Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

 

>
> For what it is worth, SageMathCloud has a buttons/lists, etc. for 
> editing markdown in Sage worksheets, and also a realtime preview 
> markdown editor for md files.  Two screenshots attached.  I also fully 
> implemented a realtime WYSIWYG editor for html and markdown a year 
> ago, but decided it wasn't up to my standards (getting realtime sync 
> to fully work well was surprisingly challenging), so I 
> disabled/removed it.   
>

Oh yeah, I forgot that I get very confused trying to make text cells in SMC 
too :(  presumably for the same reason that I don't want to type %md - but 
I don't use SMC much (yet).  I'm surprised that there isn't more stuff out 
there trying to do fully-featured wysiwyg md editing, actually, obviously 
it is completely orthogonal to the math and of very general interest.  Huh.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

> Oh yeah, I forgot that I get very confused trying to make text cells in 
> SMC too :(  presumably for the same reason that I don't want to type %md - 
> but I don't use SMC much (yet). 
>

(I.e. this might be awesome after all but I haven't gotten that far!) 

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jason Grout]

I should add that one of the major design goals for the next iteration of
the notebook is to be fully modular, so it would be easy to plug in your
own component to edit markdown, for example.

Thanks,

Jason

On Tue, Jan 26, 2016 at 10:59 PM Jason Grout  wrote:

> What if we used something like ProseMirror for the markdown editor?
> http://prosemirror.net/.  Would that help? Right now we use Codemirror.
>
> I'm writing a prototype for the next Jupyter notebook renderer as we speak
> (markdown cells are here, for example:
> https://github.com/jupyter/jupyter-js-cells/blob/master/src/widget.ts#L110).
> Pull requests welcome, as always :).
>
> Jason
>
>
> On Tue, Jan 26, 2016 at 10:55 PM kcrisman  wrote:
>
>> Another con I just discovered:
>> * You have to learn markdown to do anything useful in plain old text.
>>
>> Don't tell me this isn't a con.  (If it's not accurate, please tell me!
>> I just couldn't figure out how to get
>>
>> Now, I know enough md to get by.  Lots of people use it.  Lots of *other*
>> people (see, I used it!) would rather have at least SOME whizzy-wig
>> capability.  'Cuz why else does the interface I'm using right now in Google
>> Groups have things *like this or **this* or even bullet lists to click
>> (perhaps they use TinyMCE themselves)?  It should be just as much about
>> reducing learning curves as the "right" solution.  I hate having to
>> remember if links are [like this](url) or (this)[url] or even [url like
>> this] (oh wait, that's the Trac style).  Google lets me do this
>>  with a simple click.
>>
>> To be productive on this front and not just complain, I did a fair amount
>> of searching for wysiwyg or tinymce and jupyter and found almost nothing.
>> Could this be a replacement?  https://github.com/bollwyvl/nb-wysiwyg  I
>> also found this nice article
>> 
>> which (correctly) claims "Why is Markdown better? Well, it’s worth saying
>> that maybe it isn't. Mainly, it’s not actually a question of better or
>> worse, but of what’s in front of you and of who you are. A definitive
>> answer depends on the user and on that user’s goals and experience. These
>> Notebooks don't use Markdown because it's definitely better, but rather
>> because it's different and thus encourages users to think about their work
>> differently."  But not everyone, especially those instructors in a hurry,
>> have time to think about that on a first try.  If they end up writing a
>> book I hope they do!  But if they just want to make an example for class
>> it's a bit much.
>>
>> Hopefully Jupyter will be able to have an option to have wysiwyg
>> eventually, though I understand that might conflict with their design
>> goals. In which case their design goals are not really for
>> non-programmers.
>>
>> Practical example, lest someone think I'm beating up on a straw notebook
>> interface:
>> Someone makes an awesome 3d plot in Jupyter with vectors and parametric
>> things in red, blue, and green, labeling different things.  Now in the main
>> body of their text they want the same output, so they can talk about green
>> tangent vectors, blue normal vectors, and red curves, or something, in
>> those colors.  Lovely stuff.  They Google how to do this in md and get:
>>
>> http://stackoverflow.com/questions/19746350/how-does-one-change-color-in-markdown-cells-ipython-notebook
>> Result: the text stays all black for the presentation they have to do in
>> ten minutes.
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "sage-devel" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to sage-devel+unsubscr...@googlegroups.com.
>> To post to this group, send email to sage-devel@googlegroups.com.
>> Visit this group at https://groups.google.com/group/sage-devel.
>> For more options, visit https://groups.google.com/d/optout.
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jason Grout]

What if we used something like ProseMirror for the markdown editor?
http://prosemirror.net/.  Would that help? Right now we use Codemirror.

I'm writing a prototype for the next Jupyter notebook renderer as we speak
(markdown cells are here, for example:
https://github.com/jupyter/jupyter-js-cells/blob/master/src/widget.ts#L110).
Pull requests welcome, as always :).

Jason


On Tue, Jan 26, 2016 at 10:55 PM kcrisman  wrote:

> Another con I just discovered:
> * You have to learn markdown to do anything useful in plain old text.
>
> Don't tell me this isn't a con.  (If it's not accurate, please tell me!  I
> just couldn't figure out how to get
>
> Now, I know enough md to get by.  Lots of people use it.  Lots of *other*
> people (see, I used it!) would rather have at least SOME whizzy-wig
> capability.  'Cuz why else does the interface I'm using right now in Google
> Groups have things *like this or **this* or even bullet lists to click
> (perhaps they use TinyMCE themselves)?  It should be just as much about
> reducing learning curves as the "right" solution.  I hate having to
> remember if links are [like this](url) or (this)[url] or even [url like
> this] (oh wait, that's the Trac style).  Google lets me do this
>  with a simple click.
>
> To be productive on this front and not just complain, I did a fair amount
> of searching for wysiwyg or tinymce and jupyter and found almost nothing.
> Could this be a replacement?  https://github.com/bollwyvl/nb-wysiwyg  I
> also found this nice article
> 
> which (correctly) claims "Why is Markdown better? Well, it’s worth saying
> that maybe it isn't. Mainly, it’s not actually a question of better or
> worse, but of what’s in front of you and of who you are. A definitive
> answer depends on the user and on that user’s goals and experience. These
> Notebooks don't use Markdown because it's definitely better, but rather
> because it's different and thus encourages users to think about their work
> differently."  But not everyone, especially those instructors in a hurry,
> have time to think about that on a first try.  If they end up writing a
> book I hope they do!  But if they just want to make an example for class
> it's a bit much.
>
> Hopefully Jupyter will be able to have an option to have wysiwyg
> eventually, though I understand that might conflict with their design
> goals. In which case their design goals are not really for
> non-programmers.
>
> Practical example, lest someone think I'm beating up on a straw notebook
> interface:
> Someone makes an awesome 3d plot in Jupyter with vectors and parametric
> things in red, blue, and green, labeling different things.  Now in the main
> body of their text they want the same output, so they can talk about green
> tangent vectors, blue normal vectors, and red curves, or something, in
> those colors.  Lovely stuff.  They Google how to do this in md and get:
>
> http://stackoverflow.com/questions/19746350/how-does-one-change-color-in-markdown-cells-ipython-notebook
> Result: the text stays all black for the presentation they have to do in
> ten minutes.
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

Jason, thanks for your quick replies, very informative.

