Hi all,
Sorry for not writing a dev Meeting summary for the last 3 weeks, First there was no meeting on the last week of December and I missed the two previous one. Seem like the team was busy while I was enjoying being back in France, but still working. Hope 2017 is going well for you all, and here is again a summary of what happen, is going to happen and where we would like your feedback. Testing Etherpad We’re trying out etherpad this month for keeping our weekly meeting notes. Hackpad had enough issues that we’re trying to switch. The new meeting notes are here <https://etherpad.wikimedia.org/p/Jupyter-Weekly-Team-Meetings-2017-01-Jan> Project management (Jamie) Data Carpentry [When where] attendees made several updates to the Gallery of Interesting IPython Notebooks, please have a look at the new additions ! Also the Gallery of interesting notebook got retitled, and the URL was consequently auto changed. Please update the links if you see it somewhere. https://github.com/ipython/ipython/wiki/A-gallery-of-interesting-Jupyter-and-IPython-Notebooks GitHub Automation @MeeseeksDev (our in progress bot) is alive and well, plus already doing some good work. It has already automatically backported a few PRs and migrated a few issue. See the GitHub automation thread. If you have any feature request, questions, or is interested in having the bot on other repositories, let us know. IPython IPython 6.0 release originally schedule to be released by end of January this was planned during Spring dev meeting. Right now IPython 6.0 is a bit short on new feature and had mostly a lot of code cleanup. We think that releasing 6.0 without more new feature is not compelling enough so we’ll let IPython 6.0 release schedule to slide by a few month to have the chance to implement lot of new things ! Notebook WebPack removal PR is done, should be merged soon. This change fixes a bug that affected many existing notebook extensions (in 5.x) and ability to configure some options. https://github.com/jupyter/notebook/pull/2011 Mike asks about auto-id generation for headers which we want to update to a more modern standard mechanism. In particular one concern is synchronization between notebook release and nbconvert. - https://github.com/jupyter/notebook/issues/1961 - https://github.com/jupyter/notebook/issues/1899 - https://talk.commonmark.org/t/anchors-in-markdown/247/21 Above discussion tuned into a conversation about what kind of Markdown “spec” the Jupyter notebook should use. Right now, the current markdown spec is a bit fuzzy and no single implementation (beyond what the current notebook is) does have exactly the same set of features. Having a spec would be a necessity for alternate implementers to know what to expect. Common mark is great, but move too slowly. We don’t implement (and can’t implement) all of GitHub markdown. If you have any thoughs on this subject, please let us know. https://github.com/jupyter/nbformat/issues/80 is the issue about documenting the markdown spec we use as part of nbformat. If you are using markdown in a non standard way, please let us know. Nick wonders if we can relax the markup type from “Markdown” to just “Markup” and add an additional field and have it extensible, and let people make their own extension. That would allow to also have rich text editor like ProseMirrror. That may be a topic for later. JupyterLab Big refactor of console and notebook instantiation using factories, standardizes how editors, completers, and inspectors are passed into higher level objects. https://github.com/jupyterlab/jupyterlab/pull/1482 Tests, bugfixes, and resilience hardening of state restoration (ongoing). As a reminder, state restoration is what allows you to refresh the page and keep your layout. https://github.com/jupyterlab/jupyterlab/pull/1501 Overhaul of build system - updated for Notebook 4.3.1 and used best practices from nbdime. https://github.com/jupyterlab/jupyterlab/pull/1477 Ongoing discussion on the version compatibility of extensions with jupyterlab itself and between extensions. Issue us that node is not availlable at runtime: https://github.com/jupyterlab/jupyterlab/issues/1011 IPywidgets The widget team is continuing to work towards a 6.x release. Jason is concentrating on discovering and fixing regressions with 5.x. Mimerender extensions (Grant) PR has been submitted to jupyterlab_vega: https://github.com/altair-viz/jupyterlab_vega/pull/3 This now supports both lab and classic notebook. Grant is working with Brian and Jake to transition Altair to mimetype-based rendering: https://github.com/altair-viz/altair/issues/283 and making progress on jupyterlab_table (JSON Table Schema): https://github.com/gnestor/jupyterlab_table/tree/dataresource Pandas support is getting closer: https://github.com/pandas-dev/pandas/pull/14904 The table schema spec v1 release is imminent: https://github.com/pandas-dev/pandas/pull/14904#issuecomment-273037853 JupyterHub JupyterHub activity this past week focused on the JupyterHub repo and JupyterHub Teaching Reference Deployment repo. We triaged and closed 16 open issues on the JupyterHub repo. Two PRs related to Services were merged; one PR allows Services to request admin access. Work has continued to simplify the JupyterHub teaching deployment based on Brian’s experience at Cal Poly and to address a few minor configuration issues with nbgrader. Thanks to Jess for her support and the upcoming nbgrader 4.0 release. For full details of JupyterHub activity for this past week, see https://github.com/willingc/org-pulse/blob/jhub/2017_01_17_jhub.md. nbconvert Mike met with the group from data carpentry — great conversations and learned a lot about Rmd (R-Markdown). Customisation is important to this community, as well as hiding input. related issues are : - https://github.com/jupyter/nbconvert/issues/512 : notes on customisation (particularly for in browser customisation) - https://github.com/jupyter/nbconvert/issues/445 : notes on hiding input (in notebooks in converted formats, not live notebooks) There are currently possiblities to customize the “Download as” menu from within the notebook to get custom output but this requires shenigans for now. Apecifically, if you wanted to use a custom template this it would require: - create blah.tpl using text editor; previously editing it wouldn’t have worked, https://github.com/jupyter/notebook/pull/2038 fixes it. - create/edit jupyter_config.py & jupyter_notebook_config.py to have c.TemplateExporter.template_file="blah". - close notebook server (or open new notebook server in the same view) - Download notebook Preview and live views made possible in jupyter labs that autoupdate would be good, possibly via an “exporter extension”. Cell level tags (string based metadata) in the notebook are needed to make a lot of customisation easier. JupyterLab issue for the export menu: https://github.com/jupyterlab/jupyterlab/issues/964 Talked through custom script exporters w/ Thomas decided to use a new group of script exporter entrypoints, rather than notebook metadata mechanism. This would allow for notebook/kernels authors to provide a “preferred” exporter for notebooks of a particular language, instead of having the users configure everything. This has the advantage of working without having to update all the notebooks. Related issues: https://github.com/jupyter/nbconvert/pull/414 That’s it At least for this for this week. As usual let us know if this summary is useful and feel free to ask questions or comments. Thanks -- Matthias -- You received this message because you are subscribed to the Google Groups "Project Jupyter" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/jupyter/10eaab33-fecf-425c-b455-5056f80929c7%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
