This example is broken since commit
99c45877a9 (wrong file name passed to
'local-file'), and its functionality is superseded by the upcoming
'provenance-service-type'.
* gnu/system/examples/vm-image.tmpl (this-file): Remove.
<services>: Remove 'config-file service.
* guix/derivations.scm (derivation-input-fold): New procedure.
(substitution-oracle)[closure]: Rewrite in terms of
'derivation-input-fold'.
* tests/derivations.scm ("derivation-input-fold"): New test.
Am Donnerstag, den 05.12.2019, 12:46 -0600 schrieb Brett Gilio:
> That is why I made the note about eventually replacing the
> installation of that directory to use the system usually used by
> `emacs-build-system` because I figure this would cause less issues in
> the future if for some reason the emacs site-lisp directory structure
> were to change again.
>
> --8<---------------cut here---------------start------------->8---
> ;; TODO: Replace with `#:include' method used by
> ;; emacs-build-system.
> --8<---------------cut here---------------end--------------->8---
>
> But I don't think it is a pressing issue to do it so cleanly.
> Or maybe it is. Idk.
Perhaps you're right, but as far as I can see the Emacs build system
does not yet tell us how to proceed with data. It also has a unique
way of handling documentation, first putting it into site-lisp before
moving it to share/info where it should belong.
> > I still don't feel comfortable with data being at the top of site-
> > lisp,
> > though. Perhaps something similar to emacs-yasnippet-snippets,
> > where
> > you store the data in share/emacs/telega-data would be more
> > appropriate.
>
> I was ACTUALLY going to do it this way originally, but I was really
> uncertain of it so I just sent the patch as is. I like the idea more,
> so
> I am going to revise the patch. Otherwise, once I get this revision
> in
> everything should be good to go. I'll also substitute the functions
> to
> reflect that changed path.
>
> [...]
>
> Alright, carry on! Patch will be available soon.
Looking at your revised patch, I still feel a bit uneasy about putting
data into the load-path. I therefore proprose a revised revised (R^2,
if you will) patch, which I've attached to this message.
Regards,
Leo
From 59918d4a486c453463e6c59487047177f3b54205 Mon Sep 17 00:00:00 2001
From: Brett Gilio <brettg@posteo.net>
Date: Thu, 5 Dec 2019 13:51:14 -0600
Subject: [PATCH 3/3] gnu: emacs-telega: Install telega-data to site-lisp.
* gnu/packages/emacs-xyz.scm (emacs-telega)[arguments]: This commit revises a
previous patch and gives the source etc/ directory a unique installation pth
in the site-lisp directory stucture.
Co-authored-by: Leo Prikler <leo.prikler@student.tugraz.at>
Signed-off-by: Brett Gilio <brettg@posteo.net>
--
Brett M. Gilio
https://git.sr.ht/~brettgilio/
>From 03a2617b29c92dc83828495305d73ec4074a5102 Mon Sep 17 00:00:00 2001
From: Brett Gilio <brettg@posteo.net>
Date: Mon, 2 Dec 2019 15:26:27 -0600
Subject: [PATCH 1/2] gnu: emacs-telega: Use test suites.
* gnu/packages/emacs-xyz.scm (emacs-telega)[arguments]: Adds support for
running the ert and server test suites.
Signed-off-by: Brett Gilio <brettg@posteo.net>
* gnu/packages/compression.scm (ncompress): New variable.
* gnu/packages/patches/compress-fix-softlinks.patch: New file.
* gnu/local.mk (dist_patch_DATA): Add it.
* gnu/local.mk: Include lisp-xyz.scm.
* gnu/packages/lisp-xyz.scm: New file.
* gnu/packages/lisp.scm: Move all lisp libraries to lisp-xyz.scm, uglify-js to
javascript.scm and stumpwm to wm.scm.
* gnu/packages/javascript.scm: Add uglify-js.
* gnu/packages/wm.scm: Add stumpwm.
* gnu/packages/bioinformatics.scm: Find uglify-js in javascript.scm.
* gnu/packages/machine-learning.scm: Depend on lisp-xyz.scm instead of lisp.scm.
* gnu/packages/web.scm: Find uglify-js in javascript.scm.
* gnu/packages/web-browsers.scm: Depend on lisp-xyz.scm instead of lisp.scm.
* guix/build-system/minify.scm (default-uglify-js): Find uglify-js in
javascript module instead of lisp.
