⚠️ Dependabot is rebasing this PR ⚠️

Rebasing might not happen immediately, so don't worry if this takes some time.

Note: if you make any changes to this PR yourself, they will take precedence over the rebase.

Bumps ws from 6.1.4 to 6.2.2.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

⚠️ Dependabot is rebasing this PR ⚠️

Rebasing might not happen immediately, so don't worry if this takes some time.

Note: if you make any changes to this PR yourself, they will take precedence over the rebase.

Bumps three from 0.78.0 to 0.125.0.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Description

This will contain front-end modifications for custom software environments. So far

• show the name of the chosen image

Testing Steps

Relevant Issues

Checklist:

• [ ] No debugging console.log messages.
• [ ] All new code is actually used.
• [ ] Non-obvious code has some sort of comments.

Front end:

• [ ] Restart at least one project and check its ~/.smc/local_hub/local_hub.log
• [ ] Completely restart Webpack with ./w in /src
• [ ] Completely restart the hub by killing and restarting ./start_hub.py in /src/dev/project
• [x] Screenshots if relevant.

in project settings, explaining which features are available

default

old image, from a year ago

custom software, in that case it is just for a specific python library

Description

Jupyter notebooks without a kernel or an unknown one are a source of confusion. This patch presents a clear selection dialog with some explanations about what a kernel actually is to the user.

Reference: there are two situations the dialog shows up. either user requested, or when opening and the kernel is "bad". here is a side-by-side view of running this in two instances. The dialog only closes when a kernel is set by itself, when it wasn't opened by the user in the first place.

• no kernel
• unknown kernel

Testing Steps

important: switch to experimental image to see the suggestions.

1. existing notebook with known kernel → this doesn't show up
2. click on kernel info at the top right → kernel "change" dialog shows up
3. open a new ipynb file → dialog asks what to do (slightly different from above)
4. an unknown kernel can be created like that:

~$ cat <<EOF > unknown-7.ipynb
{
 "cells": [ ],
 "metadata": {
  "kernelspec": {
   "name": "pythonXY"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}
EOF

~$ open unknown-7.ipynb

Relevant Issues

I didn't check

Checklist:

• [x] No debugging console.log messages.
• [ ] All new code is actually used.
• [ ] Non-obvious code has some sort of comments.

Front end:

• [x] Restart at least one project and check its ~/.smc/local_hub/local_hub.log
• [x] Completely restart Webpack with ./w in /src
• [ ] Screenshots if relevant.

Back end:

• [x] it touches the "project"!

First step towards "assisted" grading. The most fundamental idea is to assign points to files (later, also automatically via some magic™) . Everything syncs as usual, which is pretty neat.

The most important aspect to review is the structure of the course file. It's modeled closely how grades are done. But additionally, there is a mapping from each file path in the collected files of the respective student to an integer (the points), and defaults to zero → see course/actions::set_points.

• On Firefox we disabled Jupyter entirely, temporarily (see #1537).

• We have had regular reports from users of issues over the years, and no amount of rewrites or subtle bug fixes has 100% eliminated these.

• Upgrading to new versions of Jupyter is terrifying since the API changes, so things randomly break (for our very brittle use of Jupyter).

• We are massively abusing and monkey patching Jupyter as it is to get sync to work.

The only viable solution I can think of is to completely rewrite Jupyter from scratch, except the backend kernel functionality. Thus this issue is labeled "insanely hard".

cling is a C++ interpreter based on clang/llvm. Would it be possible to install it on cocalc, together with its kernel? I'll be using it for teaching C++ this coming semester. We will be using Jupyter locally on the university's machines, but having a backup solution on cocalc would be helpful!

@SylvainCorlay implemented a Jupyter kernel implemented in C++ which is more robust than the one provided by cling. And will soon enable widgets and plots!

Installation instructions with conda:

conda install cling xeus-cling notebook -c QuantStack -c conda-forge

One should be able to install the kernel in another Jupyter deployment, but I haven't tried that.

