Hi Andreas, Le 03/02/2023 à 13:05, Andreas Tille a écrit :
Hi Pierre,Am Fri, Feb 03, 2023 at 11:32:07AM +0100 schrieb Pierre Gruet:[16:13:15] Found snippy-vcf_to_tab - /usr/bin/snippy-vcf_to_tab [16:13:15] Found snippy-vcf_report - /usr/bin/snippy-vcf_report [16:13:15] Checking version: samtools --version is >= 1.7 - ok, have 1.16 [16:13:15] Checking version: bcftools --version is >= 1.7 - ok, have 1.16 [16:13:15] Checking version: freebayes --version is >= 1.1 - ok, have 1.3.6 [16:13:15] Need snpEff -version >= 4.3 but you have 0 - please upgrade it. autopkgtest [16:13:16]: test run-unit-test: -----------------------] I have no idea what this might mean.In fact it turns out snpeff itself is not installable on i386. I just tried to apt-get install snpeff in an i386 chroot and I gotStrange that this log is that different from s390x where the installation issue is more direct, but in principle the same.-------------------------------8<------------------------------------ Reading package lists... Building dependency tree... Reading state information... Some packages could not be installed. This may mean that you have requested an impossible situation or if you are using the unstable distribution that some required packages have not yet been created or been moved out of Incoming. The following information may help to resolve the situation: The following packages have unmet dependencies: libngs-java : Depends: libngs-jni (>= 3.0.3+dfsg-4) but it is not installable Depends: libngs-jni (< 3.0.3+dfsg-4.1~) but it is not installable E: Unable to correct problems, you have held broken packages. Command apt-get --dry-run install -- snpeff exited with exit code 1. -------------------------------8<------------------------------------ libngs-jni and libngs-java are built from src:sra-sdk, of which binaries are only for amd64 and arm64, see [3]. Thus we could ignore autopkgtest failures on arches other than amd64 and arm64 for snippy -- although I cannot explain right now how its autopkgtests are passing on armel for instance, as libngs-jni is unavailable there.I think with this explanation the conclusion should be diff --git a/debian/tests/control b/debian/tests/control index cd113c6..fd46ed1 100644 --- a/debian/tests/control +++ b/debian/tests/control @@ -1,4 +1,4 @@ Tests: run-unit-test Depends: @ Restrictions: allow-stderr, skip-not-installable -Architecture: !s390x +Architecture: amd64 arm64 which I'll upload soon.
Looks perfectly reasonable, yes. Thanks!
Would you have time today, you can take a decision if it is obvious to you. Else I will look at it during the weekend!IMHO we need to decide whether we should ship snpeff version 5.0 rather than 5.1 to deal with bug #1029202. May be you can estimate the effort needed for this change? I have no idea how many applications besides snippy are suffering from this issue but it does not sound good.
I am frustrated to say this, but possibly the biggest difficulty is... getting the source. Version 5.0 was never tagged in the GitHub repo, it is not available anymore on the website of SnpEff, the Wayback Machine has not registered it :-(
I understand your colleagues are using version 5.0 from conda. I can find the binaries there, but do the people from conda register the sources anywhere?
Bye, -- Pierre
Kind regards Andreas.[3] https://buildd.debian.org/status/package.php?p=sra-sdk
OpenPGP_signature
Description: OpenPGP digital signature
