http://www.nature.com/authors/policies/license.html

#64

@jasongrout suggested we add license information based on #7

@jasongrout @ian-r-rose @ellisonbg @sccolbert Would like to get feedback from your experiences giving the demo at Scipy and PyData Seattle. Could you comment here or create issues as needed?

@jasongrout Also we'd like any content you ended up using in your content. Is there a video of your talk available?

#31 change Copyright date from 2016 to 2017. IIRC the copyright date should be the date of first copyright and is implicitly YYYY-Present. Should it be changed back ? @ellisonbg ?

I don't think that the invoke commands install the kernels for the julia and R notebooks.

Either document that they need to be installed separately or install if possible.

Do we want this is a wiki or in the repo in a markdown?

Tried 'invoke environment' inside the repo directory using:
miniconda 4.5.11
python 3.6.5

invoke environment fails with the following messages:
Solving environment: failed

UnsatisfiableError: The following specifications were found to be in conflict:

• scikit-learn=0.18
• xtensor-blas=0.12
  Use "conda info " to see the dependencies for each package.

Any help is appreciated!

Looks like the last build has not been passing in Travis with:

AttributeError: module 'tornado.web' has no attribute 'asynchronous' 

Possibly related to:
conda-forge/nbconvert-feedstock#27

Note that that currently this repo does not work on binder (running notebook simply hangs)

@sccolbert wants there to be a vegalite or altair example that works

@jasongrout @ellisonbg : @fperez asked me to use this demo to start putting together initial user documentation for lab. I can't write issues on the forks, but since @jasonsgrout 's fork is far ahead, is it possible we could PR the forks?

So people can walk themselves through the demo

jupyterlab/jupyterlab#3173

@jasongrout

Invalid requirement: 'bqplot=0.9.0b10' = is not a valid operator. Did you mean == ?

- Writing config: /Users/4d/anaconda/envs/jlabdemo/etc/jupyter/labconfig X is not version compatible with installed JupyterLab version 0.16.2 Expects JupyterLab version ^0.13.1 from packaged module [email protected]/lib/widget.js Expects JupyterLab version ^0.13.1 from packaged module [email protected]/lib/plugin.js

Collecting dask-labextensions Could not find a version that satisfies the requirement dask-labextensions (from versions: ) No matching distribution found for dask-labextensions

Once conda-forge/staged-recipes#14245 is merged and the packages available on conda, we should be able to remove this task:

And add the dependencies to:

Which should make the Binder build faster.

I tried to get the demos working for a presentation this morning. Even after using invoke to install the data files, the files could not be found. Are some of the datafiles missing? Or is it a path issue on my part?

@jasongrout Since a number of the Lab extensions aren't working eg widgets and geojson, should we bother with including content that relies on those extensions? For example the geojson of the museums in DC

@jasongrout suggested we convert the makefile to a Python installation script while we pin down what external content we use

Using:

• Windows 10
• Miniconda
• Anaconda Prompt to run conda

(base) C:\>conda --version
conda 4.4.10

Installation was ended successfully

(base) C:\>conda install -c conda-forge invoke pyyaml
Solving environment: done

# All requested packages already installed.

But creating the environment failed -

(base) C:\>invoke environment
Can't find any collection named 'tasks'!

Thanks.

I'm getting this result from chrome console on mybinder.org when notebook is trying to connect to kernel.

WebSocket connection to 'wss://hub.mybinder.org/user/jupyterlab-jupyterlab-demo-azimq5mq/api/kernels/267f4698-ce2a-443d-aab9-46bd61a49ffb/channels?session_id=8d0e4978-7cb6-4184-82b3-4bf6fa323698&token=Qy1Nd1A6RDWoJEL5lUqo_w' failed: Error during WebSocket handshake: Unexpected response code: 502

Meanwhile, the top right status dot keeps showing Kernel Reconnecting.

When I visit the normal notebook (path: /notebooks/demo/Lorenz.ipynb), 500 : Internal Server Error is returned.

The current CI job is not checking that JupyterLab is starting.

