Cultivating Community around software and
data
Scott Chamberlain (@sckottie/@ropensci)
UC Berkeley / rOpenSci
Open science ecosytsem
Data
People
Software
rOpenSci &
data providers
After working with
A LOT of data
providers
- API first! GUI website after (if ever)
- see #1
- see #2
-
Think about data size/use cases/how often
updated/etc.
-
Clients, clients, clients - users may use
website, but reproducbility should be paramount
- Reach out to rOpenSci - we can help
for any 1 resource, work can be distributed among groups
-
e.g.,
Fishbase
- rOpenSci built/maintains API
-
e.g.,
SeaAroundUs
- rOpenSci took over maint. R pkg
-
e.g.,
BIEN
- rOpenSci working on building API
ropensci &
software makers
How to nurture comm. around software?
-
software review!*
-
nurture comm. members at all entry points (code,
docs, bug reports, use cases, etc.)
-
hire specifically for comm. management (shutout
to ours:
Stefanie Butland)
-
helps to have people in many time zones
-
empower people to do what they love that helps
build community *
*: will talk more about these
rOpenSci Software Review
-
Completely open source tools
Free to run
-
All reviews/conversations in the open
-
Reviews are/can be linked to code changes
-
Paired with submission to a journal - JOSS (more
to come ...)
how to make a large and resilient contributor base?
-
be really nice!!!
-
> 1 contributor
-
give direct access to contributors
-
contributor diversity
turn users into contribs
make software work easier
-
make software count for tenure/grad school
decisions
-
more
-
more academic software jobs!!!!
-
drop "novel" work req.(~software maintenance)
-
provide (at least facilitate) software training
(don't make people learn on their own)
Open Questions
in the face of dwindling budgets, academia not valuing
software, etc., how do we:
-
maintain software through time, while people
come and go?
-
attract at least 2 people to all software
projects, no matter how small?
-
incentivize users and contributors to engage
more?
-
connect users to data through software in a way
that facilitates reproducble science?
-
minimize bad actors from spoiling communities?