Re: [jupyter] Re: How plausible is an emscripten jupyterlab?

2016-11-22 Thread Simon Biggs 
Another point is that ChromeOS has a lot to gain from something like this.
Might it be possible to get some funding from Google to pay the wages of
someone to implement this?

On Wed, 23 Nov 2016, 12:13 PM Simon Biggs  wrote:

As a stepping stone that would still be useful now would it be worth having
a way to have tmpnb read notebook files from the browser's IndexedDB? Or
will even that stepping stone be exceptionally difficult?




On Wed, 23 Nov 2016, 12:01 PM Simon Biggs  wrote:

I guess the answer is it is plausible, but it would be a very large amount
of work.

I don't think I could do such a large amount. If someone does want to work
on it I would love to help.

On Wed, 23 Nov 2016, 10:04 AM Simon Biggs  wrote:

Sorry about the multi post. I think it was an issue with swapping windows
to and from Google groups on Android.

On Wed, 23 Nov 2016, 9:58 AM Simon Biggs  wrote:

Flex looks like it could be amazing. I shall look into that for a project I
am working on. Thanks for the pointer.

With regards to storing the notebook data within the browser I recently
used dexiejs (http://dexie.org) for interacting with IndexedDB. Using Dexie
I was able to write an "import/export" to and from json (
http://electrons.simonbiggs.net/database). (
https://github.com/SimonBiggs/electroninserts-webapp/blob/master/src/app/services/data-services/data-persistence.service.ts
line 278 onwards). That subset of code can be released under a more
permissive license if desired.

With regards to numpy and scipy I completely understand they are a fair way
off. At least until webassembly supports compiling from Fortran.
Nevertheless a pure python static jupyterlab would be perfect for certain
use cases. In my field (health) it can be a bureaucratic nightmare to get
anything installed. Furthermore due to privacy reasons, using an off site
Python server is out if the question. To show something I've written in
python to someone else, and allow them to use it on their computer has
often currently been insurmountable.

As an aside it might also highlight the need for a Fortran to webassembly
compiler quickening the availability of the scientific computing tools
within javascript.

--
You received this message because you are subscribed to a topic in the
Google Groups "Project Jupyter" group.
To unsubscribe from this topic, visit
https://groups.google.com/d/topic/jupyter/5w4A2IOJfxE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to
jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/jupyter/4ce3396c-26e9-4f93-b98d-777589e838ac%40googlegroups.com
.
For more options, visit https://groups.google.com/d/optout.

-- 
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 jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jupyter/CAALqB0DEAs119JECvOjKdWNkoMCXcHD0ZvOzEd5k9VFukAs6JQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [jupyter] Re: How plausible is an emscripten jupyterlab?

2016-11-22 Thread Simon Biggs 
As a stepping stone that would still be useful now would it be worth having
a way to have tmpnb read notebook files from the browser's IndexedDB? Or
will even that stepping stone be exceptionally difficult?




On Wed, 23 Nov 2016, 12:01 PM Simon Biggs  wrote:

I guess the answer is it is plausible, but it would be a very large amount
of work.

I don't think I could do such a large amount. If someone does want to work
on it I would love to help.

On Wed, 23 Nov 2016, 10:04 AM Simon Biggs  wrote:

Sorry about the multi post. I think it was an issue with swapping windows
to and from Google groups on Android.

On Wed, 23 Nov 2016, 9:58 AM Simon Biggs  wrote:

Flex looks like it could be amazing. I shall look into that for a project I
am working on. Thanks for the pointer.

With regards to storing the notebook data within the browser I recently
used dexiejs (http://dexie.org) for interacting with IndexedDB. Using Dexie
I was able to write an "import/export" to and from json (
http://electrons.simonbiggs.net/database). (
https://github.com/SimonBiggs/electroninserts-webapp/blob/master/src/app/services/data-services/data-persistence.service.ts
line 278 onwards). That subset of code can be released under a more
permissive license if desired.

With regards to numpy and scipy I completely understand they are a fair way
off. At least until webassembly supports compiling from Fortran.
Nevertheless a pure python static jupyterlab would be perfect for certain
use cases. In my field (health) it can be a bureaucratic nightmare to get
anything installed. Furthermore due to privacy reasons, using an off site
Python server is out if the question. To show something I've written in
python to someone else, and allow them to use it on their computer has
often currently been insurmountable.

As an aside it might also highlight the need for a Fortran to webassembly
compiler quickening the availability of the scientific computing tools
within javascript.

