Re: [PATCH v3] meta: drop debian stretch support

[Uladzimir Bely <ubely@ilbers.de>]

In the email from Tuesday, 3 January 2023 01:03:02 +03 user Henning Schild 
wrote:
> On Mon, 2 Jan 2023 16:56:24 +0100
> 
> Henning Schild <henning.schild@siemens.com> wrote:
> > This one maybe should have waited a little longer for review, it fell
> > right into the time between christmas and new year.
> > 
> > A changelog to v1 would have been nice, or a reply to my review
> > comments i gave.
> > 
> > Am Tue, 27 Dec 2022 09:35:55 +0300
> > 
> > schrieb Anton Mikanovich <amikan@ilbers.de>:
> > > Debian Stretch state was moved to EOL since July 2020, so drop it.
> > > This will allow us to get rid of many hacks inside Isar core and
> > > single example packages.
> > > Also drop raspbian-stretch distro and rpi machine used it.
> > > Drop wic_fakeroot fsck hack because it is no more needed.
> > > 
> > > Update documentation to remove Debian Stretch there also.
> > > 
> > > Signed-off-by: Anton Mikanovich <amikan@ilbers.de>
> > > ---
> > > 
> > >  README.md                                     |  2 +-
> > >  doc/offline.md                                |  4 +-
> > >  doc/technical_overview.md                     |  6 +--
> > >  doc/user_manual.md                            | 48
> > > 
> > > +++++++------------ meta-isar/conf/distro/raspbian-stretch.conf   |
> > > 17 ------- meta-isar/conf/distro/raspbian-stretch.list   |  2 -
> > > 
> > >  meta-isar/conf/local.conf.sample              |  7 ---
> > >  meta-isar/conf/machine/rpi.conf               |  8 ----
> > >  .../multiconfig/container-amd64-stretch.conf  |  7 ---
> > >  .../conf/multiconfig/qemuamd64-stretch.conf   |  5 --
> > >  .../conf/multiconfig/qemuarm-stretch.conf     |  5 --
> > >  .../conf/multiconfig/qemuarm64-stretch.conf   |  6 ---
> > >  .../conf/multiconfig/qemui386-stretch.conf    |  5 --
> > >  .../conf/multiconfig/qemumipsel-stretch.conf  |  5 --
> > >  meta-isar/conf/multiconfig/rpi-stretch.conf   | 47
> > > 
> > > ------------------ meta/classes/buildchroot.bbclass              |
> > > 3 +- meta/classes/dpkg-gbp.bbclass                 |  7 +--
> > > 
> > >  meta/classes/image.bbclass                    |  2 -
> > >  meta/classes/sbuild.bbclass                   |  3 +-
> > >  meta/conf/distro/debian-stretch.conf          | 17 -------
> > >  meta/conf/distro/debian-stretch.list          |  8 ----
> > >  .../files/isar-apt-fallback.conf              |  4 --
> > >  .../isar-bootstrap/isar-bootstrap.inc         | 18 +------
> > >  meta/recipes-kernel/linux/linux-custom.inc    |  1 -
> > >  scripts/start_vm                              |  2 +-
> > >  scripts/wic_fakeroot                          |  8 ----
> > 
> > I think this complete file can be removed. Not sure how i said it in
> > my review to v1. Right now this wrapper does nothing but wrap. All the
> > work it used to do has been deprecated and removed.
> > 
> > I suggest to throw a patch on top to remove the file and set
> > 
> > FAKEROOTCMD ?= "true;"
> > 
> > like the removed comment suggested.
> 
> I tried dropping the now seemingly useless wrapper but that somehow did
> not work like it used to. So maybe we need to keep it after all.
> 
> But it would be nice to know why we need to have it in place and have
> that written down in a comment and a commit message.
> 
> regards,
> Henning
> 

Hello.

I've experimented with this FAKEROOTCMD a bit and it seems we really still 
need it.

It really almost dummy except removing first argument from command line passed 
(`args.pop(0)`) and it seems to be the main thing we want from it (removing 
"pseudo" or "fakeroot"?).

Simple replacing the script call with `true` command doesn't work as well as 
attempt to use this logic without `pop(0)`.


-- 
Uladzimir Bely