What if we used something like ProseMirror for the markdown editor? 
>> http://prosemirror.net/ 
>> .
>>   
>> Would that help? Right now we use Codemirror.
>>
>>
This seems like a step in the right direction.  We don't need "flawless" 
copying of TinyMCE, for instance, but no ability to do wysiwyg in the text 
cells would definitely be unfortunate for that market segment, as it were.  

This?  Even finding things between TinyMCE and md is 
http://leeoniya.github.io/reMarked.js/ 
 

> I'm writing a prototype for the next Jupyter notebook renderer as we speak 
>> (markdown cells are here, for example: 
>> https://github.com/jupyter/jupyter-js-cells/blob/master/src/widget.ts#L110). 
>>  
>> Pull requests welcome, as always :).
>>
>> Jason
>>
>>>
>>>
 

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[kcrisman]

Another con I just discovered: 
* You have to learn markdown to do anything useful in plain old text.

Don't tell me this isn't a con.  (If it's not accurate, please tell me!  I 
just couldn't figure out how to get 

Now, I know enough md to get by.  Lots of people use it.  Lots of *other* 
people (see, I used it!) would rather have at least SOME whizzy-wig 
capability.  'Cuz why else does the interface I'm using right now in Google 
Groups have things *like this or **this* or even bullet lists to click 
(perhaps they use TinyMCE themselves)?  It should be just as much about 
reducing learning curves as the "right" solution.  I hate having to 
remember if links are [like this](url) or (this)[url] or even [url like 
this] (oh wait, that's the Trac style).  Google lets me do this 
 with a simple click.

To be productive on this front and not just complain, I did a fair amount 
of searching for wysiwyg or tinymce and jupyter and found almost nothing.  
Could this be a replacement?  https://github.com/bollwyvl/nb-wysiwyg  I 
also found this nice article 

 
which (correctly) claims "Why is Markdown better? Well, it’s worth saying 
that maybe it isn't. Mainly, it’s not actually a question of better or 
worse, but of what’s in front of you and of who you are. A definitive 
answer depends on the user and on that user’s goals and experience. These 
Notebooks don't use Markdown because it's definitely better, but rather 
because it's different and thus encourages users to think about their work 
differently."  But not everyone, especially those instructors in a hurry, 
have time to think about that on a first try.  If they end up writing a 
book I hope they do!  But if they just want to make an example for class 
it's a bit much.

Hopefully Jupyter will be able to have an option to have wysiwyg 
eventually, though I understand that might conflict with their design 
goals. In which case their design goals are not really for 
non-programmers.  

Practical example, lest someone think I'm beating up on a straw notebook 
interface:
Someone makes an awesome 3d plot in Jupyter with vectors and parametric 
things in red, blue, and green, labeling different things.  Now in the main 
body of their text they want the same output, so they can talk about green 
tangent vectors, blue normal vectors, and red curves, or something, in 
those colors.  Lovely stuff.  They Google how to do this in md and get:
http://stackoverflow.com/questions/19746350/how-does-one-change-color-in-markdown-cells-ipython-notebook
Result: the text stays all black for the presentation they have to do in 
ten minutes.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Henri Girard]

%%maxima doesn't work but %maxima does 

Le vendredi 18 décembre 2015 11:35:53 UTC+1, Jeroen Demeyer a écrit :
>
> Should the Jupyter notebook be the default notebook for the next Sage 
> 7.0 release? I don't really have an opinion on the matter. 
>
>
> Pros: 
>
> * Nice tracebacks! 
> * The Jupyter notebook is a mature well-maintained project, unlike 
> SageNB. It is widely used in the "scientific Python" community. 
> * Availability of other Jupyter kernels besides Sage. 
>
>
> Cons: 
>
> * Less compatible with Sage: Sage interacts don't work, some graphics 
> don't work. 
> * Certain features of SageNB are missing: live documentation, 
> sharing/publishing of worksheets. 
> * It clutters the file system with .ipynb files. 
>
>
> Don't cares: 
>
> * It's just a default choice, both options remain available. 
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

Fixing the cell magic situation is now 
http://trac.sagemath.org/ticket/19851 (needs review)

On Sunday, January 3, 2016 at 4:33:36 PM UTC+1, Volker Braun wrote:
>
> Interactive line magics obviously can't work in the browser like on the 
> command line
>
> The lack of appropriate cell magics (%%maxima)  is just an existing bug on 
> the commandline.
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [IPython-dev] [sage-devel] Re: Jupyter notebook by default?

[MinRK]

On Sat, Jan 9, 2016 at 8:04 AM, kcrisman  wrote:

> These all sound like awesome ideas to me for moving more *robustly* to a
> sws -> ipynb transition!  Would attached files work properly (or at all)?
> I'm only asking out of ignorance.
>

No, attached files would likely be placed next to the notebook.

-MinRK


> --
> You received this message because you are subscribed to a topic in the
> Google Groups "sage-devel" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/sage-devel/8erxWppKxXM/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

Screenshot:




On Saturday, January 9, 2016 at 1:39:23 PM UTC+1, Volker Braun wrote:
>
> Fixing the cell magic situation is now 
> http://trac.sagemath.org/ticket/19851 (needs review)
>
> On Sunday, January 3, 2016 at 4:33:36 PM UTC+1, Volker Braun wrote:
>>
>> Interactive line magics obviously can't work in the browser like on the 
>> command line
>>
>> The lack of appropriate cell magics (%%maxima)  is just an existing bug 
>> on the commandline.
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [IPython-dev] [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

These all sound like awesome ideas to me for moving more *robustly* to a 
sws -> ipynb transition!  Would attached files work properly (or at all)? 
 I'm only asking out of ignorance.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [IPython-dev] [sage-devel] Re: Jupyter notebook by default?

[MinRK]

Thanks Jason for cross-posting.

Since the issue of funding was brought up, I think supporting projects like
this is exactly the sort of thing we should be doing with the funding we
have, whether the work sits on the Jupyter or Sage side (I assume there
will be both). It’s a bit tricky to keep track of all the points in an
email thread, but if we could aggregate the things that are blockers and
the things that would be nice, especially changes you need from Jupyter, we
should be able to start ticking boxes.

