galaxy-iuc/iuc

Chat with the Intergalactic Utility Commission of Galaxy tool developers

Marius van den Beek

@mvdbeek

I’ve excluded both, qiime is unable to create proper biom files

I have no idea what’s going on there

make takes forever to test

and fills up my hard disk locally

there is a chance we can get qiime to work though, since it passed travis tests prior to containerizing

we just need to figure out what the container is missing

Nicola Soranzo

@nsoranzo

So they are not included in the 18?

Marius van den Beek

@mvdbeek

(and that’s with the extended base already)

yeah

not in the list

qiime is 33 failing tests

Nicola Soranzo

@nsoranzo

OK, still pretty awesome! Thanks for the hard work on this!

Marius van den Beek

@mvdbeek

https://mvdbeek.github.io/iuc-tool-test-results/

the raceid tests just need a higher delta (it’s 15 currently ...)

scpipe seems to have trouble contacting biomart

the ebi tools use an API that seems to have changed

CoverageReport2 I would deprecate, it’s a custom tool by @Takadonet et al

I tried bumping the outdated dependencies, but the perl script just doesn’t run

and for coverage we can use multiqc

valet I tried looking into, the test aleays comes out different and with NA NA NA instead of an actual value

no idea what to do there

and finally fastx_trimmer I also spend quite some time figuring out what’s wrong, it’s just the fastq.gz test that fails

Marius van den Beek

@mvdbeek

also of note — this is with galaxyproject/galaxy, release_20.01 branch and planemo 0.70.0, so nothing from my fork/branch

I did include my open IUC PRs though

Nicola Soranzo

@nsoranzo

Speaking of Planemo, should we start dropping Py2 ? Maybe John won't notice :laughing:

Marius van den Beek

@mvdbeek

Also :tada: for Galaxy, it ran 2750 tests and no Galaxy issue as far as I can tell

if you include the upload jobs I’d guess that comes out to ~ 7500 jobs assuming about 2 input datasets per test

Nicola Soranzo

@nsoranzo

Neat

cshl_fastx_trimmer tests pass for me using Conda deps

So it's something related to the container

Marius van den Beek

@mvdbeek

Given it’s the fastq.gz test I’d assume zlib or sth. like that ?

we did bump it recently, let me find the PR

Nicola Soranzo

@nsoranzo

Or gunzip options

Marius van den Beek

@mvdbeek

bioconda/bioconda-recipes#19385

the IUC PR is here: https://github.com/galaxyproject/tools-iuc/pull/2791/files

Marius van den Beek

@mvdbeek

the weirdest thing about the fastx_trimmer test is that I can mount in the test datsest and run the command manually and all works fine

Nicola Soranzo

@nsoranzo

The tests pass locally for me with planemo test --biocontainers (for fastx_trimmer)

Wolfgang Maier

@wm75

maybe a problem with Galaxy's decompress="true" in some version?

Nicola Soranzo

@nsoranzo

I should specify, with the version on master

Marius van den Beek

@mvdbeek

possible, this is also on python 3.7

Nicola Soranzo

@nsoranzo

I'm on Python 3.6.9 for Planemo and 2.7 for Galaxy 20.01

Marius van den Beek

@mvdbeek

this is with --galaxy_python_version 3.7

Nicola Soranzo

@nsoranzo

Let me try

John Chilton

@jmchilton

I’ll notice eventually 😂

Nicola Soranzo

@nsoranzo

Damn :wink:

Marius van den Beek

@mvdbeek

what? now it passes for me too

wasn’t this the premise of containerisation ? argrgrgrgr :laughing:

Marius van den Beek

@mvdbeek

still failing on travis after triggering a new build: https://travis-ci.org/galaxyproject/tools-iuc/jobs/638930688?utm_medium=notification&utm_source=github_status

Nicola Soranzo

@nsoranzo

I can reproduce locally with planemo test --galaxy_python_version 3.7 --biocontainers on master

Marius van den Beek

@mvdbeek

Trimming: base 1 to 27
Input: 2 reads.
Output: 2 reads.

Where communities thrive

People

Repo info

Activity