Question: BioConductor Package submission
0
3.5 years ago by
thomas.nigel.lawson0 wrote:

Hi BioConductor community,

I am submitting my first R package into BioConductor and I have come across a problem that I am finding hard to solve. I don't seem to be getting a build report after submitting my package to the tracker

The story so far:

I submitted an issue using the tracker tool. Issue number 1455.

I have 2 packages to submit. A software package and experimental data package I used for unit testing and the vignette. Both packages passed R CMD check and R CMD BiocCheck on my local computer.

The experimental data package was submitted successfully to the same issue with the following message:

Dear Package contributor,

This is the automated single package builder at bioconductor.org.

Your package has been built on Linux, Mac, and Windows.

Congratulations! The package built without errors or warnings
on all platforms.


The software package however was giving an error regarding the vignette. I made a change to the code, tested locally and then reloaded to the tracker but now I do not receive any build report. I have tried reloading a few variations of the software package tar ball but I still get no build report.

Apologies if I am missing something really obvious.

modified 3.5 years ago • written 3.5 years ago by thomas.nigel.lawson0
1

You should ask this question on bioc-devel.

The package seems to hang when building the vignette on linux, I will look into it. Here is a partial build report that includes the other platforms.

http://staging.bioconductor.org:8000/job/463/

1

This is where the vignette seems to hang. I installed the package and Stangled the vignette and then sourced the resulting R file. It seems to hang here and can't be interrupted with ^C:

> xset <- xcmsSet(msmsPths, method = "centWave", ppm=5, peakwidth=c(5,20), snthresh=20, mzdiff=0.001, nSlaves = 3)     3 slaves are spawned successfully. 0 failed.

Are you able to build it yourself on a linux machine or virtual machine?

Also note the segfault on mac in the partial build report linked to above. And the "permission denied" error on windows. Those are often caused by not stopping a socket cluster that has been started.

Thank you. I will look into the clusters and then ask again on bioc-devel if I still have problems.

My local testing environment was on Linux (built and passed fine locally). I will try locally on Windows as well though.

How did you get the partial build report?

EDIT: I realise now that all build reports can be found here http://staging.bioconductor.org:8000/