Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add dashboarding section #8

Merged
merged 2 commits into from
Apr 22, 2016
Merged

Conversation

parente
Copy link
Member

@parente parente commented Mar 31, 2016

### jupyter/notebook (>4.2)

* Add a URL in the single user notebook server that shows a rendered dashboard.
* TODO: This is a straight copy/paste from the hackpad. Why / how did we say this would go into notebook server? The layout mechanism is one extension for classic and a separate plugin for Jupyter Lab. how would this server endpoint know what to serve up to render the dashboard? Or is this a completely separate renderer that has to be written (and why?)
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@ellisonbg @fperez @bollwyvl @minrk There's an action item to add a URL to the notebook server. I'm not clear on how that would work given there would be two implementations of the dashboard authoring / rendering in the notebook server, one as a classic extension and one in Jupyter Lab. Which would render the output at the URL? Do we want the server dependent on one or the other since they're both, technically, plug-ins?

@parente parente changed the title Add dashboarding section [WIP] Add dashboarding section Mar 31, 2016
@parente parente changed the title [WIP] Add dashboarding section Add dashboarding section Apr 1, 2016
@parente
Copy link
Member Author

parente commented Apr 1, 2016

Need feedback on the TODO before merging.

@rgbkrk
Copy link
Member

rgbkrk commented Apr 4, 2016

Perhaps we can iterate on the TODO with another PR? Or even delete it?

@parente
Copy link
Member Author

parente commented Apr 4, 2016

We could, but it's a pretty big question mark in the roadmap to not know what is going to do the actual dashboard rendering in the notebook server.

@rgbkrk
Copy link
Member

rgbkrk commented Apr 4, 2016

I'd make the wording allow for any of the approaches outlined to be able to handle it so it doesn't block the rest from coming in from this PR.

@parente
Copy link
Member Author

parente commented Apr 4, 2016

@rgbkrk rewritten to mark the specific renderer as TBD. Is this cool with others?

To clarify, I don't really care what renders it, I'm just sanity checking that we did agree that something in the notebook server would have this responsibility.

@rgbkrk
Copy link
Member

rgbkrk commented Apr 4, 2016

/cc @ellisonbg

@fperez
Copy link
Member

fperez commented Apr 5, 2016

Yes, I think we did agree to that :) Implementation specifics TBD, but for sharing/collaboration reasons in some contexts, that seemed like the right way to go and I don't recall any opposition.

@parente
Copy link
Member Author

parente commented Apr 5, 2016

OK. Then I've got no other outstanding edits. Someone can click the big green button.

@parente
Copy link
Member Author

parente commented Apr 7, 2016

As @bollwyvl, I and others noodle on the proposal for layout metadata in notebooks (https://jupyter.hackpad.com/Document-Layout-Metadata-infor-Notebooks-pjlprsYHvtw), I'm wondering what the broader community thinks we should be doing with respect to the layout spec:

  1. Writing a light spec about where in the notebook document layout tools (dashboards, nbpresent, RISE, ...) should store their metadata to avoid conflicts and enable simple discovery. This can be done quickly and allows anyone to write their own extensions / plugins how they see fit for whatever layout he/she is trying to support (slides, posters, mobile dashboards, ...), but gives no interoperability guarantees.
  2. Writing a detailed spec about how all layout tools should adopt a common format for specifying arbitrary layouts with sections, coordinates, dimensions, scaling, viewports, etc. This will enable interoperability across tools if/when a common schema can be designed and implemented across all tools for all use cases and layouts.
  3. Both: one then the other?
  4. Something else entirely.

@rgbkrk
Copy link
Member

rgbkrk commented Apr 7, 2016

In the short term, 1. In the long term 2. I guess that means 3/both.

Solely because 2 may take forever while we can learn how 1 shakes out.

@damianavila
Copy link
Member

In the short term, 1. In the long term 2. I guess that means 3/both.

Ditto.
Btw, PR LGTM.

@parente
Copy link
Member Author

parente commented Apr 22, 2016

We've spec'ed out the current jupyter-incubator/dashboards metadata and rendering requirements here: https://github.com/jupyter-incubator/dashboards/wiki/Dashboard-Metadata-and-Rendering

Next up is the light spec for layout tool metadata as an enhancement proposal.

Any reason not to merge this PR?

@damianavila
Copy link
Member

Any reason not to merge this PR?

No reason, I think... it was opened more than 20 days, people offered feedback and we could change things later with another PR (if there is more discussion on this topic to be landed in the roadmap). Merging...

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants