Closed dcroote closed 6 years ago
Hi Derek, Thanks for contributing, it's a good idea.
As of now Travis fails on OSX, any clue why? Also, why are you fixing a doctest in CIGAR codes? That seems unrelated to the flushing?
Hi Fabio, Looks like the OSX builds fail because of a newer version of pip that causes the build to wait for user input:
Solving package specifications: .
Package plan for installation in environment /Users/travis/miniconda:
The following packages will be UPDATED:
pip: 9.0.1-py27h61def0c_3 --> 9.0.1-py27h1567d89_4
Proceed ([y]/n)?
No output has been received in the last 10m0s, this potentially indicates a stalled build or something wrong with the build itself.
I think this can be solved with the following addition to .travis_before_install.sh before conda install pip
:
conda config --set always_yes
The doctest commit was to fix a build error I got on python2.7 due to unsorted dictionary item comparison.
great, I'm rebuilding master to fix those first, then we'll add the flush
Hi Derek,
Now master
builds, thanks to some fixes in here and some other ones. Can you please rebase and double check that travis passes?
Thanks
Fabio
Hi Fabio,
I rebased, but it looks like swig is now causing many of the builds to fail:
Reading package lists... Done
Building dependency tree
Reading state information... Done
Suggested packages:
swig3.0-examples swig3.0-doc
The following NEW packages will be installed:
swig3.0
0 upgraded, 1 newly installed, 0 to remove and 137 not upgraded.
Need to get 979 kB of archives.
After this operation, 4,762 kB of additional disk space will be used.
WARNING: The following packages cannot be authenticated!
swig3.0
E: There were unauthenticated packages and -y was used without --allow-unauthenticated
The command "./.travis_before_install.sh" failed and exited with 100 during .
Your build has been stopped
Looks like the previous error was a temporary issue- I initiated a rebuild by closing and opening the issue and all looks good.
Please keep an eye if this significantly slows down performance, but it looks good. Thank you!
When I run htseq-count on a cluster the progress doesn't get flushed to the logfile until the job completes. For larger bulk RNA-seq samples it would be nice to have the progress updated during job execution, hence the added
flush()
statements.