Gentoo Archives: gentoo-lisp

From: Leonardo Valeri Manera <l.valerimanera@×××××.com>
To: gentoo-lisp@l.g.o
Subject: [gentoo-lisp] [announce] Chicken Egg ebuilds
Date: Mon, 18 Feb 2008 16:04:39
I've completed preliminary work on an eclass and a few demo ebuilds
for installing chicken-scheme's eggs via package managers, and will be
pushing them to the gentoo-lisp overlay in a few minutes.

Testing would be greatly appreciated.

The packages live dev-chicken, if you don't like that blame hkBst, he
gave the OK ;)

The eclass automates pretty much everything regarding compilation and
testing, but it /is/ my first eclass, so please be kind; it does work
correctly with non-trivial eggs though.

Currently there are 3 optional variables that should be set before
inheriting the 'eggs' eclass:

OLD_EGGPAGE - any non-nil value will make it generate a HOMEPAGE
variable for eggs with docs haven't been ported to the wiki.
EGG_TESTABLE - enables test phase... I'm not sure if this is
necessary, but chicken-setup does make some noise if one tried to run
tests on an egg with no test phase.
NEED_CHICKEN - This might not be necessary /ever/, but I took a page
out of the elisp eclass here and decided to put it in 'just in case'.
The effect is the same.

The eclass defines the DEPEND/RDEPEND to dev-scheme/chicken, the
SRC_URI for the egg, the HOMEPAGE line, and takes care of all

Documentation is currently installed to
/usr/share/doc/chicken-eggs/${PN}... the development is fast enough
that using he version as well would mean breaking people's bookmarks
to their own filesystems every week, but if that goes against some
unbreakable rule, its easy enough to change... it would make usage of
the html docs fairly inconvenient however.

Left to the ebuild - because they would require reading a scheme file
in the egg subversion repository, or doing the same in ${S} after the
egg is unpacked - are any dependencies to other eggs, as well as
description and license.

Also, obviously, any dependencies to system packages.

Keywords should in theory be the same as those of the main chicken
package, but I'm only keywording them for x86 unstable since that's
all I can claim to have tested upon.

gentoo-lisp@l.g.o mailing list