From: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
To: buildroot@busybox.net
Subject: [Buildroot] [RFC] Removing CLEAN_CMDS and UNINSTALL_(STAGING|TARGET)_CMDS
Date: Sat, 9 Feb 2013 00:26:14 +0100 [thread overview]
Message-ID: <20130209002614.5568c197@skate> (raw)
In-Reply-To: <201302082349.42455.yann.morin.1998@free.fr>
Dear Yann E. MORIN,
On Fri, 8 Feb 2013 23:49:42 +0100, Yann E. MORIN wrote:
> > That said, it is true that the <pkg>-clean target might be useful for
> > some use cases.
>
> What about your pending out-of-tree package build? I see it as a very
> easy way to clean a package: remove the build-dir altogether. Thus, we
> still have the ability to '-clean' a package, whithout a requirement
> on the package to provide any _CMDS for it.
Agreed. If we have a infrastructure-based way of supporting make
<foo>-clean, then I'm all for it. But as you say, it still means that
the <pkg>_CLEAN_CMDS would no longer exist.
> > However, is <pkg>-uninstall really useful, considering
> > that it doesn't take into account the removal of the reverse
> > dependencies?
>
> Not easily doable, so probably error-prone on the user side. However,
> how many issues have been reported on this? I guess the volume was has
> been pretty low, so either users do not use it, or are not aware of it,
> or expect it to break and manually ensure coherency.
>
> For the two first categories, they won't miss it if/when we remove it.
> For the latter category, I doubt they are many, so will probably not
> whine too much if/when we remove it.
The main problem I see with <pkg>_UNINSTALL_..._CMDS is that we have
never been asking for them when reviewing packages. Sometimes new
packages have them, sometimes not. And I haven't seen any of these
commands being updated/fixed in any package, which to me is an
indication that nobody uses them in practice.
> My stance on the subject is:
> - pkg-clean can be achieved in a better, safer, more systematic way
> with your out-of-tree package build, so we can/should remove any
> _CLEAN_CMDS;
> - pkg-uninstall is inherently broken, and there is no easy, sustainable
> way to implement it, so we can/should remove it.
I agree with your summary.
Thanks for this feedback,
Thomas
--
Thomas Petazzoni, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
next prev parent reply other threads:[~2013-02-08 23:26 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-02-08 22:37 [Buildroot] [RFC] Removing CLEAN_CMDS and UNINSTALL_(STAGING|TARGET)_CMDS Thomas Petazzoni
2013-02-08 22:37 ` [Buildroot] [RFC 1/5] packages: remove all the <pkg>_CLEAN_CMDS commands Thomas Petazzoni
2013-02-08 22:38 ` [Buildroot] [RFC 2/5] package: remove support from CLEAN_CMDS Thomas Petazzoni
2013-02-08 22:38 ` [Buildroot] [RFC 3/5] packages: remove all the <pkg>_UNINSTALL_STAGING_CMDS commands Thomas Petazzoni
2013-02-08 22:38 ` [Buildroot] [RFC 4/5] packages: remove all the <pkg>_UNINSTALL_TARGET_CMDS commands Thomas Petazzoni
2013-02-08 22:38 ` [Buildroot] [RFC 5/5] package: remove support for UNINSTALL_(TARGET|STAGING)_CMDS Thomas Petazzoni
2013-02-08 22:49 ` [Buildroot] [RFC] Removing CLEAN_CMDS and UNINSTALL_(STAGING|TARGET)_CMDS Yann E. MORIN
2013-02-08 23:26 ` Thomas Petazzoni [this message]
2013-02-12 6:32 ` Arnout Vandecappelle
2013-02-12 8:38 ` Stephan Hoffmann
2013-02-12 13:25 ` Luca Ceresoli
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
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20130209002614.5568c197@skate \
--to=thomas.petazzoni@free-electrons.com \
--cc=buildroot@busybox.net \
/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.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.