From: Baurzhan Ismagulov <ibr@radix50.net>
To: isar-users@googlegroups.com
Cc: Henning Schild <henning.schild@siemens.com>
Subject: Re: [PATCH v1 0/6] Migrate to mmdebstrap
Date: Mon, 29 May 2023 16:06:32 +0200 [thread overview]
Message-ID: <ZHSxaEmyWaFuiPV9@ilbers.de> (raw)
In-Reply-To: <20230512215029.69351b38@md1za8fc.ad001.siemens.net>
On 2023-05-12 21:50, 'Henning Schild' via isar-users wrote:
> Maybe you can add a few words on the relevance and the sustainability
> of mmdebstrap and compare it to debootstrap. Is it an official debian
> supported tool, where does upstream use it, do we align with debian
> strategy here?
>
> As far as i remember we came from multistrap and eventually switches to
> debootstrap. And the reasons have been around multistrap being somehow
> not official or sustainable.
Some points that quickly come to mind are:
* mmdebstrap is part of Debian and claims e.g. the following advantages
compared to debootstrap: Direct debootstrapping from multiple repos,
reproducible, unprivileged operation, apt-less chroots (and more in [1]).
* What we'd like to have:
* Introspection ("what will be installed") before actually installing stuff,
necessary for upfront base-apt and Debian dependency support in bitbake.
* Unprivileged bootstrapping, for later unprivileged building in unprivileged
containers.
* Bootstrapping from multiple repos in one pass (installing security and
custom-built packages doesn't require updating and upgrading).
* Go away from shell, preferably towards Python.
* Replace qemu-debootstrap which is deprecated.
mmdebstrap satisfies this except that it is written in Perl.
* Regarding multistrap, the author said he didn't maintain it anymore. Moving
back to Bourne shell, no introspection and two-stage bootstrap + upgrade to
security repo was a regression which I didn't like. Retrospectively, I think
we could have maintained multistrap as well -- there are people in emdebian
community who need it and we've discussed reimplementing it in Python (in the
end, no changes which would require toucing multistrap have been introduced
till today). Given mmdebstrap, we're postponing the reimplementation for an
indefinite period of time -- the same policy as with bitbake which suits our
needs as is to 80 %.
1. https://gitlab.mister-muffin.de/josch/mmdebstrap#the-sales-pitch-in-comparison-to-debootstrap
With kind regards,
Baurzhan
next prev parent reply other threads:[~2023-05-29 14:06 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-09 8:21 Anton Mikanovich
2023-05-09 8:21 ` [PATCH v1 1/6] meta: Switch " Anton Mikanovich
2023-05-09 8:21 ` [PATCH v1 2/6] isar-bootstrap: Move preparations to hooks Anton Mikanovich
2023-05-09 8:21 ` [PATCH v1 3/6] isar-bootstrap: Use tar output instead of directory Anton Mikanovich
2023-05-09 8:21 ` [PATCH v1 4/6] user_manual.md: Replace debootstrap with mmdebstrap Anton Mikanovich
2023-05-10 7:09 ` Jan Kiszka
2023-05-10 7:15 ` Anton Mikanovich
2023-05-18 6:31 ` Anton Mikanovich
2023-05-09 8:21 ` [PATCH v1 5/6] isar-bootstrap: Remove unused code Anton Mikanovich
2023-05-09 8:21 ` [PATCH v1 6/6] CI: Install mmdebstrap Anton Mikanovich
2023-05-12 19:50 ` [PATCH v1 0/6] Migrate to mmdebstrap Henning Schild
2023-05-29 14:06 ` Baurzhan Ismagulov [this message]
2023-05-12 20:07 ` Henning Schild
2023-05-18 6:39 ` Anton Mikanovich
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=ZHSxaEmyWaFuiPV9@ilbers.de \
--to=ibr@radix50.net \
--cc=henning.schild@siemens.com \
--cc=isar-users@googlegroups.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.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox