Binderhub: BinderHub + JupyterHub Scipy 2018 Talk/Poster Proposal
Click here to make additions/comments/edits on the SciPy proposal
The scipy 2018 deadline for talks/proposals is February 15th.
Submission guidelines page: https://scipy2018.scipy.org/ehome/299527/648140/
This issue is meant for discussing a submission for a talk/proposal for SciPy 2018.
choldgraf
All 21 comments
Don't think I will make it to SciPy. Going to fly the binder flag in Europe though.
betatim
on 22 Jan 2018
sounds good! I think that a lot of work can also just go into general tutorial materials etc that we can reuse for a number of different events
choldgraf
on 22 Jan 2018
Submissions are actually February 9th. I'm interested in being involved with this.
mpacer
on 23 Jan 2018
Happy to help with specific questions, but I don't think I'll have time to be super involved!
yuvipanda
on 24 Jan 2018
Happy to help on the slide deck and review proposal. I'm not sure if I will make SciPy this year. We may be doing some outreach classes at Cal Poly this July :-)
willingc
on 24 Jan 2018
Hmmm, I'm not sure we'll have the representation of people at scipy to be able to do this. I assume we need at least 2 folks there.
I might go, but it depends on how much other stuff I have going on that summer. Is anybody on the Binder team _definitely_ going to scipy that would like to help give a tutorial? @mpacer you mentioned you're interested in being involved, does that == "helping giving the tutorial" or are you going to be busy with organizing the conference?
choldgraf
on 29 Jan 2018
going once, going twice? I don't think we should commit ourselves to giving a tutorial at a conference if we aren't sure we'll have the #s to attend...
choldgraf
on 1 Feb 2018
Alright. So I was talking with @choldgraf on slack and had an idea that given the limited number of people who may be in attendance, it may make more sense to propose a talk/poster on Binderhub/repo2docker.
In addition to the talk/poster, we would submit a paper to the proceedings.
The topic of the paper would differ from a JOSS paper, instead it would focus on the research angles of the projects. The nature of the problem and the design decisions that were made when organising the libraries in order to achieve the desired performance characteristics. Analytics (including some nice figures) would be good for that as well.
mpacer
on 3 Feb 2018
I think that sounds like a good idea to me - we don't have a reliable guess on whether enough people will attend for a tutorial, and could use a longer form paper for people to cite Binder. I'm happy to help out with writing etc, and if I end up attending scipy could give a talk or something (we already have a few talks sitting around that we could use).
To me the big question is what kind of scope to put around a paper or talk. There's JupyterHub / repo2docker / the helm charts / the specific instance on mybinder.org. What would make for a compelling narrative? Anybody have ideas? Or maybe we can discuss @ the team meeting?
choldgraf
on 3 Feb 2018
To me the big question is what kind of scope to put around a paper or talk. There's JupyterHub / repo2docker / the helm charts / the specific instance on mybinder.org. What would make for a compelling narrative? Anybody have ideas? Or maybe we can discuss @ the team meeting?
I was specifically suggesting that we focus on the design decisions of the two libraries (binderhub and repo2docker). It probably wouldn't make sense to talk about those without covering some aspects of JupyterHub as well… but I figured that that could be (and has been?) a talk in its own right and so shouldn't be the main focus. I'm thinking the analysis of the specific service itself plays the role of the "results" section of the paper. I don't know enough about the helm charts to know what role they would have in this narrative.
mpacer
on 3 Feb 2018
Updated the top-level comment with more up-to-date info. Specifically, note that since we're now planning on a talk/poster we have until February 15th to submit. I can put together sometime that we can discuss @ the group meeting and we can plan to submit on the 9th in true fly-by-the-seat-of-our-pants academic fashion :-D If somebody else would like to put together this proposal, that'd be great too!
choldgraf
on 5 Feb 2018
Note we now have until February 15th to submit (I'll update dates above accordingly)
choldgraf
on 6 Feb 2018
Hey all - I've created a hackmd where people can make additions/edits/comments to the SciPy proposal:
https://hackmd.io/MYDg7AzAnALAjBAtAVgKYCMqJhAJgNkRH2QAZFVgpcoRlh9c4Yg=
anybody should feel free to contribute!
If someone would like to start a draft, you might wanna check out the JOSS proposal to mine some content.
choldgraf
on 7 Feb 2018
Just a note that @mpacer added a bunch of great content in the form of an outline! I took a quick pass at re-organizing today, can take another pass tomorrow. If folks have thoughts etc, please do jump in!
Note: we have 3 days to submit this proposal!
https://hackmd.io/MYDg7AzAnALAjBAtAVgKYCMqJhAJgNkRH2QAZFVgpcoRlh9c4Yg=
choldgraf
on 12 Feb 2018
The notes were quickly jotted down during a conversation with @minrk (and he was the source for all that good content). I'm sorry you had to wade through that which I had written in a fairly unstructured form.
I think some of the stuff that had been there wasn't useful due to the later portions of the conversation. I had intended to sort through that…
(edits)
Is there a way to look at the history so I can see the state from before your changes?
Answer: Yes. However, I looked at the revision history and it seems like it ran into some weird versioning problems during the middle of the conversation.
@choldgraf thank you for making it far more digestible by other people… I'll take a pass at transforming the outline content to the long description, _maybe_ the short description. My guess is that the outline will continue to be useful in designing the talk/poster (assuming it's accepted).
mpacer
on 13 Feb 2018
ah my bad, I didn't realize you didn't want folks to make edits to that - was just trying to be helpful :-/
I'll hold off on any more edits until somebody mentions that it's open for a round of edits etc!
choldgraf
on 13 Feb 2018
No, it's good now & open for edits! Because you had edited it, it's now much more interpretable! Thank you @choldgraf for all your hard work (& I'm still sorry for having made it necessary).
mpacer
on 13 Feb 2018
Hi @choldgraf @minrk I took a stab at reworking the the description to be further away from a generic Binder talk (which wasn't obviously differentiated from the original SciPy Binder talk) and tried to focus more on repo2docker and BinderHub as being newly developed components of the larger system which has been improved for large scale maintainability and deployment improvements.
I would appreciate feedback!
If people have ideas as to what details to go into for the long description, that would also be great.
Also, @choldgraf will you be at SciPy this year? If so I'd imagine putting credentials for you, @minrk (who mentioned explicitly that he would be at SciPy), and I… is there anyone else? I don't want to leave anyone out.
mpacer
on 14 Feb 2018
Hello all, please see the hackmd at https://hackmd.io/MYDg7AzAnALAjBAtAVgKYCMqJhAJgNkRH2QAZFVgpcoRlh9c4Yg=?
Unless someone wishes to make specific changes, I'm going to be submitting this version around 23:00 PST tonight.
mpacer
on 16 Feb 2018
I have made a https://github.com/jupyterhub/jhub-proposals/pull/1 to record the version that I am preparing to submit. Any changes will be made as a modification to the one in that PR.
mpacer
on 16 Feb 2018
Congrats on the submission! Fingers crossed it gets picked. Going to close this.
betatim
on 18 Feb 2018
Related issues
jpivarski
·
6Comments
choldgraf
·
6Comments
yuvipanda
·
3Comments
melaniewalsh
·
4Comments
rprimet
·
5Comments