On Tue, Mar 22, 2016 at 11:20:37PM +0300, Alex Kost wrote: > Leo Famulari (2016-03-22 01:52 +0300) wrote: > > > On Mon, Mar 21, 2016 at 06:35:19PM -0400, Leo Famulari wrote: > >> I asked upstream about some warnings during the docs generation, and > >> they pointed out that I needed to add a command line. I've attached a > >> revised patch. > >> > >> Can I get some feedback on it? > > > > Based on feedback on #guix, I've revised it again. Attached... > > > From: Leo Famulari <l...@famulari.name> > > Date: Mon, 14 Mar 2016 19:34:25 -0400 > > Subject: [PATCH 5/6] gnu: Add borg. > > > > * gnu/packages/backup.scm (borg): New variable. > > [...] > > +(define-public borg > > + (package > > + (name "borg") > > + (version "1.0.0") > > + (source (origin > > + (method url-fetch) > > + (uri (pypi-uri "borgbackup" version)) > > + (sha256 > > + (base32 > > + "0wa6cvqs3rni5nwrgagigchcly8a53rxk56z0zn8iaii2cqrw2sh")))) > > + (build-system python-build-system) > > + (arguments > > + `(#:phases > > + (modify-phases %standard-phases > > + (add-after 'unpack 'set-env > > + (lambda* (#:key inputs #:allow-other-keys) > > + (let ((openssl (assoc-ref inputs "openssl")) > > + (lz4 (assoc-ref inputs "lz4"))) > > + (setenv "BORG_OPENSSL_PREFIX" openssl) > > + (setenv "BORG_LZ4_PREFIX" lz4) > > + (setenv "PYTHON_EGG_CACHE" "/tmp") > > + #t))) > > If you keep #t inside 'let', then align it to (setenv ...), or move it > away from 'let': > > (lambda* (#:key inputs #:allow-other-keys) > (let ((openssl (assoc-ref inputs "openssl")) > (lz4 (assoc-ref inputs "lz4"))) > (setenv "BORG_OPENSSL_PREFIX" openssl) > (setenv "BORG_LZ4_PREFIX" lz4) > (setenv "PYTHON_EGG_CACHE" "/tmp")) > #t) > > Note the difference: another ')' after the last (setenv ...). > > > + (add-after 'install 'docs > > Perhaps it is better to call this phase 'install-doc'. It's a usual > name that can be met in other recipes. There is also > 'install-man-pages' phase name, but I prefer 'install-doc'.
Sure, I'll take your advice! > > The rest looks good to me, so if you checked that this program works, I > think you can push this patchset. I've been using this program for a few months without problems (yes, I do restore from time to time). It's being developed very rapidly for something so critical as a backup program, so I think that it's still a bit experimental. I could allude to that in the description. What do you think?
