[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Package r-cran-rdpack single test fail blocks migration of package r-base



One-line summary: could it be due to the latest upgrade of r-cran-rstudioapi?

Le Thu, Nov 14, 2024 at 01:10:21PM -0600, Dirk Eddelbuettel a écrit :
> 
> Could a group maintainer have a look at package Rdpack (aka r-cran-rdpack) at
>   https://qa.debian.org/excuses.php?package=r-base
> which is being held up by a single test failing on multiple arches, see eg  
>   https://ci.debian.net/packages/r/r-cran-rdpack/testing/arm64/54324186/#S2
> 
> This may be a good time to make contact with upstream as the package is also
> having some minor issues at CRAN
>   https://cloud.r-project.org/web/checks/check_results_Rdpack.html
> so modifying a test to be more tolerant may be advisable.

Hi Dirk and everybody,

I have inspected the diff of the autopkgest runs for testing (2024-11-14
17:07:34 UTC, pass),  r-base/4.4.2-1 migration (2024-11-15 00:10:43 UTC,
fail) and unstable (2024-10-24 04:53:49 UTC, pass).

First, I have to confess that I do not understand why the tests pass
with R 4.4.1 in testing, with R 4.4.2 in unstable, but not with R 4.4.2
in testing.

In particular the diff between the runs in testing with R 4.4.1 (pass)
and R 4.4.2 (fail) have very few differences.

 1) r-base-core (4.4.1-3 for testing, 4.4.2-1 for migration in testing)
 2) libsemanage2 (3.7-2 in testing, 3.7-2+b1, for migration in tesing)
 3) test reports [ FAIL 0 | WARN 12 | SKIP 1 | PASS 41 ]     
    versus [ FAIL 1 | WARN 13 | SKIP 1 | PASS 40 ]                                                                                                     
 4) Failed tests reported in migration to testing:
  ══ Failed tests ════════════════════════════════════════════════════════════════
  ── Failure ('test-bib.R:90:5'): bib works fine ─────────────────────────────────
  `insert_ref("xxx", package = "rbibutils")` did not throw an error.

That's all!

Then I compared the failed migration to testing with the successful run
in unstable.

 1) Apart from r-base-core itself, all r-* packages are at exactly
    the same version number, except for r-cran-rstudioapi which is
    at version 0.17.1-1 in testing and 0.17.0-1 in the CI runs on
    unstable as they were made before Dirk's latest upload.

This may explain why the package's tests also fail on CRAN but did not
on unstable.  I have rescheduled new CI runs in Unstable, let's see if
they fail now.

Have a nice day,

Charles

-- 
Charles Plessy                         Nagahama, Yomitan, Okinawa, Japan
Debian Med packaging team         http://www.debian.org/devel/debian-med
Tooting from home                  https://framapiaf.org/@charles_plessy
- You  do not have  my permission  to use  this email  to train  an AI -


Reply to: