These are the beginnings of an entry for openscienceprize.org. Join in if you are interested!
Participants: Tim Head, Titus Brown, et al.
There is quite a bit of work left to do to have a complete proposal. If you have comments, please create an issue.
Build the infrastructure required to create and publish scientific output that is more than a simple, static document.
A toolbox that makes reusable science a first class citizen in the academic workflow by being useful from day one of a research project. Reusability as a side effect.
Combining ideas from mybinder, everware, thebe, github, docker, travis, and zenodo. Useful for researchers from day one. Runs locally on your laptop or lab's cluster. Provide a way to automate every step. Opinionated. Helps setup structure of a research project. A toolbox, not a framework. Most individual steps already have fantastic solutions out there, people should use them! Final step: interactive, re-mixable papers.
Read the full proposal.
Check out issue #14 if you want to collaborate or become a co-proponent.
Everyone is welcome to contribute. Discussion happens in issues, and changes are handled via pull requests.
You can contribute by filing an issue on something you find unclear, needs improving, is missing, is incorrect, or add an idea of your own. Or directly create a pull request with your suggested edits.
As this is about creating a entry for the open science prize, creating the entry should happen in the open as well. Licensing details and a list of contributors can be found in LICENSE.
Watch the repository - this way, you will get email notifications of new conversations. You can do this by clicking on 'Watch' at the top of the repository:
This is an unordered collection of thoughts. Feel free to add to it.
Build a tool that universities/labs/publishers can use to encourage re-use of software and provide a completely new kind of research article, one that is interactive!
- preserving and open access of data is becoming mainstream
- to make sense of data we need lots of documentation and experience
- one source of documentation and experience is the code used to analyse the data in the first place
- sharing code is not enough, code only makes sense in the environment it runs in
- so far very little effort on sharing code and its environment
- adding re-usability/re-producibility as an after thought does not work (misaligned incentives)
- make re-usability a waste product, provide tools that are useful to researchers while developing their analysis pipeline that produce re-usability as a side effect
- bit rot is everywhere, undetected bit rot is a disaster
- periodically and automatically run a subset of analyses to detect bit rot
- researchers use very diverse set of tools, not interested in changing them
- need a flexible system that allows composing
- machine readable environment and entry point allow for completely new kinds of scholarly communication (-> interactive articles)
- simple syste that can be setup at universities or labs or publishers
- lots of problems already solved (sharing -> git(hub), workflow -> make, snakemake, etc), do not reinvent them, instead allow composing of these tools
- do not build a platform that people have to adopt to, or at least reduce the adoption as much as possible and provide benefits to researchers from day one.
- support a model of "bring your own computing", allows users to provide their own computing resources instead of beign tied to host.
- BYOC is inherently flexible in terms of how compute power is provided