feat: ✨ add initial collaborate on GenomeDK section

[signekb]

Description

This adds some points about how to work together in a shared folder on GenomeDK that @martonvago mentioned when we talked earlier today. This might/probably will be elaborating on as we go, but wanted to add these points as a starting point.

Needs a quick review.

[lwjohnst86]

Nice start!

[lwjohnst86]

Could you include a concrete folder structure here for "shared folder"? Are you working in the exactly the same folder? I made a comment or issue about this I thought, but I can't find it. Do you work like:

onlimit-study/
└── feasibility-data/
    ├── main/
    ├── signe/
    └── marton/

So you have a clone in signe/ and Marton works in marton/? Or do you all work in feasibility-data/? If the second, we should probably work like the above to avoid this "shared editing" situation.

[signekb]

Ah, yes, it's the second situation; there's one feasibility-data/ that everyone works in. With the first, you mean we each have a branch, or that we have n_users + 1 clones of the repo (including the data)?

[lwjohnst86]

We each have clones (with our name for the clone), where we all push and pull from GitHub.

[signekb]

Do we then only have the checked-out data files in the main/ repo? Otherwise, we would have multiple copies of the raw data - one copy for each local repo - right?

[lwjohnst86]

This should be solved by using the build and release process I wrote about. No data is saved outside of the main repo on genomedk

[signekb]

Alright. I'll update this with my understanding of how we'll work 👍

[martonvago]

I had a look at this and wrote a comment.

Just two points here:

• We don't even need to nest the individual clones under a feasibility-data/ folder, everyone can clone where they want. Or is there some benefit to having the clones in a shared folder?
• I think Signe is right about the data duplication. If a clone wants to do something with the data, it has to smudge it, and then there will be duplication. If the clones never smudge, that is the same as working outside of GenomeDK and then I'm not sure what the clones on GenomeDK are for.