From: Maxim Cournoyer <maxim.cournoyer@gmail.com>
To: Liliana Marie Prikler <liliana.prikler@ist.tugraz.at>
Cc: 53514@debbugs.gnu.org
Subject: bug#53514: Guix should not set global variables that may affect host
Date: Tue, 25 Jan 2022 18:14:19 -0500 [thread overview]
Message-ID: <87a6fj4ddg.fsf@gmail.com> (raw)
In-Reply-To: <e7aac16cd2f922576d3c8da29bece57aa6b3c353.camel@ist.tugraz.at> (Liliana Marie Prikler's message of "Tue, 25 Jan 2022 08:21:59 +0100")
Hi Liliana,
Liliana Marie Prikler <liliana.prikler@ist.tugraz.at> writes:
> Hi,
>
> Am Montag, dem 24.01.2022 um 17:24 -0500 schrieb Maxim Cournoyer:
>> Hello!
>>
>> There are multiple reports about the negative effects of Guix setting
>> variables such as XDG_DATA_DIRS on foreign distributions, that may
>> cause problems a severe as locking users out of their graphical
>> session [0].
>>
>> In my opinion, we should pursue patching every application/library to
>> use a Guix-specific variant, e.g. GUIX_XDG_DATA_DIRS instead of
>> XDG_DATA_DIRS, to avoid interfering with the host system, as was done
>> for GUIX_PYTHONPATH.
>>
>> This is a big task in itself; we can open more focused/actionable
>> tasks for each environment variable, starting with those causing the
>> most serious issues.
>>
>> Any takers?
> I'm not convinced that patching XDG_DATA_DIRS is a good solution here.
> Even if we go forward and implement this for each and every
> library/application, (it would be reasonably simple to do so for glib
> and qt at least, but there's many more consumers, including Guix
> itself), we'd just force users on foreign distros to set up their
> XDG_DATA_DIRS for us if they e.g. want to have desktop icons available,
> so they'd quickly encounter the same issue on their own.
They may end up doing so, but at least they wouldn't be locked out of
their graphical session just for installing a package from Guix, which
is a clear improvement over the current situation.
> I see two ways forward for this: First, "ignore it" and just document
> the behaviour. This isn't just a bug Guix is suffering from, it also
> affects other third-party package installers like Flatpak and Snap.
>
> Since distros increasingly become aware of them, this will soon no
> longer be an issue for most our users. Second, extend search-paths
> with a way of enforcing a default value when none is set. This way,
> Guix will still override XDG_DATA_DIRS, but since
> $HOME/.local/share:/usr/share is set as the default as per spec, it
> will do what the distro expected.
That sounds good to, and should be easier to achieve.
Thanks,
Maxim
prev parent reply other threads:[~2022-01-25 23:16 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-01-24 22:24 bug#53514: Guix should not set global variables that may affect host Maxim Cournoyer
2022-01-25 7:21 ` Liliana Marie Prikler
2022-01-25 23:14 ` Maxim Cournoyer [this message]
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=87a6fj4ddg.fsf@gmail.com \
--to=maxim.cournoyer@gmail.com \
--cc=53514@debbugs.gnu.org \
--cc=liliana.prikler@ist.tugraz.at \
/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).