Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.debian.maint.python > #17223 > unrolled thread

Weird status of pdm

Started bySantiago Vila <sanvila@debian.org>
First post2025-12-04 11:30 +0100
Last post2025-12-04 23:50 +0100
Articles 3 — 2 participants

Back to article view | Back to linux.debian.maint.python


Contents

  Weird status of pdm Santiago Vila <sanvila@debian.org> - 2025-12-04 11:30 +0100
    Re: Weird status of pdm Soren Stoutner <soren@debian.org> - 2025-12-04 18:00 +0100
      Re: Weird status of pdm Santiago Vila <sanvila@debian.org> - 2025-12-04 23:50 +0100

#17223 — Weird status of pdm

FromSantiago Vila <sanvila@debian.org>
Date2025-12-04 11:30 +0100
SubjectWeird status of pdm
Message-ID<LYeL0-hQ5-21@gated-at.bofh.it>
Hello.

I would like to get some help with "pdm". The package builds fine in salsa ci:

https://salsa.debian.org/python-team/packages/pdm/-/pipelines

but it fails in the buildds:

https://buildd.debian.org/status/package.php?p=pdm

Recently, Graham Inggs told me that many python packages are expected
to FTBFS during the transition (to avoid MBF for issues that will fix
by themselves), so I wonder if this is one of them or not.

I feel that there must be something subtle about this case and I don't
know what it could be.

Thanks.

[toc] | [next] | [standalone]


#17225

FromSoren Stoutner <soren@debian.org>
Date2025-12-04 18:00 +0100
Message-ID<LYkQp-lRT-9@gated-at.bofh.it>
In reply to#17223

[Multipart message — attachments visible in raw view] — view raw

On Thursday, December 4, 2025 3:26:25 AM Mountain Standard Time Santiago Vila 
wrote:
> Hello.
> 
> I would like to get some help with "pdm". The package builds fine in salsa 
ci:
> 
> https://salsa.debian.org/python-team/packages/pdm/-/pipelines
> 
> but it fails in the buildds:
> 
> https://buildd.debian.org/status/package.php?p=pdm

You have a number of tests that are failing on the buildds.

One of the differences between the buildds and Salsa CI is that the buildds do 
not allow any internet access during build, whereas Salsa CI does not enforce 
that.  So, the first place I would recommend looking is to see if the tests 
that are failing require internet access.

If that is the case, you can disable just those tests during the build, while 
enabling them in autopkgtest (which does allow internet access).

If these tests do not require internet access, then you might dig deeper into 
the error message for each test to see why they are failing.

-- 
Soren Stoutner
soren@debian.org

[toc] | [prev] | [next] | [standalone]


#17228

FromSantiago Vila <sanvila@debian.org>
Date2025-12-04 23:50 +0100
Message-ID<LYqj7-pze-7@gated-at.bofh.it>
In reply to#17225
[ trimming Cc list ]

On Thu, Dec 04, 2025 at 09:59:28AM -0700, Soren Stoutner wrote:

> You have a number of tests that are failing on the buildds.
> 
> One of the differences between the buildds and Salsa CI is that the buildds do 
> not allow any internet access during build, whereas Salsa CI does not enforce 
> that.  So, the first place I would recommend looking is to see if the tests 
> that are failing require internet access.
> 
> If that is the case, you can disable just those tests during the build, while 
> enabling them in autopkgtest (which does allow internet access).
> 
> If these tests do not require internet access, then you might dig deeper into 
> the error message for each test to see why they are failing.

Thanks a lot, there is indeed an upstream commit regarding network
access, which I will try next, because it will disable at least some
of the failing tests.

[ However, both the buildd network and Salsa CI recently switched to
  sbuild+unshare and neither of them currently allow network access,
  so it is certainly surprising that they do not behave the same
  in this respect ].

Thanks.

[toc] | [prev] | [standalone]


Back to top | Article view | linux.debian.maint.python


csiph-web