A summary of what I’ve seen so far:

   - sage interacts
   - language cells
   - document conversion from sagenb to ipynb
   - low-level output capturing
   - gracefully handling large output

Some comments:

Re: language cells, I assume it’s referring to things like %%bash, %%R, and
%%cython. While these *look* similar, there is a significant difference in
how they are implemented. For instance, the R magic (provided by rpy2) runs
an R interpreter in-memory, and talks to it, capturing output, etc.. Where
many of these magics, such as bash, ruby, perl, come from is some “script
magic” machinery in IPython, which populates the default magics with
shortcuts to running a script in a given interpreter. They are essentially
shortcuts to cat  | . It’s not a fundamental limitation,
or anything dire like that. If sage has an implementation of running code
in a persistent alternate interpreter, then it should not be much work to
represent that in magics, since cell magics are any Python functions called
with two string arguments (the rest of the line and the cell), and can be
defined at any time, for instance:

def mymagic(line, cell):
do_stuff_with(cell)

get_ipython().register_magic_function(mymagic, 'cell')

Re: output capturing, Thomas Kluyver and I were at CERN last month working
on the Cling kernel, and one of the things we did was C-level capturing of
output. Now that we have that working, integrating it into the IPython
kernel should not be much work, and if it’s really important, libraries can
use the same technique themselves without waiting for IPython to catch up.

Interacts are perhaps the hardest piece. I think it should be doable to get
sage’s own interacts working in the notebook, rather than forcing people to
adopt the much more basic interact provided by the IPython widgets.

I can’t speak to the UI transition part of the problem whenever you change
defaults, which is a big challenge, but I think we can at least mitigate
most of the things on the Jupyter side that are getting in your way.

-MinRK
​

On Tue, Jan 5, 2016 at 8:19 PM, Jason Grout <ja...@jasongrout.org> wrote:

> (cross-posting to ipython-dev)
>
> Jon,
>
> At the recent San Francisco meetings, we talked about this.  What do you
> think about:
>
> 1. keeping track of the size of the io messages sent from any specific
> kernel execution
> 2. When the total size of io reaches some specific size
> (user-configurable), transmitting a special "throwing away output, but
> here's how to save the output to a file if you want in the future, or how
> to increase the limit" message
> 3. keep a running buffer of the last bit of output attempted to be sent,
> and send it when the execution finishes (so basically a ring buffer that
> overwrites the oldest message)
>
> This:
>
> * allows small output through
> * provides an explanatory message
> * provides the last bit of output as well
>
> One thing to figure out: a limit on size of output that is text may not be
> appropriate for output that is images, etc.
>
> Thanks,
>
> Jason
>
>
> On Tue, Jan 5, 2016 at 12:11 PM, Jason Grout <ja...@jasongrout.org> wrote:
>
>>
>> ------ Forwarded message ------
>> From: Jonathan Frederic <jon.fre...@gmail.com>
>> Date: Tue, Jan 5, 2016 at 11:42 AM
>> Subject: Re: [sage-devel] Re: Jupyter notebook by default?
>> To: Jason Grout <grout.ja...@gmail.com>
>> Cc: sage-devel <sage-devel@googlegroups.com>
>>
>>
>> Jason,
>>
>> Thanks for pulling me in on this.
>>
>> William,
>>
>> I agree, getting a bunch of people to agree on stuff can seem
>> impossible.  However, you mention Sage offers a couple options to mitigate
>> output overflows, can you point me to those options?  The Jupyter Notebook
>> should provide multiple options too - this will also make it easier for
>> everyone to agree.
>>
>> Also, in you experience, which of these options work the best?
>>
>> I was thinking initially of doing something simple, like hard limiting
>> data/time, then printing an error if that's exceeded.  In the Jupyter
>> Notebook, we have to worry about
>> - Too many messages sent on the websocket
>> - The notebook json file growing too large and consequentl

Re: [IPython-dev] [sage-devel] Re: Jupyter notebook by default?

[Thierry]

Hi,

On Thu, Jan 07, 2016 at 11:48:05AM +0100, MinRK wrote:
[...]
>- document conversion from sagenb to ipynb

We should probably better focus on a good rst2ipynb translator.

Indeed, there is a "sage -sws2rst" command already. This will have the
benefit to be helpful for more people than just Sage users, and help us
going towards Jupyter live documentation i.e. ipynb files connected
together with a bit of sphinx (i dont' know if it is possible to add a
navigation bar to ipynb worksheets, or use frames, or whatever ?).

Moreover, all the developpers i met during tutorials and schools write
their tutorials in .rst, simply because it is human writeable, it is easy
to copy-paste from existing Sage doctests and tutorials, and can be easily
versionned/shared/imported.

Though json is far better than tar.bz or xml from the human point of view,
it is still too verbose to be considered as a source format for
worksheets, moreover from my little experience, people who write tutorials
prefer to use an editor/IDE/whatever than a webpage.

With pandoc and notedown, it is easy to do basic translation rst -> md ->
ipynb. The main remaining task is to let the code blocks clever, that is:

::
sage: a = 1
sage: 2*a
2
sage: 1+1
2

should be translated in ipynb as (simplified)

cell:
type: code
language: sage
input:
a = 1
2*a
output:
2
cell:
type: code
language: sage
input:
1+1
output:
2

which is not the case when piping pandoc and notedown (we obtain a single
input block, and the `sage:` are not removed).

Another separate hook could deal with Sage specific tweaks (e.g. `blah`
for :math:`blah`)

Ciao,
Thierry

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

IMHO output capture into a web browser isn't really different from the 
scrollback buffer of a terminal. We obviously enjoy the infinite scrollback 
but do not want an unbounded drawing surface in the terminal (= dom nodes 
in the web browser). And certainly nobody wants a piece of their output 
discarded in a long-running computation. 

The technical implementation is virtual scrolling, this is what the 
terminal does and this is how the browser should do it, too.


On Tuesday, January 5, 2016 at 8:21:53 PM UTC+1, Jason Grout wrote:
>
> (cross-posting to ipython-dev)
>
> Jon,
>
> At the recent San Francisco meetings, we talked about this.  What do you 
> think about:
>
> 1. keeping track of the size of the io messages sent from any specific 
> kernel execution
> 2. When the total size of io reaches some specific size 
> (user-configurable), transmitting a special "throwing away output, but 
> here's how to save the output to a file if you want in the future, or how 
> to increase the limit" message
> 3. keep a running buffer of the last bit of output attempted to be sent, 
> and send it when the execution finishes (so basically a ring buffer that 
> overwrites the oldest message)
>
> This:
>
> * allows small output through
> * provides an explanatory message
> * provides the last bit of output as well
>
> One thing to figure out: a limit on size of output that is text may not be 
> appropriate for output that is images, etc.
>
> Thanks,
>
> Jason
>
>
> On Tue, Jan 5, 2016 at 12:11 PM, Jason Grout <ja...@jasongrout.org 
> > wrote:
>
>>
>> ------ Forwarded message --
>> From: Jonathan Frederic <jon.f...@gmail.com >
>> Date: Tue, Jan 5, 2016 at 11:42 AM
>> Subject: Re: [sage-devel] Re: Jupyter notebook by default?
>> To: Jason Grout <grout...@gmail.com >
>> Cc: sage-devel <sage-...@googlegroups.com >
>>
>>
>> Jason,
>>
>> Thanks for pulling me in on this.  
>>
>> William,
>>
>> I agree, getting a bunch of people to agree on stuff can seem 
>> impossible.  However, you mention Sage offers a couple options to mitigate 
>> output overflows, can you point me to those options?  The Jupyter Notebook 
>> should provide multiple options too - this will also make it easier for 
>> everyone to agree.
>>
>> Also, in you experience, which of these options work the best?  
>>
>> I was thinking initially of doing something simple, like hard limiting 
>> data/time, then printing an error if that's exceeded.  In the Jupyter 
>> Notebook, we have to worry about
>> - Too many messages sent on the websocket
>> - The notebook json file growing too large and consequently becoming 
>> unopenable
>> - Too much data being appended to the DOM, crashing the browser
>>
>>
>> Thanks!
>> -Jon
>>
>> On Tue, Jan 5, 2016 at 10:19 AM, Jason Grout <grout...@gmail.com 
>> > wrote:
>>
>>>
>>>
>>> On Tuesday, January 5, 2016 at 8:17:45 AM UTC-7, William wrote:
>>>>
>>>>
>>>> One example of a subtle feature in Sage (notebook and worksheets) not 
>>>> in Jupyter, which I was just reminded of, is output limiting.  In Sage 
>>>> there are numerous rules/options to deal with people doing stuff like: 
>>>>
>>>> while True: 
>>>>print "hi!" 
>>>>
>>>> ... which is exactly what students will tend to do by accident... 
>>>> Jupyter doesn't deal with this, but it might not be too hard to 
>>>> implement in theory.  One of the main problems is figuring out what 
>>>> the arbitrary rate limiting defaults "should" be; it's arbitrary, and 
>>>> depends a lot on whether everything is local, over the web, etc. so 
>>>> getting a bunch of people to agree is hard, which might mean they will 
>>>> never implement anything. 
>>>>
>>>
>>>
>>> William,
>>>
>>> Jon Frederic in the Jupyter dev meeting happening right now said that he 
>>> will be working on output limiting as one of his next things.
>>>
>>> Jason
>>>
>>
>>
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jeroen Demeyer]

On 2016-01-06 10:05, Volker Braun wrote:

And certainly nobody wants a piece of
their output discarded in a long-running computation.


+1. Actually throwing away output is the worst solution. The full output 
should still be available somehow.


--
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Min RK]

On Wednesday, January 6, 2016 at 10:05:32 AM UTC+1, Volker Braun wrote:
>
> IMHO output capture into a web browser isn't really different from the 
> scrollback buffer of a terminal. We obviously enjoy the infinite scrollback 
> but do not want an unbounded drawing surface in the terminal (= dom nodes 
> in the web browser). And certainly nobody wants a piece of their output 
> discarded in a long-running computation. 
>
> The technical implementation is virtual scrolling, this is what the 
> terminal does and this is how the browser should do it, too.
>

Jon mentioned that there are a few levels for large output to cause 
problems. The lowest bar is putting the output on the page, which is by far 
the easiest to hit, causing an unresponsive browser. This is the level that 
can be addressed by virtual scrolling / truncating output in UI. 
Fortunately, it’s also the easiest one to implement.


If we truncate instead of virtual-scroll, then we have a choice for whether 
truncated output is included in the document or not, which alleviates the 
problem of opening notebooks that have a problematic amount of output. But 
it’s putting that on the page that’s ~always the problem, not loading the 
notebook JSON itself, so I’m somewhat less concerned about that.


The next level where it can cause problems is the output coming over the 
network in the first place. We can throttle this in the notebook server, as 
was implemented for 4.2 <https://github.com/jupyter/notebook/pull/506> months 
ago. Again, this moves the bar for when output causes trouble, but isn’t a 
complete solution. Dumping truncated output to a file is complicated a bit 
by the separations we have in place, but it should be doable.


-MinRK
 

>
>
> On Tuesday, January 5, 2016 at 8:21:53 PM UTC+1, Jason Grout wrote:
>>
>> (cross-posting to ipython-dev)
>>
>> Jon,
>>
>> At the recent San Francisco meetings, we talked about this.  What do you 
>> think about:
>>
>> 1. keeping track of the size of the io messages sent from any specific 
>> kernel execution
>> 2. When the total size of io reaches some specific size 
>> (user-configurable), transmitting a special "throwing away output, but 
>> here's how to save the output to a file if you want in the future, or how 
>> to increase the limit" message
>> 3. keep a running buffer of the last bit of output attempted to be sent, 
>> and send it when the execution finishes (so basically a ring buffer that 
>> overwrites the oldest message)
>>
>> This:
>>
>> * allows small output through
>> * provides an explanatory message
>> * provides the last bit of output as well
>>
>> One thing to figure out: a limit on size of output that is text may not 
>> be appropriate for output that is images, etc.
>>
>> Thanks,
>>
>> Jason
>>
>>
>> On Tue, Jan 5, 2016 at 12:11 PM, Jason Grout <ja...@jasongrout.org> 
>> wrote:
>>
>>>
>>> -- Forwarded message --
>>> From: Jonathan Frederic <jon.f...@gmail.com>
>>> Date: Tue, Jan 5, 2016 at 11:42 AM
>>> Subject: Re: [sage-devel] Re: Jupyter notebook by default?
>>> To: Jason Grout <grout...@gmail.com>
>>> Cc: sage-devel <sage-...@googlegroups.com>
>>>
>>>
>>> Jason,
>>>
>>> Thanks for pulling me in on this.  
>>>
>>> William,
>>>
>>> I agree, getting a bunch of people to agree on stuff can seem 
>>> impossible.  However, you mention Sage offers a couple options to mitigate 
>>> output overflows, can you point me to those options?  The Jupyter Notebook 
>>> should provide multiple options too - this will also make it easier for 
>>> everyone to agree.
>>>
>>> Also, in you experience, which of these options work the best?  
>>>
>>> I was thinking initially of doing something simple, like hard limiting 
>>> data/time, then printing an error if that's exceeded.  In the Jupyter 
>>> Notebook, we have to worry about
>>> - Too many messages sent on the websocket
>>> - The notebook json file growing too large and consequently becoming 
>>> unopenable
>>> - Too much data being appended to the DOM, crashing the browser
>>>
>>>
>>> Thanks!
>>> -Jon
>>>
>>> On Tue, Jan 5, 2016 at 10:19 AM, Jason Grout <grout...@gmail.com> wrote:
>>>
>>>>
>>>>
>>>> On Tuesday, January 5, 2016 at 8:17:45 AM UTC-7, William wrote:
>>>>>
>>>>>
>>>>> One example of a subtle feature in Sage (notebook 

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

On Wednesday, January 6, 2016 at 11:55:36 AM UTC+1, Min RK wrote:
>
> If we truncate instead of virtual-scroll, then we have a choice for 
> whether truncated output is included in the document or not, which 
> alleviates the problem of opening notebooks that have a problematic amount 
> of output
>

There is no fundamental problem with large amounts of output (really, any 
content), and there is essentially only a single way to do it right:

The view (dom) needs only a fixed number of dom nodes for a virtual scroll.

