On 22/05/2014 14:39, David Colling wrote:
> Just a reminder that we have a meeting tomorrow and one of the things
> that will appear on the agenda is the attached document which is a draft
> response to the meeting on HeP software at CERN
This sounds very sensible. Someone suggested somewhere that writing code
and publishing it even counts as a publication, and could be referenced
with a DOI or something. If in the Glorious Future(tm) all data
associated with a publication is published, so should be the codes used
to analyse it, so people can verify the results. It was even suggested
that software be peer reviewed.
Apologies for being vague in the attributions; I genuinely cannot
remember where the idea came from - I don't think it was one of my own.
One other thing to note is the choice of language. I have sort of taken
the afternoon off to do some Real Work(tm), and am trying again to build
the EGI cloud CDMI server. It is written in python, with all the
disadvantages of python - not portable between versions, needs to fetch
stuff through the firewall during build, remote repositories having the
wrong urls, etc. Grr.
But you might have some prototype code written in (say) Ruby, or some
other proper language in which the developer can prototype rapidly but
the language is not widely understood in the community, or the
performance is not sufficient for production runs, so the code is later
replaced with high performance code written in (say) Fortran, and IMHO
it would be like citing an earlier publication if you reimplement
something. The prototype should get credit too, even if it doesn't run
in production.
-j
--
Scanned by iCritical.
|