[DIYbio] Re: Request for Feedback: PySplicer 0.2

Random nitpick stuff:

You could release executables which might be easier to run for many bio types.

In my python coding standards, all filenames must match class names.   In my source code control standards, all filenames must be lowercase, words separated by underscore.  In python.org's recommendations, they suggest not using py prefix for anything.

Your self's internal variables should start with underscore.

Your verbose messages should go to stdout and errors to stderr, if insisting on using the console.  Otherwise write both to a file with a timestamp'ed name which you create on startup, which is written as .csv using the csv package.

Likely you should time each run in microseconds.  This has the beneficial side effect of pointing out areas of needed optimization (perhaps in C).

Each file should have your author comment-banner.

I never use "import xyz from abc" because I think it confuses the namespace.  I only use "import xyz" and use full class names where necessary.


Go ahead and include XNA, you know you want to.

You made sure your use of random is in fact random right?  I haven't used random in python but assume you need a seed and the quick glance doesn't show you seeding.

I only use python 3. I didn't notice a comment anywhere in the source referring to required python version.  Python version changes are a pain, I always list compatibility in every source file.



Tangent.
Your constant PGP signage is bad forum practice imho, similar to having an 8-line signature or etc.  Doubt anyone is impersonating you.  If you'd like authentication then repost all your posts to your own site for verification.


 ## Jonathan Cline
## jcline@ieee.org
## Mobile: +1-805-617-0223
########################