The in-browser view model can lazily load the current scroll position, with 
a suitable cache. Fixed amount of browser JS memory.

The server can just mmap the output file, or alternatively seek around in 
the file. With a suitable index. Fixed amount of server-side memory.

The kernel has to block if the notebook server can't append output fast 
enough, thats normal flow control just like in a pipe. Fixed memory usage 
in the kernel.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[MinRK]

On Wed, Jan 6, 2016 at 1:02 PM, Volker Braun  wrote:

> On Wednesday, January 6, 2016 at 11:55:36 AM UTC+1, Min RK wrote:
>>
>> If we truncate instead of virtual-scroll, then we have a choice for
>> whether truncated output is included in the document or not, which
>> alleviates the problem of opening notebooks that have a problematic amount
>> of output
>>
>
> There is no fundamental problem with large amounts of output (really, any
> content), and there is essentially only a single way to do it right:
>
> The view (dom) needs only a fixed number of dom nodes for a virtual scroll.
>
> The in-browser view model can lazily load the current scroll position,
> with a suitable cache. Fixed amount of browser JS memory.
>
> The server can just mmap the output file, or alternatively seek around in
> the file. With a suitable index. Fixed amount of server-side memory.
>

Files aren't used for output. The filesystem should only be involved, if at
all, in the exceptional case of output overflow.


>
> The kernel has to block if the notebook server can't append output fast
> enough, thats normal flow control just like in a pipe. Fixed memory usage
> in the kernel.
>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "sage-devel" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/sage-devel/8erxWppKxXM/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

On Wednesday, January 6, 2016 at 1:17:12 PM UTC+1, Min RK wrote:
>
> Files aren't used for output. The filesystem should only be involved, if 
> at all, in the exceptional case of output overflow.
>

Everything is a file of sorts... map is just ram with filesystem backing. 

You can put large stuff that you don't continuously access into ram (which 
will then be paged out -> swapfs) or you put it into a temp file (either 
tmpfs->swap or disk). However you call it, large data that is not 
continuously accessed must end up on the disk because thats what it is good 
for.


-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

I'd also be more than happy to ship the personal SMC edition with sage when 
its ready; But sticking with the essentially unsupported SageNB for 1+years 
just to wait is not a sane plan. Even then, jupyter notebooks are a 
forward-safe choice so we have nothing to gain from waiting while people 
write new SageNB notebooks.

As for the Jupyter wishlist, proper output capture would also be nice. 
Right now only the Python-internal stdout is captured, but for example

sage: cython(r'printf("test\n")')
test

yields no output in Jupyter.

There are at least two different multi-user Jupyter versions that are of 
interest; the authenticated (via unix account, much better than SageNB) 
jupyterhub and the anonymous https://tmpnb.org (try it now if you haven't 
seen it)


On Tuesday, January 5, 2016 at 4:17:45 PM UTC+1, William wrote:
>
> On Tue, Jan 5, 2016 at 5:57 AM, kcrisman  
> wrote: 
> > The question wasn't for you, but for all those who early in this thread 
> said 
> > how awesome Jupyter was.  But thank you for confirming. 
>
> Some people tend to use/develop either Jupyter or SageMath notebooks 
> exclusively, and remain a little ignorant of the other one.  There are 
> notable exceptions though, like Jason Grout who works a huge amount on 
> Jupyter development now (for pay!), but is also very knowledgable 
> about Sage.Personally, I'm looking at Jupyter-related stuff today, 
> since I'm rewriting (again) how synchronized editing works, and trying 
> to address various issue... 
>
> One example of a subtle feature in Sage (notebook and worksheets) not 
> in Jupyter, which I was just reminded of, is output limiting.  In Sage 
> there are numerous rules/options to deal with people doing stuff like: 
>
> while True: 
>print "hi!" 
>
> ... which is exactly what students will tend to do by accident... 
> Jupyter doesn't deal with this, but it might not be too hard to 
> implement in theory.  One of the main problems is figuring out what 
> the arbitrary rate limiting defaults "should" be; it's arbitrary, and 
> depends a lot on whether everything is local, over the web, etc. so 
> getting a bunch of people to agree is hard, which might mean they will 
> never implement anything. 
>
> Another basic -- and much harder to implement(!) -- subtle feature of 
> the sage notebook (and SMC) that Jupyter doesn't have is the 
> following.  Try typing 
>
> import time 
> for i in range(10): 
>  time.sleep(1) 
>  print i 
>
> and closing your browser half way through.   In Sagenb (and sagews) 
> it'll compute all the output and put it in the browser, where you'll 
> find it later when you visit the page.  In Jupyter, all the output 
> that appears when you aren't observing the computation is lost.   I 
> remember in maybe 2006 or 2007 implementing this and that it was very 
> important to researchers -- you can just start: 
>
> for n in range(100): 
> print n, important_function_of(n) 
>
> and come back tomorrow and see the result -- researchers *love* to be 
> able to do  that without having to worry.  With Jupyter, you have no 
> choice but to create a file, and output each result to that file, then 
> look in the file later; this is a bigger cognitive load. 
>
> Implementing the above (recording all output without the browser 
> client open) requires adding a slightly nontrivial  idea to how 
> Jupyter is implemented, so I don't think it's likely to be really 
> easy. 
>
> Don't worry -- I've repeatedly mentioned the above differences to many 
> Jupyter developers, and I'm sure they will get addressed, since there 
> are a ton of people working on Jupyter. 
>
> Anyway, there are many subtle differences...  Everything can be worked 
> around, of course. 
>
> > 
> >>> 
> >>> And what of the long-term in Sage itself - would an eventual "SMC 
> >>> personal edition" become the default?  [Not rhetorical but probably 
> too far 
> >>> in the future to speculate] 
> >> 
> >> 
> >> No clue.  There's no legal or technical reason it couldn't happen... 
> This 
> >> year it likely won't as SMC is still changing way too much. 
> > 
> > 
> > So technical reason in the sense that even if someone did it, it would 
> > require a lot of maintenance to keep up with official SMC. 
> > 
> > +++ 
> > 
> > On a less sarcastic note (and my apologies for that) I'm wondering what 
> the 
> > status of Jupyterhub (the multi-user Jupyter, right?) is right now. 
>  Active 
> > development, but so is HURD... not that I am expecting it to take 30 
> years 
> > to produce!  Just curious if there are any inside scoops.  Proper 
> migration 
> > of entire servers being possible would be a much bigger reason to change 
> the 
> > default. 
> > 
> > -- 
> > You received this message because you are subscribed to the Google 
> Groups 
> > "sage-devel" group. 
> > To unsubscribe from this group and stop receiving emails from it, send 
> an 
> > email to sage-devel+...@googlegroups.com . 
> > To post to this group, 

