Binderhub: Extending Binder or JupyterHub for Scientific Publishing

[jzf2101]

@kyook and @natesjacobs and I discussed the idea at FORCE today of preventing automatic image builds of binder so that people could peer review a repository. What do people think of this?

[choldgraf]

interesting - why would "preventing automatic image builds" help out here? (as opposed to, say, forcing repositories to specify a commit hash or a tag associated w/ the submission date)

[jzf2101]

Because the form currently builds whatever it's given. They're interested in reviewing a repo before pushing the image to the public.

[betatim]

I'm not sure I understand this use case. Wouldn't the review happen on GitHub or where ever the code is hosted? You could even integrate a repo2docker run in your CI to verify that the repository will build.

How would this flow? You submit a link to a repo in the form on the frontpage of your BinderHub and then an admin has to review it before it gets built?

On a small scale you could handle this by adding/removing repositories to the list of banned repos.

[jzf2101]

So for example, in a journal, the authors submit a manuscript for review. It undergoes a review and the authors revise the manuscript before it is finalized and published.

In something like SciPy - this happens in a PR - https://github.com/scipy-conference/scipy_proceedings/pull/386

In both these cases though, we're talking about a finalized PDF that is produced.

What if people wanted to submit the the repo to be built and shared as a live Jupyter Notebook? Can people use binder for this case?

My thought is that an author can submit the repo on the form and then a review can come in as issues that can be reviewed until a specific commit is selected to be built.

[jzf2101]

@tompollard @alistairewj this is one thing we were talking about for binder with scientific publishing

[betatim]

Maybe this could be built as a separate project that implements this reviewing and commenting workflow (based on GitHub as backend?)? Something like the process that https://joss.theoj.org/ uses maybe.

It isn't clear to me why we should prevent people from launching various versions of a repository, even if they aren't the "reviewed" or "published" version. This is why it sounds like this is more about facilitating the review workflow and then publicising the particular binder URL that has the right commit in it.

[kyook]

Sure allowing people to release versions regardless of review should be
fine. However, it seems good to have a way to review and vet a version as
well as mark/highlight a version that has passed all reviews.

~~~~
Karen Yook

Curator
WormBase Caltech
Tel: [phone]
e-mail: [email protected]
e-mail: [email protected]
skype name: wbkaren

On Wed, Oct 17, 2018 at 2:13 AM Tim Head [e-mail] wrote:

Maybe this could be built as a separate project that implements this
reviewing and commenting workflow (based on GitHub as backend?)? Something
like the process that https://joss.theoj.org/ uses maybe.

It isn't clear to me why we should prevent people from launching various
versions of a repository, even if they aren't the "reviewed" or "published"
version. This is why it sounds like this is more about facilitating the
review workflow and then publicising the particular binder URL that has the
right commit in it.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/jupyterhub/binderhub/issues/686#issuecomment-430502394,
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAlVGhkiMslaiHBAHkfjDeK2jXE2Wz0oks5ulsqEgaJpZM4XWs47
.

[choldgraf]

cc @jbpoline who has also discussed this in the context of a project he's working on called Aperture

[jbpoline]

@jzf2101 @choldgraf @betatim
Aperture will be the OHBM publishing platform, with the Canadian Open Neuroscience Platform (CONP) providing back end for it, using the compute canada infrastructure. We hope to have something started in june 2019

[kyook]

@natesjacobs perhaps Aperture is what you are looking for?

[jzf2101]

@jbpoline were you at FORCE11? We were talking about publishing with Binder there a few days ago

[jbpoline]

@jzf2101 : I was the chair of the local organizing committee - (but still I had to miss some of it because of local commitments).

[jzf2101]

Did we meet at Open Beer Science then?

[meeseeksmachine]

This issue has been mentioned on Jupyter Community Forum. There might be relevant details there:

https://discourse.jupyter.org/t/guidelines-for-submitting-a-notebook-for-peer-review-today/3533/10