Request for comments on interchangeable bioinformatics containers
People
Repo info
Activity
pbelmann
@pbelmann
@michaelbarton could you take a look at my PR with the developer guide?
Paolo Di Tommaso
@pditommaso
Hi @michaelbarton
I was wondering if the yaml input/output definition is going to be required in all kind of containers that will make part of the bioboxes specification
Michael Barton
@michaelbarton
Yes with the current specification. It’s the best way we have come up with for transferring data between the user and container.
Do you have an alternative suggestion?
Paolo Di Tommaso
@pditommaso
that mean that one will have to create a yaml file in order to launch a container
Michael Barton
@michaelbarton
Yes
Paolo Di Tommaso
@pditommaso
it's that quite unhandy, don't you think that?
Michael Barton
@michaelbarton
Yes I do
However it is the only sane way to deal with complicated data we could think of.
Paolo Di Tommaso
@pditommaso
the beauty of Docker is that you can interact quickly with containers with cmd line
Michael Barton
@michaelbarton
We originally started with using environment variables passed to the container, but it quickly became complicated.
Paolo Di Tommaso
@pditommaso
in fact, I was remembering that
why don't you use command line parameters?
Michael Barton
@michaelbarton
because if you are specifying multiple fasta files with different insert sizes it quickly becomes just as difficult.
The issue is #61
It was the longest discussion we had.
I would be happy to read a proposal for a simpler interface, however I would ask that you read that discussion first.
Paolo Di Tommaso
@pditommaso
ok, now that you pointed out the issue I will read that thanks
Michael Barton
@michaelbarton
Thanks. We do welcome input from everyone. I know you have a lot experience from nextflow.io, so I would be glad to read any feedback or suggestions you have.