Biopython2

Ideas for Biopython 2

This page is mostly a collection of ideas from discussions on the mailling list regarding a possible version 2 of Biopython.

If I missed something, please accept my apologies and change accordingly.

I am trying to keep a neutral tone, but sometime we might want to include major arguments in favor and against some option so that readers are made aware of the gist of main arguments

Dependencies

• Depend on matplotlib, numpy and scipy or
• Create a list of acceptable dependencies (e.g. requests or pandas) that developers could use

Documentation

• There seems to be a consensus that API documentation takes precedence
• numpydoc and Sphinx proposed
• Juptyer notebook vs HTML proposed for tutorials.
• Latex to be discontinued?

Deprecate Python 2?

Ongoing discussion

Which Python 3 version?

• All of them? (no one defended this)
• 3.5
• Most recent version when we start developing Bioython 2 (e.g. now would mean 3.6)

Lower case package and module naming

Lower case for packages and modules seems consensual.

Top-level policies

• Name this biopy.? biopython.? biop.*?
• How many modules on the top-level? None or some basic stuff (life exceptions, abstract file management)
• Automatic import of everything, or just partial? or even nothing?

Package organization

Which sub-packages? Tightly related to the modular approach (see below)

Anyone wants to write a bit here about options?

Module system

• Should there be a module system with core modules allowing third parties to to make extensions, a la Biogems
• Bow’s work might be a starting point

Removal of outdated sub-packages

Discuss which packages to remove

Relationship between Biopython 1 and 2

• Is there any code sharing? If so how? Code sharing should jeopardize desired Bioypthon 2 features and architecture
• APIs would probably not be compatible

Feasability

Is there even time to do this?