Document our general approach to building software

[adamltyson]

We detail a lot of the technical elements like CI etc, but we haven't documented the "philosophy". I think this would be useful for:

• Guiding new members of the team/collaborators
• General communication with the wider scientific community

I'm thinking things like:

• Building tools informed by the needs of the SWC/GCNU but in a way that is maximally useful to the wider community
• Collaborating with existing tools/groups where possible
• Making tools as general purpose as possible (e.g. species agnostic) to build a bigger community of users/developers
• Using well established tools and frameworks to aid maintenance
• Integrating as far as possible with existing community efforts e.g. napari
• Standardising as much as possible across the team (e.g. choice of tools/frameworks) to aid maintenance
• Prioritising contributions from researchers as well as SWEs (e.g. by prioritising well known languages like Python)

Interested in any more ideas @neuroinformatics-unit/neuroinformatics-all

[niksirbi]

I like this "NIU manifesto" 😄 .

I would may be add:

• prioritising reliance on FOSS tools as much as possible, and contributing upstream where it makes sense (maybe this can be merged with the point about "well established tools and frameworks").
• developing in the open, as in, not only the outputs are open, but also the process. I personally think this is key to our approach.
• Focusing on "software as infrastructure", instead of "single-use software" (may be merged with "Making tools as general as possible")

[lauraporta]

I've some notes on this subject, taken when chatting with @niksirbi some time ago.
These notes are a draft recipe on how to develop an open source project, not exactly a manifesto, but I think they're a good starting point.

Attaching them as the notes are a bit long.
interview to Niko.pdf