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Tue, Jan 5, 2016 at 8:14 AM, Volker Braun  wrote:
> I'd also be more than happy to ship the personal SMC edition with sage when
> its ready; But sticking with the essentially unsupported SageNB for 1+years
> just to wait is not a sane plan. Even then, jupyter notebooks are a
> forward-safe choice so we have nothing to gain from waiting while people
> write new SageNB notebooks.

+1  and "personal SMC" is going to fully support using Jupyter no
matter what, which makes Jupyter notebooks even safer.

> As for the Jupyter wishlist, proper output capture would also be nice. Right
> now only the Python-internal stdout is captured, but for example
>
> sage: cython(r'printf("test\n")')
> test
>
> yields no output in Jupyter.
>
> There are at least two different multi-user Jupyter versions that are of
> interest; the authenticated (via unix account, much better than SageNB)
> jupyterhub and the anonymous https://tmpnb.org (try it now if you haven't
> seen it)

And SageMathCloud, which also provides multi-user Jupyter.  At the
moment I write this, there are 59 jupyter notebooks running on SMC --
see "Running Instances" here:

 
https://cloud.sagemath.com/b97f6266-fe6f-4b40-bd88-9798994a04d1/raw/metrics/metrics.html

>
>
> On Tuesday, January 5, 2016 at 4:17:45 PM UTC+1, William wrote:
>>
>> On Tue, Jan 5, 2016 at 5:57 AM, kcrisman  wrote:
>> > The question wasn't for you, but for all those who early in this thread
>> > said
>> > how awesome Jupyter was.  But thank you for confirming.
>>
>> Some people tend to use/develop either Jupyter or SageMath notebooks
>> exclusively, and remain a little ignorant of the other one.  There are
>> notable exceptions though, like Jason Grout who works a huge amount on
>> Jupyter development now (for pay!), but is also very knowledgable
>> about Sage.Personally, I'm looking at Jupyter-related stuff today,
>> since I'm rewriting (again) how synchronized editing works, and trying
>> to address various issue...
>>
>> One example of a subtle feature in Sage (notebook and worksheets) not
>> in Jupyter, which I was just reminded of, is output limiting.  In Sage
>> there are numerous rules/options to deal with people doing stuff like:
>>
>> while True:
>>print "hi!"
>>
>> ... which is exactly what students will tend to do by accident...
>> Jupyter doesn't deal with this, but it might not be too hard to
>> implement in theory.  One of the main problems is figuring out what
>> the arbitrary rate limiting defaults "should" be; it's arbitrary, and
>> depends a lot on whether everything is local, over the web, etc. so
>> getting a bunch of people to agree is hard, which might mean they will
>> never implement anything.
>>
>> Another basic -- and much harder to implement(!) -- subtle feature of
>> the sage notebook (and SMC) that Jupyter doesn't have is the
>> following.  Try typing
>>
>> import time
>> for i in range(10):
>>  time.sleep(1)
>>  print i
>>
>> and closing your browser half way through.   In Sagenb (and sagews)
>> it'll compute all the output and put it in the browser, where you'll
>> find it later when you visit the page.  In Jupyter, all the output
>> that appears when you aren't observing the computation is lost.   I
>> remember in maybe 2006 or 2007 implementing this and that it was very
>> important to researchers -- you can just start:
>>
>> for n in range(100):
>> print n, important_function_of(n)
>>
>> and come back tomorrow and see the result -- researchers *love* to be
>> able to do  that without having to worry.  With Jupyter, you have no
>> choice but to create a file, and output each result to that file, then
>> look in the file later; this is a bigger cognitive load.
>>
>> Implementing the above (recording all output without the browser
>> client open) requires adding a slightly nontrivial  idea to how
>> Jupyter is implemented, so I don't think it's likely to be really
>> easy.
>>
>> Don't worry -- I've repeatedly mentioned the above differences to many
>> Jupyter developers, and I'm sure they will get addressed, since there
>> are a ton of people working on Jupyter.
>>
>> Anyway, there are many subtle differences...  Everything can be worked
>> around, of course.
>>
>> >
>> >>>
>> >>> And what of the long-term in Sage itself - would an eventual "SMC
>> >>> personal edition" become the default?  [Not rhetorical but probably
>> >>> too far
>> >>> in the future to speculate]
>> >>
>> >>
>> >> No clue.  There's no legal or technical reason it couldn't happen...
>> >> This
>> >> year it likely won't as SMC is still changing way too much.
>> >
>> >
>> > So technical reason in the sense that even if someone did it, it would
>> > require a lot of maintenance to keep up with official SMC.
>> >
>> > +++
>> >
>> > On a less sarcastic note (and my apologies for that) I'm wondering what
>> > the
>> > status of Jupyterhub (the multi-user Jupyter, right?) is right now.
>> > Active
>> > development, but so is 

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

Indeed, there are at least *three* different multi-user Jupyter servers.