--
You received this message because you are subscribed to a topic in the
Google Groups "Project Jupyter" group.
To unsubscribe from this topic, visit
https://groups.google.com/d/topic/jupyter/5w4A2IOJfxE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to
jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/jupyter/4ce3396c-26e9-4f93-b98d-777589e838ac%40googlegroups.com
.
For more options, visit https://groups.google.com/d/optout.

-- 
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 jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jupyter/CAALqB0DXJv1Cp7cGeSyeL9YuYOCCNPW-bJQpCuMPO2XwUKQKjQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [jupyter] Re: How plausible is an emscripten jupyterlab?

2016-11-22 Thread Simon Biggs 
I guess the answer is it is plausible, but it would be a very large amount
of work.

I don't think I could do such a large amount. If someone does want to work
on it I would love to help.

On Wed, 23 Nov 2016, 10:04 AM Simon Biggs  wrote:

> Sorry about the multi post. I think it was an issue with swapping windows
> to and from Google groups on Android.
>
> On Wed, 23 Nov 2016, 9:58 AM Simon Biggs  wrote:
>
> Flex looks like it could be amazing. I shall look into that for a project
> I am working on. Thanks for the pointer.
>
> With regards to storing the notebook data within the browser I recently
> used dexiejs (http://dexie.org) for interacting with IndexedDB. Using
> Dexie I was able to write an "import/export" to and from json (
> http://electrons.simonbiggs.net/database). (
> https://github.com/SimonBiggs/electroninserts-webapp/blob/master/src/app/services/data-services/data-persistence.service.ts
> line 278 onwards). That subset of code can be released under a more
> permissive license if desired.
>
> With regards to numpy and scipy I completely understand they are a fair
> way off. At least until webassembly supports compiling from Fortran.
> Nevertheless a pure python static jupyterlab would be perfect for certain
> use cases. In my field (health) it can be a bureaucratic nightmare to get
> anything installed. Furthermore due to privacy reasons, using an off site
> Python server is out if the question. To show something I've written in
> python to someone else, and allow them to use it on their computer has
> often currently been insurmountable.
>
> As an aside it might also highlight the need for a Fortran to webassembly
> compiler quickening the availability of the scientific computing tools
> within javascript.
>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "Project Jupyter" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/jupyter/5w4A2IOJfxE/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> jupyter+unsubscr...@googlegroups.com.
> To post to this group, send email to jupyter@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/jupyter/4ce3396c-26e9-4f93-b98d-777589e838ac%40googlegroups.com
> .
> For more options, visit https://groups.google.com/d/optout.
>
>

-- 
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 jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jupyter/CAALqB0Bg0-E2EhsSBvtpci2JVYZM0Cf39nnNRf2JXy9hdN%3D0tA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [jupyter] Re: How plausible is an emscripten jupyterlab?

2016-11-22 Thread Simon Biggs 
Sorry about the multi post. I think it was an issue with swapping windows
to and from Google groups on Android.

On Wed, 23 Nov 2016, 9:58 AM Simon Biggs  wrote:

> Flex looks like it could be amazing. I shall look into that for a project
> I am working on. Thanks for the pointer.
>
> With regards to storing the notebook data within the browser I recently
> used dexiejs (http://dexie.org) for interacting with IndexedDB. Using
> Dexie I was able to write an "import/export" to and from json (
> http://electrons.simonbiggs.net/database). (
> https://github.com/SimonBiggs/electroninserts-webapp/blob/master/src/app/services/data-services/data-persistence.service.ts
> line 278 onwards). That subset of code can be released under a more
> permissive license if desired.
>
> With regards to numpy and scipy I completely understand they are a fair
> way off. At least until webassembly supports compiling from Fortran.
> Nevertheless a pure python static jupyterlab would be perfect for certain
> use cases. In my field (health) it can be a bureaucratic nightmare to get
> anything installed. Furthermore due to privacy reasons, using an off site
> Python server is out if the question. To show something I've written in
> python to someone else, and allow them to use it on their computer has
> often currently been insurmountable.
>
> As an aside it might also highlight the need for a Fortran to webassembly
> compiler quickening the availability of the scientific computing tools
> within javascript.
>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "Project Jupyter" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/jupyter/5w4A2IOJfxE/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> jupyter+unsubscr...@googlegroups.com.
> To post to this group, send email to jupyter@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/jupyter/4ce3396c-26e9-4f93-b98d-777589e838ac%40googlegroups.com
> .
> For more options, visit https://groups.google.com/d/optout.
>

-- 
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 jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jupyter/CAALqB0DTedR9S1fW6OG3g8DDTLSb3e9ob25dhjU9seU0hE%3DYjw%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [jupyter] Re: How plausible is an emscripten jupyterlab?

2016-11-22 Thread Simon Biggs 
Flex looks like it could be amazing. I shall look into that for a project I am 
working on. Thanks for the pointer.

With regards to storing the notebook data within the browser I recently used 
dexiejs (http://dexie.org) for interacting with IndexedDB. Using Dexie I was 
able to write an "import/export" to and from json 
(http://electrons.simonbiggs.net/database). 
(https://github.com/SimonBiggs/electroninserts-webapp/blob/master/src/app/services/data-services/data-persistence.service.ts
 line 278 onwards). That subset of code can be released under a more permissive 
license if desired.

With regards to numpy and scipy I completely understand they are a fair way 
off. At least until webassembly supports compiling from Fortran. Nevertheless a 
pure python static jupyterlab would be perfect for certain use cases. In my 
field (health) it can be a bureaucratic nightmare to get anything installed. 
Furthermore due to privacy reasons, using an off site Python server is out if 
the question. To show something I've written in python to someone else, and 
allow them to use it on their computer has often currently been insurmountable. 

As an aside it might also highlight the need for a Fortran to webassembly 
compiler quickening the availability of the scientific computing tools within 
javascript.

-- 
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 jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jupyter/4ce3396c-26e9-4f93-b98d-777589e838ac%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [jupyter] Re: How plausible is an emscripten jupyterlab?

2016-11-22 Thread Simon Biggs 
Flex looks like it could be amazing. I shall look into that for a project I am 
working on. Thanks for the pointer.

With regards to storing the notebook data within the browser I recently used 
dexiejs (http://dexie.org) for interacting with IndexedDB. Using Dexie I was 
able to write an "import/export" to and from json 
(http://electrons.simonbiggs.net/database). 
(https://github.com/SimonBiggs/electroninserts-webapp/blob/master/src/app/services/data-services/data-persistence.service.ts
 line 278 onwards). That subset of code can be released under a more permissive 
license if desired.

With regards to numpy and scipy I completely understand they are a fair way 
off. At least until webassembly supports compiling from Fortran. Nevertheless a 
pure python static jupyterlab would be perfect for certain use cases. In my 
field (health) it can be a bureaucratic nightmare to get anything installed. 
Furthermore due to privacy reasons, using an off site Python server is out if 
the question. To show something I've written in python to someone else, and 
allow them to use it on their computer has often currently been insurmountable. 

As an aside it might also highlight the need for a Fortran to webassembly 
compiler quickening the availability of the scientific computing tools within 
javascript.

-- 
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 jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jupyter/ebcae057-10ad-48b1-9610-5d6f9acd07e8%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [jupyter] Re: How plausible is an emscripten jupyterlab?

2016-11-22 Thread Simon Biggs 
Flex looks like it could be amazing. I shall look into that for a project I am 
working on. Thanks for the pointer.

With regards to storing the notebook data within the browser I recently used 
dexiejs (http://dexie.org) for interacting with IndexedDB. Using Dexie I was 
able to write an "import/export" to and from json 
(http://electrons.simonbiggs.net/database). 
(https://github.com/SimonBiggs/electroninserts-webapp/blob/master/src/app/services/data-services/data-persistence.service.ts
 line 278 onwards). That subset of code can be released under a more permissive 
license if desired.

With regards to numpy and scipy I completely understand they are a fair way 
off. At least until webassembly supports compiling from Fortran. Nevertheless a 
pure python static jupyterlab would be perfect for certain use cases. In my 
field (health) it can be a bureaucratic nightmare to get anything installed. 
Furthermore due to privacy reasons, using an off site Python server is out if 
the question. To show something I've written in python to someone else, and 
allow them to use it on their computer has often currently been insurmountable. 

As an aside it might also highlight the need for a Fortran to webassembly 
compiler quickening the availability of the scientific computing tools within 
javascript.

-- 
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 jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jupyter/36d601d9-4a0b-4fdc-9837-aba8eaecb12e%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [jupyter] Re: How plausible is an emscripten jupyterlab?

2016-11-22 Thread Ian Rose 
It should be possible to get a static version of JupyerLab with CPython
compiled with emscripten (with this setup you should also be able to get a
Javascript kernel up and running, I think).
However, I think that if you wanted any of the scientific stack (numpy,
scipy, etc) you will have a much more difficult time of it. These all rely
extensively on compiled C and Fortran code, and have pretty complicated
build processes. All of these compiled modules would also need to be passed
through emscripten.

Ian

On Tue, Nov 22, 2016 at 6:58 AM, Thomas Kluyver  wrote:

> I guess that Simon was thinking of including an emscripten-ed cpython to
> run the user's code in the notebook. I think it's quite plausible, if quite
> a bit of work, to do that, and it would be a really fascinating tool for
> certain situations, like teaching basic programming. Turning it into a
> totally static webpage would probably require some more work, to save
> notebooks in local storage instead of on the server.
>
>
> On 22 November 2016 at 14:46, Steven Silvester 
> wrote:
>
>> Hi Simon,
>>
>> An individual plugin could conceivably be written in CPython and compiled
>> to JavaScript using emscripten.  Almar Klein is working on a CPython ->
>> JupyterLab plugin generator right now using the pyscript compiler in Flexx
>> (not yet published).   https://github.com/zoofIO/flexx.
>>
>>
>> Regards,
>>
>> Steve
>>
>>
>> On Sunday, November 20, 2016 at 6:50:25 PM UTC-6, Simon Biggs wrote:
>>>
>>> Supposedly http://repl.it was able to use emscripten to compile CPython
>>> to javascript. How plausible might it be to have the base jupyterlab also
>>> compiled and have a minimal working version of jupyterlab hosted as a
>>> static web page? Am I just talking nonsense, or does the fact the
>>> webassembly is a fast approaching standard make this a project worth
>>> investigating?
>>
>> --
>> 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 jupyter+unsubscr...@googlegroups.com.
>> To post to this group, send email to jupyter@googlegroups.com.
>> To view this discussion on the web visit https://groups.google.com/d/ms
>> gid/jupyter/d3a82f6f-9348-4262-9d22-ee99a935614d%40googlegroups.com
>> 
>> .
>>
>> For more options, visit https://groups.google.com/d/optout.
>>
>
> --
> 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 jupyter+unsubscr...@googlegroups.com.
> To post to this group, send email to jupyter@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/
> msgid/jupyter/CAOvn4qi0NUSGgbdoyNqg2MMpzKuf75q0had80EEA2dxOVQFWmw%40mail.
> gmail.com
> 
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
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 jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jupyter/CAEm%3D8TSu07knwW-8Qcr1utCJahLrFf5U5ot0K%2BZ14DDnNpnxKg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [jupyter] Re: How plausible is an emscripten jupyterlab?

2016-11-22 Thread Thomas Kluyver 
I guess that Simon was thinking of including an emscripten-ed cpython to
run the user's code in the notebook. I think it's quite plausible, if quite
a bit of work, to do that, and it would be a really fascinating tool for
certain situations, like teaching basic programming. Turning it into a
totally static webpage would probably require some more work, to save
notebooks in local storage instead of on the server.

On 22 November 2016 at 14:46, Steven Silvester 
wrote:

> Hi Simon,
>
> An individual plugin could conceivably be written in CPython and compiled
> to JavaScript using emscripten.  Almar Klein is working on a CPython ->
> JupyterLab plugin generator right now using the pyscript compiler in Flexx
> (not yet published).   https://github.com/zoofIO/flexx.
>
>
> Regards,
>
> Steve
>
>
> On Sunday, November 20, 2016 at 6:50:25 PM UTC-6, Simon Biggs wrote:
>>
>> Supposedly http://repl.it was able to use emscripten to compile CPython
>> to javascript. How plausible might it be to have the base jupyterlab also
>> compiled and have a minimal working version of jupyterlab hosted as a
>> static web page? Am I just talking nonsense, or does the fact the
>> webassembly is a fast approaching standard make this a project worth
>> investigating?
>
> --
> 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 jupyter+unsubscr...@googlegroups.com.
> To post to this group, send email to jupyter@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/ms
> gid/jupyter/d3a82f6f-9348-4262-9d22-ee99a935614d%40googlegroups.com
> 
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
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 jupyter+unsubscr...@googlegroups.com.
To post to this group, send email to jupyter@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jupyter/CAOvn4qi0NUSGgbdoyNqg2MMpzKuf75q0had80EEA2dxOVQFWmw%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.