Re: Bug#949658: Different tests for r-cran-recipes - "wrong" test prevents migration of r-cran-testthat
Hi Andreas
On Sun, 26 Jan 2020 at 10:27, Andreas Tille <tille@debian.org> wrote:
> I fail to see how this will finally help since the strange regression
> of
>
> autopkgtest for r-cran-recipes/0.1.9+dfsg-2: amd64: Regression ♻ , arm64: Test in progress
>
> for r-cran-testthat seems to remain. So r-cran-testthat will not
> migrate and the specified versioned Depends will not be fullfilled.
I see r-cran-testthat migrated yesterday [1].
So whichever package that was causing r-cran-recipes to pass in
unstable, has now migrated to testing, allowing it to pass there.
Looking at britney's migrations jobs for r-cran-recipes [2], the first
passing test with r-cran-testthat/2.3.1-1 occurred at 2020-01-30
06:43:35 UTC.
The debci log for that test [3], shows the following were among the
changes in test packages since the last test run:
@@ -136,7 +136,7 @@
r-cran-rcppeigen 0.3.3.7.0-1
r-cran-rcpproll 0.3.0-1
r-cran-recipes 0.1.9+dfsg-2
-r-cran-rlang 0.4.1-1
+r-cran-rlang 0.4.3-1
r-cran-robustbase 0.93-5-2
r-cran-rpart 4.1-15-2
r-cran-rprojroot 1.3-2-2
@@ -150,7 +150,7 @@
r-cran-survival 3.1-8-1
r-cran-testthat 2.3.1-1
r-cran-tibble 2.1.3-2
-r-cran-tidyr 1.0.0-1
+r-cran-tidyr 1.0.2-1
r-cran-tidyselect 0.2.5-1
r-cran-timedate 3043.102-1+b1
r-cran-utf8 1.1.4-1+b1
So I suspect either r-cran-rlang or r-cran-tidyr was the culprit here.
I think r-cran-recipes missed a versioned dependency or test
dependency, or r-cran-testthat missed a breaks.
Regards
Graham
[1] https://alioth-lists.debian.net/pipermail/r-pkg-team/2020-January/009852.html
[2] https://ci.debian.net/user/britney/jobs?package=r-cran-recipes&suite[]=testing&arch[]=amd64
[3] https://ci.debian.net/data/packages/testing/amd64/r/r-cran-recipes/4145364.log
Reply to: