From: "Tolga Hoşgör" <tlghosgor@gmail.com>
To: claudius.heine.ext@siemens.com
Cc: jan.kiszka@siemens.com, isar-users@googlegroups.com
Subject: Re: [PATCH] isar-bootstrap: debootstrap --include flag to support https:// sources
Date: Wed, 5 Dec 2018 13:28:22 +0300 [thread overview]
Message-ID: <CAEK19EnJf1B_H=crmwqm+faZsTqdSENXk1Ed0+hsBUz-NJotNQ@mail.gmail.com> (raw)
In-Reply-To: <45323f2d-7dac-9b9d-6382-013b9ae1242c@siemens.com>
On Tue, Dec 4, 2018 at 5:46 PM Claudius Heine
<claudius.heine.ext@siemens.com> wrote:
>
> Hi Jan,
>
> On 04/12/2018 15.28, Jan Kiszka wrote:
> > On 04.12.18 15:23, [ext] Claudius Heine wrote:
> >>> @@ -138,6 +154,12 @@ def get_distro_components_argument(d, is_host):
> >>> else:
> >>> return ""
> >>> +def get_debootstrap_includes(d, is_host=False):
> >>> + if get_distro_have_https_source(d, is_host):
> >>> + return "locales,apt-transport-https,ca-certificates"
> >>
> >> Those package names should be configured in the distro config.
I partly agree. It makes sense that it is the responsibility of distro
config but "debootstrap" does this automatically [1]. This
intervention by "debootstrap" creates some erratic behavior.
Scenario 1:
- No special bootstrap config on distro.
- Distro sources have an "https" and "http" in that order.
Scenario 2:
- No special bootstrap config on distro.
- Distro sources have an "http" and "https" in that order.
While we were happily building on scenario 1, with a minor change now
the build magically fails on scenario 2.
An acceptable solution for consistency would be to disable this
behavior [1] but "debootstrap" does not seem to have an option to
disable this magic-package-adding behavior. How about explicitly
`--exclude` the packages for consistency?
"debootstrap" also did this in its distro-specific scripts [1], like
your approach. But now it is on a common script "debian-common" [2].
So now they think this is more or less a standard package I guess.
Actually what is needed is something the output of:
"DEBOOTSTRAP_DIR=/usr/share/debootstrap source
/usr/share/debootstrap/scripts/<distro>; MIRRORS=<mirrors>
work_out_debs && echo $base" but this is not publicly exposed in any
way. Could call `debootstrap --print-debs <mirror>` for each mirror
and sort/uniq the result but it would be ugly.
> >>
> >
> > Means you would suggest to introduce two variables, e.g.
> > DEBOOTSTRAP_INCLUDE and DEBOOTSTRAP_INCLUDE_HTTPS to encode those
> > packages? Those could then be set in distro confs and simply used here.
>
Now I am completely lost. I suppose there is no need to create the
logic to to differentiate HTTP and HTTPS in ISAR if distro conf will
set the packages. Distro configuration knows whether if it is using
HTTP or HTTPS (or in theory something else) and it should be enough to
define/extend _one_ variable accordingly.
> Not sure about those names, but in principle yes. Package names are
How about BOOTSTRAP_PACKAGES as the variable name?
> distro specific and should not be part of isar-bootstrap IMO.
> Not sure why the 'locales' package made it through review, but that
> might just be because its just one package that is probably named like
> this in every distribution, past and future.
>
> Claudius
>
> --
> DENX Software Engineering GmbH, Managing Director: Wolfgang Denk
> HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
> Phone: (+49)-8142-66989-54 Fax: (+49)-8142-66989-80 Email: ch@denx.de
Do you agree on adding `--exclude=apt-transport-https,ca-certificates`
and introducing a `BOOTSTRAP_PACKAGES` variable to be configured in
distro?
[1] https://salsa.debian.org/installer-team/debootstrap/blob/08cfced48bdf7bc1240efb63f69cde7ed385038b/scripts/sid#L36-40
[2] https://salsa.debian.org/installer-team/debootstrap/blob/4a4e7186de7237c58eae4f0b2de3ad670ad7436f/scripts/debian-common#L30-34
next prev parent reply other threads:[~2018-12-05 10:28 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-12-04 13:35 Hosgor, Tolga (CT RDA DS EU TR MTS)
2018-12-04 14:11 ` Jan Kiszka
2018-12-04 14:23 ` Claudius Heine
2018-12-04 14:28 ` Jan Kiszka
2018-12-04 14:46 ` Claudius Heine
2018-12-05 10:28 ` Tolga Hoşgör [this message]
2018-12-05 12:27 ` Claudius Heine
2018-12-05 12:55 ` Claudius Heine
2018-12-05 12:58 ` Jan Kiszka
2018-12-05 13:04 ` Claudius Heine
2018-12-05 17:09 ` Henning Schild
2018-12-05 17:12 ` Henning Schild
2018-12-11 7:17 ` Henning Schild
2018-12-11 17:08 ` Tolga Hoşgör
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='CAEK19EnJf1B_H=crmwqm+faZsTqdSENXk1Ed0+hsBUz-NJotNQ@mail.gmail.com' \
--to=tlghosgor@gmail.com \
--cc=claudius.heine.ext@siemens.com \
--cc=isar-users@googlegroups.com \
--cc=jan.kiszka@siemens.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