Write a manifest for the Initiative

[EiffL]

I'm thinking we can have a statement describing the initiative and the guidelines for collaborators. It should include at least the following points:

• Develop high quality scientific software
• Commitment to openness, both in terms of projects and software
• Positive and welcoming collaborative environment
• Scientific goal is to develop tools for joint forward simulation of various probes, in a differentiable framework for efficient inference

@changhoonhahn what do you think?

[changhoonhahn]

@EiffL I highly support this!

[EiffL]

Sience goal:

• Develop tools for joint forward simulation of various probes, in a differentiable framework for efficient inference

Founding principles:

• Develop high quality scientific software for high quality science
• Openness and transparency, at all levels of software and project development
• Positive and welcoming collaborative environment

Proposed conveners (more converners can be added later on, choosing from active contributors):

• @EiffL
• @changhoonhahn
• @sferraroastro

Convener roles:

• Manage the community: Ensure that the terms of collaboration are respected, facilitate discussions through telecons and open channels, maintain membership status for collaborators.
• Maintain the project roadmap: Propose and define in accordance to the community the project goals and define a roadmap. Gently push contributions towards meeting the community defined project goals, for instance in the initial inception phase of a new project. Maintain list of active projects, and advertise required action items.

Membership:

• Anyone can join, provided that they agree to the terms. Pull Requests on codes are most welcome, but agreeing to the terms is required.

Proposal for terms of collaboration:
1 - Code of conduct:
a - We strive for and require the respectful treatment of all of our colleagues. Therefore, we do not tolerate any kind of bullying or harassment in our community. This is a community dedicated to inclusive behavior, which acknowledges the need, as well as advocates for, the equitable treatment of all members of our community. This includes all axes of human identity related to age, race, ethnicity, nationality, gender, sexuality, physical ability/accessibility. It also includes equal respect and equitable support for early-career participants.
b - If any problem arises, please contact any of the conveners.

2 - Project activity:
a - All projects should be discussed and announced to the project community, early on in the development. When proposing a new project idea, the proposal should be communicated to the conveners, who review it, and officially announce it.
b - Projects should be on public repositories, and discussions about projects should be preferentially on public channels. we strongly encourage members to discuss the project on GitHub and Gitter whenever possible. The motivation for this is to foster an open and collaborative environment.
c - Leaders of announced projects are responsible for their projects, we only ask two things: 1. make the best possible effort to welcome contributions and co-authors. 2. The paper draft needs to receive approval from the "project" to be submitted under a D UP affiliation

3 - Authorship:
1. Authorship shall be divided in three tiers: Major authors, minor authors, and colloration (offer an opt-in opportunity to co-author the work).
2. Authors opting into co-authorship should provide a brief description of their contributions for tier 1 and 2. See example in Mao+18 before appendix that begins, “The contributions from the authors are listed below…”

[sferraroastro]

Looks great to me!

[EiffL]

So, after further thought, this proposal enforces that members should always divulge any projects they work on that are related to the Initiative, and make best effort to open it to other people.

In practice, and in particular at the beginning, this means that if people have pre-established projects they are developing with close collaborators, they can't become members.
Also, if say a professor is a member, and likes, forgets to mention that one of their student is working on a tangentially related project, they would be expelled.

1- That may be too restrictive, people might not be easily convinced to join under these conditions
2- Might be difficult to enforce, someone will have to make a judgement call on whether some previously un-disclosed projects fall under the initiative umbrella.

So, I think we should have a slightly better phrasing concerning project activites, something like:

a. All members are welcome to propose new projects under the Initiative, but these projects should be discussed and announced to the community early on in the development. When proposing a new project idea, the proposal should be communicated to the conveners, who review it, and officially announce it.

This implies that as a member you retain the choice of what project you actually want to cover under the Initiative or not.

But in this case, we can also add a community guideline point like:

Community guidelines:
  - In an effort to foster trust and an open collaborative environment, members are strongly encouraged to communicate about their own related research activities, even if they are not declared as DUI projects. 

Finally, we should have a provision for pre-existing projects, that people may want to move under the DUI, like FlowPM ( ;-) @modichirag ). We don't want to encourage people to do things on their own and only join the DUI once it's published, but at the same time, we want a mechanism to allow people to integrate their work. So, I'm thinking we can have a distinction between "DUI projects and papers" and "endorsed by DUI projects and paper". And what could be considered an endorsed project:

• No major objection from members, in particular not a project that has scooped by surprised other DUI members.
• Relevant to the DUI ecosystem

Thoughts?