On Tuesday, January 5, 2016 at 5:26:24 PM UTC+1, William wrote:
>
> > There are at least two different multi-user Jupyter versions that are of 
> > interest; the authenticated (via unix account, much better than SageNB) 
> > jupyterhub and the anonymous https://tmpnb.org (try it now if you 
> haven't 
> > seen it) 
>
> And SageMathCloud, which also provides multi-user Jupyter.  At the 
> moment I write this, there are 59 jupyter notebooks running on SMC -- 
> see "Running Instances" here: 
>
>  
> https://cloud.sagemath.com/b97f6266-fe6f-4b40-bd88-9798994a04d1/raw/metrics/metrics.html
>  
>
> > 
> > 
> > On Tuesday, January 5, 2016 at 4:17:45 PM UTC+1, William wrote: 
> >> 
> >> On Tue, Jan 5, 2016 at 5:57 AM, kcrisman  wrote: 
> >> > The question wasn't for you, but for all those who early in this 
> thread 
> >> > said 
> >> > how awesome Jupyter was.  But thank you for confirming. 
> >> 
> >> Some people tend to use/develop either Jupyter or SageMath notebooks 
> >> exclusively, and remain a little ignorant of the other one.  There are 
> >> notable exceptions though, like Jason Grout who works a huge amount on 
> >> Jupyter development now (for pay!), but is also very knowledgable 
> >> about Sage.Personally, I'm looking at Jupyter-related stuff today, 
> >> since I'm rewriting (again) how synchronized editing works, and trying 
> >> to address various issue... 
> >> 
> >> One example of a subtle feature in Sage (notebook and worksheets) not 
> >> in Jupyter, which I was just reminded of, is output limiting.  In Sage 
> >> there are numerous rules/options to deal with people doing stuff like: 
> >> 
> >> while True: 
> >>print "hi!" 
> >> 
> >> ... which is exactly what students will tend to do by accident... 
> >> Jupyter doesn't deal with this, but it might not be too hard to 
> >> implement in theory.  One of the main problems is figuring out what 
> >> the arbitrary rate limiting defaults "should" be; it's arbitrary, and 
> >> depends a lot on whether everything is local, over the web, etc. so 
> >> getting a bunch of people to agree is hard, which might mean they will 
> >> never implement anything. 
> >> 
> >> Another basic -- and much harder to implement(!) -- subtle feature of 
> >> the sage notebook (and SMC) that Jupyter doesn't have is the 
> >> following.  Try typing 
> >> 
> >> import time 
> >> for i in range(10): 
> >>  time.sleep(1) 
> >>  print i 
> >> 
> >> and closing your browser half way through.   In Sagenb (and sagews) 
> >> it'll compute all the output and put it in the browser, where you'll 
> >> find it later when you visit the page.  In Jupyter, all the output 
> >> that appears when you aren't observing the computation is lost.   I 
> >> remember in maybe 2006 or 2007 implementing this and that it was very 
> >> important to researchers -- you can just start: 
> >> 
> >> for n in range(100): 
> >> print n, important_function_of(n) 
> >> 
> >> and come back tomorrow and see the result -- researchers *love* to be 
> >> able to do  that without having to worry.  With Jupyter, you have no 
> >> choice but to create a file, and output each result to that file, then 
> >> look in the file later; this is a bigger cognitive load. 
> >> 
> >> Implementing the above (recording all output without the browser 
> >> client open) requires adding a slightly nontrivial  idea to how 
> >> Jupyter is implemented, so I don't think it's likely to be really 
> >> easy. 
> >> 
> >> Don't worry -- I've repeatedly mentioned the above differences to many 
> >> Jupyter developers, and I'm sure they will get addressed, since there 
> >> are a ton of people working on Jupyter. 
> >> 
> >> Anyway, there are many subtle differences...  Everything can be worked 
> >> around, of course. 
> >> 
> >> > 
> >> >>> 
> >> >>> And what of the long-term in Sage itself - would an eventual "SMC 
> >> >>> personal edition" become the default?  [Not rhetorical but probably 
> >> >>> too far 
> >> >>> in the future to speculate] 
> >> >> 
> >> >> 
> >> >> No clue.  There's no legal or technical reason it couldn't happen... 
> >> >> This 
> >> >> year it likely won't as SMC is still changing way too much. 
> >> > 
> >> > 
> >> > So technical reason in the sense that even if someone did it, it 
> would 
> >> > require a lot of maintenance to keep up with official SMC. 
> >> > 
> >> > +++ 
> >> > 
> >> > On a less sarcastic note (and my apologies for that) I'm wondering 
> what 
> >> > the 
> >> > status of Jupyterhub (the multi-user Jupyter, right?) is right now. 
> >> > Active 
> >> > development, but so is HURD... not that I am expecting it to take 30 
> >> > years 
> >> > to produce!  Just curious if there are any inside scoops.  Proper 
> >> > migration 
> >> > of entire servers being possible would be a much bigger reason to 
> change 
> >> > the 
> >> > default. 
> >> > 
> >> > -- 
> >> > You received 

Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

>  
>>
>>> > As stated, that would break the compatibility with SMC magics that 
>>> William 
>>> > just declared intangible (and always "cell magic"). 
>>>
>>> Just to be clear:  I  don't care whether or not whatever design you 
>>> guys are coming up with in this thread is compatible with SMC Sage 
>>> worksheets.  SMC provides both Sage worksheets and Jupyter notebooks 
>>> on an equal footing; each has its pros and cons, and will continue to 
>>> be fully supported (likely with a converter between them, that at 
>>> least converts what can be converted). 
>>>
>>>  
>> I would be interested in why SMC provides Sage worksheets if Jupyter is 
>> so *clearly* the most superior option.  [Rhetorical question]
>>
>
> Weird question, given that I literally just wrote "Sage worksheets and 
> Jupyter notebooks on an equal footing; each has its pros and cons, and will 
> continue to be fully supported". There are many ways in which Sage 
> worksheets are superior to Jupyter notebooks, and some very nontrivial 
> subtle new things I've been working on the last month that will lead to 
> more...
>  
>

The question wasn't for you, but for all those who early in this thread 
said how awesome Jupyter was.  But thank you for confirming.


>> And what of the long-term in Sage itself - would an eventual "SMC 
>> personal edition" become the default?  [Not rhetorical but probably too far 
>> in the future to speculate]
>>
>
> No clue.  There's no legal or technical reason it couldn't happen... This 
> year it likely won't as SMC is still changing way too much.
>

So technical reason in the sense that even if someone did it, it would 
require a lot of maintenance to keep up with official SMC.

+++

On a less sarcastic note (and my apologies for that) I'm wondering what the 
status of Jupyterhub (the multi-user Jupyter, right?) is right now.  Active 
development, but so is HURD... not that I am expecting it to take 30 years 
to produce!  Just curious if there are any inside scoops.  Proper migration 
of entire servers being possible would be a much bigger reason to change 
the default.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Tue, Jan 5, 2016 at 11:19 AM, Jason Grout <ja...@jasongrout.org> wrote:

> (cross-posting to ipython-dev)
>
> Jon,
>
> At the recent San Francisco meetings, we talked about this.  What do you
> think about:
>
> 1. keeping track of the size of the io messages sent from any specific
> kernel execution
> 2. When the total size of io reaches some specific size
> (user-configurable), transmitting a special "throwing away output, but
> here's how to save the output to a file if you want in the future, or how
> to increase the limit" message
> 3. keep a running buffer of the last bit of output attempted to be sent,
> and send it when the execution finishes (so basically a ring buffer that
> overwrites the oldest message)
>
> This:
>
> * allows small output through
> * provides an explanatory message
> * provides the last bit of output as well
>
> One thing to figure out: a limit on size of output that is text may not be
> appropriate for output that is images, etc.
>

The above strategy is a good start, but I've found it to be too naive in
practice.  For example, in SMC there is a different limit on the amount of
output that will be rendered with mathjax, versus the amount of stdout
output.  This is because mathjax rendering is vastly more resource
intensive than text rendering.  Also, if your graphics images are just
contents of messages in base64 (say), they can be relatively large, but
easy to render quickly.

