Introducing a Dev branch

[Blastorios]

Introducing a Dev branch

[BradyAJohnston]

While I am in general agreement with what you are suggesting, Molecular Nodes is very much an add-on first and is only experimenting with being a pip installable python package second.

I am in agreement to follow more strictly with not introducing breaking API changes on 'patch' versions, but I would consider and have tried to state that none of the API should be considered stable currently. I have deliberately not published any documentation on the API so as not to give the idea that it could be relied upon. I welcome discussion and suggestions in #359, but like I say in that thread, the API should not be considered stable. The versions are published to pypi to enable some tinkering and testing, but this is very much a secondary consideration.

While the versioning might be a bit misleading, given that we are on v4.0, like you got from the documentation this is to follow the versioning of Blender the priority is ease of use as an add-on.

I do not have the time / development resources to focus on backwards compatibility or different branches, especially to do with API changes that are currently intended only for internal usage and improvements.

I would certainly welcome efforts from anybody who wanted to help with stablising and maintaining the API going forward, and I would like to formalise more official guidelines about what could be considered 'supported' and for how long in the future. The focus for me is to develop for the currently available version of Blender and improve what I can for that.

[Blastorios]

Thanks for the clear explanation! It makes a lot of sense to focus your efforts on the add-on capability first, with the pypi package as a nice follow-up.

I am also looking at #359 in the meantime, as I think MN is a very valuable asset for both the bioinformatics and blender community.