unofficial mirror of guix-devel@gnu.org 
 help / color / mirror / code / Atom feed
From: Leo Famulari <leo@famulari.name>
To: Hartmut Goebel <h.goebel@crazy-compilers.com>
Cc: guix-devel@gnu.org
Subject: Re: [PATCH 0/2] Update Python to current versions
Date: Wed, 27 Apr 2016 12:55:57 -0400	[thread overview]
Message-ID: <20160427165557.GA13823@jasmine> (raw)
In-Reply-To: <571F34D5.9060805@crazy-compilers.com>

On Tue, Apr 26, 2016 at 11:28:53AM +0200, Hartmut Goebel wrote:
> Am 24.04.2016 um 02:18 schrieb Leo Famulari:
> > If we can successfully update to 3.5, that would be best. Otherwise, we
> > should consider using 3.4.4.
> 
> I disagree.
> 
> While Python typically is backward-compatible (except for the big change
> between Python 2 and Python 3), some minor incompatibilities are common
> when upgrading from e.g. 3.4 to 3.5. E.g. in 3.5 some deprecated things
> have been removed [1].

Okay.

> I suggest:
> 
> - Rename "python" (python@3.4.x) to "python-3.4"
> - Add a new variable "python" as an alias for "python-3.4"
> - Add "python-3.5" (@3.5.1) (for now independent from "python"

It should be possible to do these 3 actions without causing any packages
to be rebuilt. I think the first 2 steps should be in one commit, and
the second step in a second commit.

> - Update python-3.4 to 3.4.5

This will require every python-3 package to be rebuilt, so it should be
done on a python-updates branch.

That branch could also include python-3.5, updates to setuptools, sphinx
[0], and other core Python packages.

It would also be good to address these reproducibility bugs in Python's
bytecode compiler:
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=22010
http://debbugs.gnu.org/cgi/bugreport.cgi?bug=22533

What does everyone think of this plan?

[0] We need to make Sphinx stop embedding build timestamps:
https://wiki.debian.org/ReproducibleBuilds/TimestampsInManpagesGeneratedBySphinx

  reply	other threads:[~2016-04-27 16:56 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-04-21 20:18 [PATCH 0/2] Update Python to current versions Hartmut Goebel
2016-04-21 20:18 ` [PATCH 1/2] gnu: python-2: Update to 2.7.11 Hartmut Goebel
2016-04-21 21:20   ` Efraim Flashner
2016-04-22  6:42     ` Hartmut Goebel
2016-04-22  6:52       ` Ricardo Wurmus
2016-04-26  9:11         ` Hartmut Goebel
2016-04-27 12:19           ` Ludovic Courtès
2016-04-21 20:18 ` [PATCH 2/2] gnu: python: Update to 3.4.4 Hartmut Goebel
2016-04-21 23:12 ` [PATCH 0/2] Update Python to current versions Cyril Roelandt
2016-04-22  1:57   ` Efraim Flashner
2016-04-22  6:41   ` Hartmut Goebel
2016-04-24  0:18     ` Leo Famulari
2016-04-24 21:43       ` Ludovic Courtès
2016-04-25 19:07         ` Leo Famulari
2016-04-26  9:28       ` Hartmut Goebel
2016-04-27 16:55         ` Leo Famulari [this message]
2016-04-27 18:05           ` Hartmut Goebel

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=20160427165557.GA13823@jasmine \
    --to=leo@famulari.name \
    --cc=guix-devel@gnu.org \
    --cc=h.goebel@crazy-compilers.com \
    /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).