aidanbudd / bosc2015 Goto Github PK

OK now with the list of features?

It might be too long, but I struggle to shorten it by merging/removing redundancy - some features are similar/related, but still I find valuably distinct, and don't want to loose that.

I like that several features are software-specific, i.e. it's not just a list of features of great communities in general.

Only a few of us contributed so far - perhaps if we move now to start collecting specific examples/anecdotes/evidence from specific projects, to support these features as important, more people can more easily contribute?

Let me know which of these you think's the best way to move forward:

1. discuss (by mail? here in github's issues?) and agree together changes to current set of features
2. move on now to collecting examples from specific OSBC's we've been involved in, and reaching out/inviting others to contribute

Just dropping a idea here about estimating and talking about the bus factor in OS projects:
https://en.wikipedia.org/wiki/Bus_factor

Data Carpentry seems to be doing a bangup job of developing bioinformatics training resources --

https://github.com/datacarpentry/?utf8=%E2%9C%93&query=genomics

It has a different approach than Software Carpentry in that DC is doing discipline/domain specific workshops. My outsider perspective is that they're doing some good community building around bioinformatics training resources, and I thought you might be unaware. So, you might take another look and consider including. No worries if not, just pointing you at those resources ;)

Here's a first suggestion for a time plan for the project - if you have any suggestions/edits/etc. for it, add them below. If we get a lot, I'll add this instead as a file in the repo for us to edit as we want/need:

• mid September: agree a list of features of great open source scientific communities
• end September: agree a set of questions to use to reach out to relevant communities
• mid October: agree a list of communities and people to reach out to for anecdotes etc.
• end October: contact people in relevant communities to get their feedback
• end November: "deadline" for feedback from other communities
• end January: first draft of manuscript
• end March: submitting manuscript

Based on the BioJS eLife article, and the content collected during the Open Space session at BOSC2015, I've started adding more flesh to the features document.

I'm trying that, so we can get a feeling of how is a good way to add/introduce examples/anecdotes from many different example OSBCs. Hopefully, when others see how we've added such content, it'll be clearer to them how to do the same.

As I don't actually contribute any code, or use any of these myself at the moment (as a project manager I hardly ever write any code any more), would be great if some of you could have a go at adding this kind of thing based on your much more current experiences of this kind of thing!

I may need, once several of us have added something, discuss how to most usefully do this, but we won't know that until we start looking at what different people put in.

Several issues were linked to ways of collecting examples/anecdotes from others re: their experiences working in OS scientific communities. To (hopefully) make it easier to work through this issue list, I'm closing some of them and linking to them here:

• #21 "Adding examples/experience/anecdotes from real communities to our collected features."
• #18 "Which kind of communities to include in discussion/examples?"
• #13 "Collecting list of successful Open Source communities"
• #9 "Attracting more contributors"
• #15 "Revisit Data Carpentry, maybe?"

We've collected a list of potential communities to reach out to in a file in the repo: listOfOSBCs.md

I think it'd be good to work with a transparent, agreed-upon set of criteria, for deciding which ones to contact directly for their input, and that's been discussed already in several of these issues (and in the Hangouts).

Wanted to check we have consensus to this - please give your opinions on this, in particular letting us know if you agree/disagree with it.

• scope to include any scientific open-source communities - we assume that most of the relevant issues/features are common across all of them, and that we'll be biased towards bioinformatics, given our backgrounds (see issue #13)
• we want to include only communities that are open source, using the the OSI definition and their list of approved Open Source licenses. #18

Questions:

• do we want to restrict our outreach to communities we decide (using the criteria we've already collected in this repo, but perhaps needing to be quantitative/more specific or prescriptive about how we measure/decide that a community indeed has such a feature) as being "great"? Benefit I see with this option, is that we 'quality control' that the anecdotes are indeed representative of the kinds of communities we perceive others are likely to want to emulate the success of. Disadvantage is that there may be useful insights and lessons to be learn from communities that are less successful. @bgruening discusses this in #13
• do we consider a 'community' project to be one that has more than 1 collaborator? i.e. how many people need to be involved for us to consider it to be a 'community'?

This post seems relevant, and has a lot of good stuff. Not sure how best to integrate what's in there - perhaps just in the category of prior (or concurrent) art?

Sarah Sharp (2015) What makes a good community?

The author was until recently a Linux Kernel developer.

In #17 @bgruening suggested spinning out some of his questions in there to separate issues - here I'm doing that with something several of us have discussed.

My first call would be to include things that are definitely "Open Source". Although I don't have a definition of that - someone with more knowledge of this could perhaps propose/suggest one/some generally used ones?