Just like

+New --> Jupyter Classic Server, and +New --> JupyterLab server 

are 1-click ways to work with a CoCalc project through the Jupyter class and JupyterLab interfaces, we should also have a 1-click way to work with a CoCalc through VS Code. This is technically possible, and probably pretty easy, due to the open source MIT licensed code-server project. Here's an example of how that might feel:

https://github.com/betatim/vscode-binder/tree/master

• How To Set Up the code-server Cloud IDE Platform on Ubuntu 18.04 or maybe this
• maybe this

NOTE that evidently a possible shortcoming of code-server is Microsoft's licensing for extensions, which "prohibits use with any non Microsoft product." Since probably any interesting use of CoCalc + VS Code would involve extensions -- e.g., LEAN or Jupyter/Python -- this might be a big problem. Or maybe not. I don't know.

Description

This is for making smc_sagews compatible with Python 2&3 (Sage 9.0).

My strategy is:

1. make this work with sage 9 / py3 – but be intelligent to not break too much, aim for dual compatibility
2. once it works, fix sage 8.9 running py2. having a diff of all changes at hand here will help to iron out the remaining problems.
3. of course, check again for sage 9

Testing Steps

Relevant Issues

Checklist:

• [ ] Run eslint on new and edited files
• [ ] All new code is actually used.
• [ ] Non-obvious code has some sort of comments.

Front end:

• [ ] Restart at least one project and check its ~/.smc/local_hub/local_hub.log
• [ ] Completely restart Webpack with ./w in /src
• [ ] Completely restart the hub by killing and restarting ./start_hub.py in /src/dev/project
• [ ] Screenshots if relevant.

here is the first step: distinguish 2 cases

then one need to know what format should be send to the next step in the toolchain

could you remind me where to find that ?

Currently many of commonly used Julia packages are missing, and have to be installed on per-project basis. That makes it impossible to use them in free projects without internet access.

Suggestion: Install PyPlot and Gadfly for plotting, and SymPy for symbolic math.

PyPlot:

Pkg.add("PyPlot")

When done from a notebook, restart the kernel before proceeding!

To test:

using PyPlot
ion()
plot([1,2,3,4],[3,1,4,2],"-o")

Gadfly:

Pkg.add("Cairo")
Pkg.add("Fontconfig")
Pkg.add("Gadfly")

When done form a notebook, restart the kernel before proceeding!

To test:

using Gadfly
plot(x=[1,2,3,4],y=[3,1,4,2],Geom.point(),Geom.line())

SymPy

Pkg.add("SymPy")

Test

using SymPy
x = symbol("x")
factor(x^2 - 3x + 2)

Number of other useful packages will be installed as dependencies.

Description

Before merging #6284 it makes sense to introduce a stable version of 22.04, which is not the default yet. Also, this adds recent snapshots.

Checklist:

• [ ] Testing instructions are provided, if not obvious
• [ ] Release instructions are provided, if not obvious

"Surrogate pairs are strings that contain a supplemental code point (especially emojis) that cause diff indices to be offset. It can either mess up the text or cause DMP to error (within toDelta/fromDelta)." See

• https://github.com/google/diff-match-patch/issues/10
• https://github.com/google/diff-match-patch/issues/59
• https://github.com/google/diff-match-patch/issues/68

I don't know whether, or to what extent, this might impact cocalc. I've so far never been aware of such an issue. Maybe (?) when CoCalc hits it, an error is thrown, and our diff algorithm generates a very large diff that is just "replace the entire document by this other one", so for us things are not efficient, but not broken either. I don't know. It also might be very unlikely to hit in the context of Jupyter notebooks, where most text is ascii, and markdown where we usually write emojis as :thing: instead of unicode.

In any case, I'll look into this and report back here. Since the original author of DMP doesn't maintain it anymore, it could also make sense to try to modernize the library and make a new independent supported version, which contains fixes for the above issue. As some motivation, the @cocalc/util package has a copy of dmp with at least one important bugfix (from my point of view). That's in https://github.com/sagemathinc/cocalc/blob/master/src/packages/util/dmp.js

NOTE: the file dmp.js had a GPL header applied to it by some automated script that @haraldschilly wrote. However, I just fixed that and reverted the license header back to the original Apache V2 license.

1. Create a chat message
2. Send it
3. Double click on it
4. Edit, then hit control/cmd+z to undo
5. See this

Undo in general might be difficult to implement in exactly this setting. However, it shouldn't show a crash warning!

It doesn't really work and/or is slow and isn't really meaningful since there are so many very different ways to print anyways. Instead use Save and Download As.