[khmer] Guiding principles for khmer?

[C. Titus Brown]

On Sat, Nov 15, 2014 at 12:44:44PM -0800, Travis Collier wrote:
> >From a rather outside POV, the entrance point to khmer for most folks will
> probably be the protocols.  Expanding those, including perhaps adding
> smaller recipes, will be important IMO.
> 
> An API / tools distinction like Qingpeng spoke of makes a lot of sense.
> The protocols/recipes is a natural hub of a more tools focused community.
> It is also a real(ish) world testbed and place to find interesting
> problems/questions for folks more focused on the CS / software engineering.
> The forum coordinating the nitty gritty work on the API level would
> constitute another community hub, not entirely separate of course, but
> different.
> 
> Anyways, it might make sense to make a formal distinction between
> khmer-tools (scripts, protocols, ect.) and khmer-lib (API).

Thanks, Travis --

we already have formal distinctions between khmer itself, the recipes,
and the protocols:

github.com/ged-lab/khmer
github.com/ged-lab/khmer-recipes
github.com/ged-lab/khmer-protocols

what's missing is the distinction between the API and the scripts.  This is
partly because the library API is something that, to a first approximation,
only my lab works on; as that matures, maybe it'll become something separate
from the scripts.  Good to keep in mind.

Is the suggestion then to add

"A set of of command line tools for various kinds of data transformations"

to this list:

* a stable research platform for novel CS/bio research on data structures and algorithms, mostly k-mer based;
* a test bed for software engineering practice in science;
* a Python library for working with k-mers and graph structures;
* an exercise in community building in scientific software engineering;
* an exercise in ecosystem participation in scientific software engineering;

?

cheers,
--titus
-- 
C. Titus Brown, ctb at msu.edu