[Bioperl-l] extra rel. 1.6 suggestion

Chris Fields cjfields at uiuc.edu
Thu Aug 23 14:14:51 UTC 2007


Some interesting points by Sendu:

http://www.bioperl.org/wiki/Release_Schedule#Need_tests

which I agree with completely.

Maybe the best way out if this is a variation on something that was  
suggested before, which was 'splitting' the code into groups.  What  
if we set up a way to automatically gauge test coverage,  
documentation, etc.?  If I remember correctly Nathan had something  
running at one point which did this.

If so, we could determine which code is potentially 'non-compliant'  
and needs to be fixed (tests added, docs brought up to spec, so on),  
and thus prioritize at the minimum what needs to be done for a 1.6  
release.  If it's deemed not worth worrying about (no active  
development, author is out of contact, we have more important  
priorities) we split that code off into a separate 'dev' package.   
That would save some of the headache of trying to split maintenance  
of ~1000 modules up on only a few devs.

Thoughts?

chris



More information about the Bioperl-l mailing list