweb.cgi/pkgsrc/mk/configure/gnu-configure.mk?rev=1.17&content-type=text/x-cvsweb-markup

This happens to make the builds and filenames more predicatable.
So your initial assumption about NetBSD tripplets was right,
but similar to guix, pkgsrc needs a different treatment here.

Note that this is the last dependency I need to package before
can start packaging guix for pkgsrc, so it would be nice to get a patch
upstream instead of hacking around it.

The original error post:
https://lists.gnutls.org/pipermail/gnutls-help/2019-February/004488.html

I hope this is enough context to get a reply.

> Thanks,
> ng0
> 
> 

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: gnutls guile
  2019-03-13 17:59 ` ng0
@ 2019-05-17  6:54   ` Ricardo Wurmus
  2019-05-17  8:53     ` ng0
  0 siblings, 1 reply; 5+ messages in thread
From: Ricardo Wurmus @ 2019-05-17  6:54 UTC (permalink / raw)
  To: ng0; +Cc: help-guix


Hi ng0,

> The original error post:
> https://lists.gnutls.org/pipermail/gnutls-help/2019-February/004488.html
>
> I hope this is enough context to get a reply.

from the original post I see that you don’t expect that triplet.  From
your last two emails I don’t see why you can’t just patch it out and be
on your merry way.  I also don’t see why Netbsd would change the
triplet.  You wrote “This happens to make the builds and filenames more
predicatable”, but I guess you can just change this, no?

Have you attempted to stop the triplet from being changed?

--
Ricardo

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: gnutls guile
  2019-05-17  6:54   ` Ricardo Wurmus
@ 2019-05-17  8:53     ` ng0
  2019-05-19 14:42       ` Ludovic Courtès
  0 siblings, 1 reply; 5+ messages in thread
From: ng0 @ 2019-05-17  8:53 UTC (permalink / raw)
  To: Ricardo Wurmus; +Cc: help-guix

Hi Ricardo,

thanks for getting back to me.

One detail I'm still not clear about:
Ludovic said the guile bindings are maintained by Ludovic.
But recent activity in the guile parts is done by other people.
Since guix depends on this it's still okay to ask for help here,
but would it be more appropriate to report elsewhere? Or do you
still have some way to contribute changes back to GnuTLS?

Ricardo Wurmus transcribed 627 bytes:
> 
> Hi ng0,
> 
> > The original error post:
> > https://lists.gnutls.org/pipermail/gnutls-help/2019-February/004488.html
> >
> > I hope this is enough context to get a reply.
> 
> from the original post I see that you don’t expect that triplet.  From
> your last two emails I don’t see why you can’t just patch it out and be
> on your merry way.  I also don’t see why Netbsd would change the
> triplet.  You wrote “This happens to make the builds and filenames more
> predicatable”, but I guess you can just change this, no?
> 
> Have you attempted to stop the triplet from being changed?

No, this is done by pkgsrc facility, which targets more than just NetBSD.
Consequentially this would be against the pkgsrc guidelines (and probably
hard to do).
I think ( / hope ) I mentioned in my previous email(s) the location of
the mk file which handles this. The tl;dr is that this is at the core
of pkgsrc.
I can look into wether this is possible, but in my opinion there should
be an exception for pkgsrc: GnuTLS is not part of the base system, so
the only way to get GnuTLS (and its bindings) is via pkgsrc even on NetBSD;
ie pkgsrc is the defacto way to install gnutls with guile bindings.
I will gather more information, but
GSoC is upcoming and it might take some time until I get back to you.

In case I can provide more useful information in addition to asking the right
people wether this is possible, let me know.

Cheers,
ng0
 
> --
> Ricardo
> 

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: gnutls guile
  2019-05-17  8:53     ` ng0
@ 2019-05-19 14:42       ` Ludovic Courtès
  0 siblings, 0 replies; 5+ messages in thread
From: Ludovic Courtès @ 2019-05-19 14:42 UTC (permalink / raw)
  To: Ricardo Wurmus; +Cc: help-guix

Hi ng0,

ng0@n0.is skribis:

> Ludovic said the guile bindings are maintained by Ludovic.
> But recent activity in the guile parts is done by other people.
> Since guix depends on this it's still okay to ask for help here,
> but would it be more appropriate to report elsewhere? Or do you
> still have some way to contribute changes back to GnuTLS?

I think we should focus on fixing what needs to be fixed, and then I
don’t doubt the patches will be applied somehow.

However, in this case, it’s not clear to me that there’s a bug in
GnuTLS; rather, my understanding is that the triplet being passed is not
a valid triplet, leading to breakage down the road.

Did I get this right?

Thanks,
Ludo’.

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2019-05-19 14:42 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-03-12  0:20 gnutls guile ng0
2019-03-13 17:59 ` ng0
2019-05-17  6:54   ` Ricardo Wurmus
2019-05-17  8:53     ` ng0
2019-05-19 14:42       ` Ludovic Courtès

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).