unofficial mirror of guix-devel@gnu.org 
 help / color / mirror / code / Atom feed
From: Andy Patterson <ajpatter@uwaterloo.ca>
To: Pierre Neidhardt <mail@ambrevar.xyz>
Cc: guix-devel <guix-devel@gnu.org>
Subject: Re: [Next browser] Common Lisp: mgl-pax: Package SWANK-BACKEND does not exist.
Date: Tue, 28 Aug 2018 03:36:35 -0400	[thread overview]
Message-ID: <20180828033635.6c07c762@uwaterloo.ca> (raw)
In-Reply-To: <871sakp864.fsf@ambrevar.xyz>

Hi Pierre,

On Mon, 27 Aug 2018 09:42:27 +0200
Pierre Neidhardt <mail@ambrevar.xyz> wrote:

[...]

> One thing I'd like to understand about Common Lisp packages in Guix,
> if you can shed some light on this: the build system uses ASDF to
> "bundle compile" the entire package into a
> single /gnu/store/…-PACKAGE/lib/PACKAGE--system.fasl file.  This file
> is then referenced in /gnu/store…-PACKAGE/lib/PACKAGE.asd. So when a
> Common Lisp package looks for its dependencies, it searches for a
> PACKAGE.asd file in the LIBRARY_PATH environment variable. Is this
> correct?

It's actually using XDG_DATA_DIRS.  In Guix, we set up our lisp
implementations to check for share/{lisp}-bundle-systems for system
definitions. The links there are followed to the lib directory which is
a bit arbitrary, but was already being used for built-in libraries by
sbcl.

> 
> If so, what would be the steps to package a Common Lisp library
> without ASDF? 

I guess you'd have to basically do what asdf does - call the compiler
on the source files with all the right ceremony.

> Is it possible to create a bundle without ASDF?

I'm guessing it is but it's probably implementation specific.

> How
> are multiple .fasl files loaded when put in a folder pointed by
> LIBRARY_PATH?  Do we absolutely need a .asd file?
>

It could be done with output translations but it's super brittle in my
experience.  I think some asd file is required regardless.

> I have skimmed over the ASDF documentation but I am not sure I can
> find answers there.
> 
> > The log listed in the report when using that PR doesn't show the
> > full details - but I've just found out that some warnings are being
> > treated as errors by sbcl.  
> 
> Which log?  If it's one of mine, I can post the full backtrace.  As
> I'm not too familiar (yet) with Common Lisp and SBCL, I might have
> missed important parts of the backtrace.  Let me know.

I was referring to the log you posted at [1]. I'm guessing that there's
a warning somewhere just up above.

> 
> Regarding your additions: Have a look at my wip-next-browser branch,
> in case it's overlapping with your work.  I've borrowed one or two of
> the packages you had sent earlier on this list.  I've also "fixed"
> one thing in the build system (more of a quick & dirty workaround).

Sure thing.

> 
> Cheers!
> 

--
Andy

[1]<https://github.com/slime/slime/issues/457#details-issuecomment-415824038>

  reply	other threads:[~2018-08-28  7:37 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-08-23 16:24 [Next browser] Common Lisp: mgl-pax: Package SWANK-BACKEND does not exist Pierre Neidhardt
2018-08-24 10:32 ` Pierre Neidhardt
2018-08-24 12:56 ` Ludovic Courtès
2018-08-24 13:15   ` Pierre Neidhardt
2018-08-27  7:24   ` Andy Patterson
2018-08-27  7:42     ` Pierre Neidhardt
2018-08-28  7:36       ` Andy Patterson [this message]
2018-08-28 19:18         ` Pierre Neidhardt
2018-08-29  6:18           ` Andy Patterson
2018-09-02 16:59             ` Pierre Neidhardt
2018-09-08 21:49               ` Andy Patterson
2018-09-11 21:01                 ` Pierre Neidhardt
2018-09-11 21:17                   ` Leo Famulari
2018-09-12  8:10                     ` Pierre Neidhardt

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://guix.gnu.org/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180828033635.6c07c762@uwaterloo.ca \
    --to=ajpatter@uwaterloo.ca \
    --cc=guix-devel@gnu.org \
    --cc=mail@ambrevar.xyz \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/guix.git

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).