Running on Windows -

(base) C:\Code\jupyterlab-demo>invoke demofiles
cleaning demofiles
creating demofolder
cloning repos into demo folder demofiles
Cloning into 'PythonDataScienceHandbook'...
Cloning into 'Urban-Data-Challenge'...
Cloning into 'altair'...
Cloning into 'QuantEcon.notebooks'...
Cloning into 'TCGA'...
Cloning into 'TensorFlow-Examples'...
Cloning into 'bqplot'...
'touch' is not recognized as an internal or external command,
operable program or batch file.

It happens because this line:
ctx.run('touch move_this_file.txt; mkdir move_it_here')

'touch' program, doesn't exists in Windows, that why it fails.

Maybe, the code can be updated to check if it is Windows,
and run equivalent command, like:
type nul > move_this_file.txt

The same way the code does for rmdir.

Thanks

See #2, #9, and #10

We could include a geojson file of lightning strikes, and maybe a large csv of them, you know, for lightning talks :). It'd be fun to give a talk where the data was all puns on lightning talks.

How about this one?

https://commons.wikimedia.org/wiki/Category:Hubble_images_of_interacting_galaxies#/media/File:Hubble_Interacting_Galaxy_AM_0500-620_(2008-04-24).jpg

Hi, jupyterlab-plotly extension is installed, but plotly itself is missing.

### ###

Based on the documentation I've read so far from invoke, it's unclear to me if the shell commands in the invoke script will run on Windows. Could someone confirm?

@willingc @jasongrout @ian-r-rose @ellisonbg @sccolbert @blink1073

Do you have any additional feedback from the conferences you recently spoke at?

Right now the documentation for JupyterLab points to a Binder link that installs 2.X, not 3.X!

Following #115, the R kernel is no available - we should fix this.

Originally posted by @fcollonval in #115 (comment)

@ian-r-rose @ellisonbg We've been using dummy files and folder movethis and movethishere to show drag and drop. Do we want to keep these in the repo? They're currently not in master or the invoke script?

@willingc @ellisonbg @jasongrout Based on an earlier conversation, we think it would make sense to have profiles for different talks in a yml file. An invoke script would copy relevant files into a folder for the talk. This way we can customize talks to different audiences.

@jasongrout @fperez : @ellisonbg suggested we think about ideas for content and pin down the provenance of our content in the demo. I've started by creating a dropbox paper document that lists where everything comes from. Please fill in relevant information if I've determined you've made the first commit

See https://github.com/youtube/api-samples/blob/master/python/captions.py

Hello from mybinder.org,

we had a bit of a rough evening since you merged #73. We aren't quite sure why this caused us problems or what is wrong with the build for this repository. However we observed very high CPU load, and eventually noticed that lots of ipython demo pods were running but none from jupyterlab. The first time our CPU usage went through the roof was exactly when you merged #73. So for the moment we banned this repository. After the first few minutes it seems like this resolved our problem.

You can try and build the repository locally with repo2docker or making a fork which should not be banned.

You can find us on gitter: https://gitter.im/jupyterhub/binder or we discuss in this issue

What if any relationship is there between jupyterlab/examples and this repo?

We should add https://github.com/phosphorjs/phosphor/tree/master/examples/example-datagrid to the repo so we have a standalone version of http://phosphorjs.github.io/examples/datagrid/index.html available.

@ellisonbg Since we moved the repo, should we also reassign the license?

"invoke environment" fails:

(base) <repo>\jupyterlab-demo-master>invoke environment --env-name=jupyterlab-demo
creating environment jupyterlab-demo
Collecting package metadata (repodata.json): ...working... done
Solving environment: ...working... failed

ResolvePackageNotFound:
  - xeus-cling[version='<=0.11']

btw:

(base) ... >conda config --add channels conda-forge --yes
usage: conda-script.py [-h] [-V] command ...
conda-script.py: error: unrecognized arguments: --yes

@ellisonbg we discussed moving repo Can you confirm when the repo is moved?

Per suggestion from @fperez , I tried installing the fork from @jasongrout but make demo doesn't handle the environment setup from INSTALL.txt Should there be changes to the makefile or readme to take care of environment setup for the demo?

An enhancement for places in the world that have less than optimal connectivity and internet speeds.

@jasongrout suggested we set up CI for the demo