Could we comment on the redundant points? for example:
Following points more or less indicate to same feature "Open transparent effective efficient respectful communication"
Extensive access to and communication with users
Quick responses to bug fixes/feature requests
Easy to contribute to

Ismail and I have updated the list of features of great OS scientific communities, to reflect the GoogleDoc we wrote at the Sep 10th 2015 GoogleHangout.

Looking through the text again, I think a useful way forward could be for different people to take on the content of different features ("Accessibility", "Communicativeness", etc.). Each of these people, could try and write "their" feature up, to include:

• a short explanation of what is meant by it
• an explanation of how this feature is of benefit to a community
• a list of practical tips for achieving this feature
• a list of challenges for establishing this feature

If you think a different focus/summary/structure would work better, please let us know, we can try and choose something together.

If we find this list OK, we may still struggle to use it in each case - but we'll have to try to find out.

Once several people (or groups of people) have had a go at several of the features, we can look at what we've come up with, compare the texts, and see if we can use these to decide on a general style/focus/tone for these texts.

Just as reference, here is a link to the shared Google Doc that everyone is currently editing...

https://docs.google.com/document/d/1jIYaBLoW02sCDNDgikYNWipaQwfVrRQI5QmhM0H-wHc/edit

Any of you up for meeting online to talk about this?

I'd be really interested to discuss like that to hear what you'd like to do with what we have so far.

We discussed and (I think) agreed that a good way to add content linked to experiences from many different OS scientific communities, would be to directly contact people working on those communities, asking for their feedback.

We thought we could create a file in the repo for those people to fork and pull request on to collect the info - while also providing a googleform for them to use if they prefer to contribute in that way - then, if they're interested in being included as an author on the article, they'll need to give us a way of contacting them to follow up.

We thought a good way to collect this feedback would be to ask them a few (max. 3?) questions linked to their community.

How about these three questions? Please send any comments/edits/changes to these, I just put them forward as a way of kick-starting discussion. I found I wanted to provide also a text contextualising the questions - otherwise it was too hard to sum up the details of what I think we're looking for in just a couple of sentences. Very glad for feedback also on the context text.

"We would like to understand better, the features of great open source scientific communities. We've written together a list of features we think are shared by most or all communities of this kind that many of their participants (and others) would consider great:

• give
• the
• list

We'd like to illustrate how these features are an important for making these communities great, and give practical tips on how to achieve them, using anecdotes taken from the activities of communities of this kind.

With this in mind, please send us examples of anecdotes from the activities of your communities, that illustrate:

• challenges that you overcame, and how you overcame them, to build and maintain a great community
• ways in which your community benefits from displaying some of the features from this list

We're particular interested in collecting anecdotes linked to the open source, software-focused, natures of these communities - we hope that will make what we collect as useful as possible as a source of inspiration and advice for others working in other OS scientific communities.

We've created a file in the github repo we're using to collect these anecdotes - you are very welcome to fork the file and make your contribution using a pull request - that makes it easy to keep track of who's contributing what to the project (if we are able to turn the resulting information into a published manuscript, we'll contact everyone who's contributed to it, and invite them to join us as authors on it). Or, if you prefer, you can just send your feedback to me here at this email address. In that reply, please indicate whether or not you want us to make a note in the repo that you've contributed to it, and if so by what name."

The idea is to have a list of communities and write a short paragraph about them.
Moreover, we can contact each of them to get opinions and solve: #9

• Galaxy
  • GalaxyP
  • IUC
• BioJS (community paper)
• BioConductor
• BioPython
• BioPerl
• GMOD
• Education
  • data carpentry
  • software carpentry
  • Goblet

Please feel free to add more.

Hi Bjoern, Hi Malvika,

Thanks so much for contributing to this!

Would be great to have some more people giving their insights.

I guess the ways to get more are:

• wait (someone mailed me to say they'll do something next week)
• ask people we know (particularly those at the sessions in BOSC) to contribute
• ask/contact some of the 'obvious' communities directly, asking for their input (I don't know, bioperl, biojs, etc.)

If we make a list of projects here, would you be up for dividing them up to contact them?

frustratingly, my laptop has just died, so I'll be slower than planned the next days :-/

With this issue I will collect some questions I would like to discuss/clarify in the manuscript + some random thoughts. Please feel free to take a few of them out and create a new separated issue to start the actual discussion.

• We are not talking about Open Source in general, I assume, but about scientific Open Source communities?
  • Linux/Debian/LibreOffice are out?
• Which role does a community manager play?
• Which role does the license play?
  • GATK example
  • copy-left licences
  • academic-free licence only?
  • copyright assignment
• What are appropriate levels to acknowledge committers?
  • Travel support
  • Co-authorship
• What role does a funding agency play?
  • A feature that was in integral part of a grant, was developed not by the grant-writing lab but by by the community.
  • travel sponsorship of foreign community members?
  • How should a letter of support from the community look like so that a funding agency takes it seriously?