Extending sagecell to work with all jupyter widgets

[Kwankyu Lee]

Hi Andrey,

It would be nice if sagecell could work with arbitrary jupyter widgets. I have some questions.

(1) Do you think that this is technically possible at all?

(2) Is someone already working toward this goal? 

(3) Or conversely, are you aware of any project in jupyter side developing a sagecell-like server software that works with all jupyter widgets?

Thank you for your attention and reply in advance!

Kwankyu

[Andrey Novoseltsev]

On Friday, 30 November 2018 01:17:58 UTC-7, Kwankyu Lee wrote:

Hi Andrey,

It would be nice if sagecell could work with arbitrary jupyter widgets. I have some questions.

(1) Do you think that this is technically possible at all?

It is definitely possible as SageMathCell builds on IPython kernels, mostly it is just about passing messages between the JS client side and Python server kernels. Done "the right way" it may be even simpler than the current framework. But I don't know this right way yet, so it is not something quick for me.

(2) Is someone already working toward this goal? 

As fas as I know nobody is working on any part of SageMathCell except for me. I do have this as an important long term goal but hesitate to give any completion schedule. There are still some things regarding kernel management that I want to implement first. Would you like to volunteer to work on widgets?-)

(3) Or conversely, are you aware of any project in jupyter side developing a sagecell-like server software that works with all jupyter widgets?

I am not and I guess the answer depends on definition of "sagecell-like" ;-) SageNB and CoCalc are also web-interfaces to Sage.

[Kwankyu Lee]

It would be nice if sagecell could work with arbitrary jupyter widgets. I have some questions.

(1) Do you think that this is technically possible at all?

It is definitely possible as SageMathCell builds on IPython kernels, mostly it is just about passing messages between the JS client side and Python server kernels. Done "the right way" it may be even simpler than the current framework. But I don't know this right way yet, so it is not something quick for me.

I see.

(2) Is someone already working toward this goal? 

As fas as I know nobody is working on any part of SageMathCell except for me. I do have this as an important long term goal but hesitate to give any completion schedule. There are still some things regarding kernel management that I want to implement first. Would you like to volunteer to work on widgets?-)

I don't know the right way, like you, and worse still I don't understand sagemathcell, unlike you :-)

I would set as my own long term goal to understand both sides, and could help you someday. Currently I cannot afford adequate time to do any significant work on that.

 

(3) Or conversely, are you aware of any project in jupyter side developing a sagecell-like server software that works with all jupyter widgets?

I am not and I guess the answer depends on definition of "sagecell-like" ;-) SageNB and CoCalc are also web-interfaces to Sage.

I meant a single cell server to embed sage code, interacts, and jupyter widgets into any web page.

As far as I know, presently Jason Grout seems the only person that has expertise on all things for this mission...

Thanks! 

Kwankyu

[Jason Grout]

Min Ragan-Kelly (one of the original developers in the Jupyter/IPython community) built ThebeLab, which has some similar features to the Sage cell server and supports ipywidgets: https://github.com/minrk/thebelab

Examples: https://minrk.github.io/thebelab/

Widgets support: https://minrk.github.io/thebelab/widgets.html

Thanks, (and thanks again to Andrey for the incredible amount of time and expertise he brings to the sage cell server for many years!)

Jason

--
You received this message because you are subscribed to the Google Groups "sage-cell" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-cell+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-cell/ff616ad0-ece9-47dc-b49d-7db9f7cbc0bf%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Jason Grout]

Some more information about thebe vs sage cell server: https://sage-package.readthedocs.io/en/latest/sage_package/thebe.html#thebe-versus-sagemath-cell

Jason

[Kwankyu Lee]

On Saturday, December 1, 2018 at 10:21:47 AM UTC+9, Jason Grout wrote:

Some more information about thebe vs sage cell server: https://sage-package.readthedocs.io/en/latest/sage_package/thebe.html#thebe-versus-sagemath-cell

So now the mission is to combine the two forces of Thebe and SMC :-)

Many thanks to Andrey and Jason!

Kwankyu

[Andrey Novoseltsev]

On Friday, 30 November 2018 18:34:58 UTC-7, Kwankyu Lee wrote:

So now the mission is to combine the two forces of Thebe and SMC :-)

My hope was rather to reuse the code in Sage for Jupyter interacts - I think "all that has to be done" is to route communication through SageMathCell proxies.

Thanks for the thank yous! Often it would be easy to think that nobody really cares, if it were not for the usage stat ;-)

Andrey

[Jason Grout]

Just curious, what are the usage stats these days?

Jason

--

You received this message because you are subscribed to the Google Groups "sage-cell" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-cell+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-cell/dfc44a8a-fef0-415b-b7ff-a1f6e7ff96ba%40googlegroups.com.

[Kwankyu Lee]

 Often it would be easy to think that nobody really cares, if it were not for the usage stat ;-)

Also there are people running their own sagecell servers. I am one of them.

[Andrey Novoseltsev]

On Friday, 30 November 2018 22:16:37 UTC-7, Jason Grout wrote:

Just curious, what are the usage stats these days?

40-50k HAProxy sessions to compute kernels per day. 87 active sessions right now (Sunday), well over a hundred regularly and occasional higher spikes. 2 servers with 4 cores each and 22/32GB RAM are doing more or less OK.

[Andrey Novoseltsev]

On Friday, 30 November 2018 22:25:42 UTC-7, Kwankyu Lee wrote:

Also there are people running their own sagecell servers. I am one of them.

Indeed, although I have no clue how many are out there besides you. When I was running my own server Jason heard complaints quite regularly. I don't get much, but straightening and automatizing deployment was the first thing I tried to do. Either I have succeeded and it works for those who try, or failed and they give up completely ;-)