You might also want to distinguish between local users and people using
Jupyter via a remote server.  If everything is running on your laptop, the
network situation is completely different than a remote server talking to a
cell phone.

I know you wrote "user-configurable" above, but it's a possibly bad sign
when user configuration is required.

Have fun at the Jupyter dev meeting!

William



>
> Thanks,
>
> Jason
>
>
> On Tue, Jan 5, 2016 at 12:11 PM, Jason Grout <ja...@jasongrout.org> wrote:
>
>>
>> ------ Forwarded message --
>> From: Jonathan Frederic <jon.fre...@gmail.com>
>> Date: Tue, Jan 5, 2016 at 11:42 AM
>> Subject: Re: [sage-devel] Re: Jupyter notebook by default?
>> To: Jason Grout <grout.ja...@gmail.com>
>> Cc: sage-devel <sage-devel@googlegroups.com>
>>
>>
>> Jason,
>>
>> Thanks for pulling me in on this.
>>
>> William,
>>
>> I agree, getting a bunch of people to agree on stuff can seem
>> impossible.  However, you mention Sage offers a couple options to mitigate
>> output overflows, can you point me to those options?  The Jupyter Notebook
>> should provide multiple options too - this will also make it easier for
>> everyone to agree.
>>
>> Also, in you experience, which of these options work the best?
>>
>> I was thinking initially of doing something simple, like hard limiting
>> data/time, then printing an error if that's exceeded.  In the Jupyter
>> Notebook, we have to worry about
>> - Too many messages sent on the websocket
>> - The notebook json file growing too large and consequently becoming
>> unopenable
>> - Too much data being appended to the DOM, crashing the browser
>>
>>
>> Thanks!
>> -Jon
>>
>> On Tue, Jan 5, 2016 at 10:19 AM, Jason Grout <grout.ja...@gmail.com>
>> wrote:
>>
>>>
>>>
>>> On Tuesday, January 5, 2016 at 8:17:45 AM UTC-7, William wrote:
>>>>
>>>>
>>>> One example of a subtle feature in Sage (notebook and worksheets) not
>>>> in Jupyter, which I was just reminded of, is output limiting.  In Sage
>>>> there are numerous rules/options to deal with people doing stuff like:
>>>>
>>>> while True:
>>>>print "hi!"
>>>>
>>>> ... which is exactly what students will tend to do by accident...
>>>> Jupyter doesn't deal with this, but it might not be too hard to
>>>> implement in theory.  One of the main problems is figuring out what
>>>> the arbitrary rate limiting defaults "should" be; it's arbitrary, and
>>>> depends a lot on whether everything is local, over the web, etc. so
>>>> getting a bunch of people to agree is hard, which might mean they will
>>>> never implement anything.
>>>>
>>>
>>>
>>> William,
>>>
>>> Jon Frederic in the Jupyter dev meeting happening right now said that he
>>> will be working on output limiting as one of his next things.
>>>
>>> Jason
>>>
>>
>>
>>
>


-- 
William (http://wstein.org)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

> I'd also be more than happy to ship the personal SMC edition with sage 
> when its ready; But sticking with the essentially unsupported SageNB for 
> 1+years just to wait is not a sane plan. Even then, jupyter notebooks 
> are a forward-safe choice so we have nothing to gain from waiting while 
> people write new SageNB notebooks.
>
> As for the Jupyter wishlist, proper output capture would also be nice. 
> Right now only the Python-internal stdout is captured, but for example
>
> sage: cython(r'printf("test\n")')
> test
>
> yields no output in Jupyter.
>
> There are at least two different multi-user Jupyter versions that are of 
> interest; the authenticated (via unix account, much better than SageNB) 
> jupyterhub and the anonymous https://tmpnb.org (try it now if you haven't 
> seen it)
>
>
>>
It's nice, though it didn't compute anything for me (likely I have a bad 
connection). 

> Don't worry -- I've repeatedly mentioned the above differences to 
many Jupyter developers, and I'm sure they will get addressed, since 
there are a ton of people working on Jupyter. 


Right.  

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jason Grout]

On Tuesday, January 5, 2016 at 8:17:45 AM UTC-7, William wrote:
>
>
> One example of a subtle feature in Sage (notebook and worksheets) not 
> in Jupyter, which I was just reminded of, is output limiting.  In Sage 
> there are numerous rules/options to deal with people doing stuff like: 
>
> while True: 
>print "hi!" 
>
> ... which is exactly what students will tend to do by accident... 
> Jupyter doesn't deal with this, but it might not be too hard to 
> implement in theory.  One of the main problems is figuring out what 
> the arbitrary rate limiting defaults "should" be; it's arbitrary, and 
> depends a lot on whether everything is local, over the web, etc. so 
> getting a bunch of people to agree is hard, which might mean they will 
> never implement anything. 
>


William,

Jon Frederic in the Jupyter dev meeting happening right now said that he 
will be working on output limiting as one of his next things.

Jason

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jason Grout]

FYI, Sylvain Corlay is making some changes to ipywidgets to bring them more 
into line with the Sage syntax.  He said in the Jupyter dev meeting just 
now that he'll be making a PR today.

Thanks,

Jason

On Sunday, January 3, 2016 at 7:45:19 AM UTC-7, Jeroen Demeyer wrote:
>
> On 2016-01-03 14:41, Emmanuel Charpentier wrote: 
> >   * Interacts : In a Python worksheet, ipywidgets offers a nice 
> > interactive framework, ... which turns out no be non obvious to use 
> > from a Sage worksheet. A replacement and/or a compatibility layer 
> > are needed. (Even if we offer a replacement, a compatibility layer 
> > might be useful to ease transition/conversion of old sagenb 
> worksheets). 
>
> During Sage Days 70, some work was done to make ipywidgets interacts 
> usable within Sage. This now works using the git master for ipywidgets 
> (which is not in Sage yet). 
>
> Unfortunately, the API of @interact has some significant differences 
> between SageNB and Jupyter. IMHO, that of SageNB makes more sense but I 
> have not been able to convince upstream Jupyter of this fact. Jupyter 
> might have more fancy stuff (like a game controller widget), but fewer 
> basic "just works" stuff. 
>
> I think that a compatibility layer might be possible, but I don't know 
> if it can cover all features of SageNB interacts. 
>
> Jeroen. 
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jason Grout]

(cross-posting to ipython-dev)

Jon,

At the recent San Francisco meetings, we talked about this.  What do you
think about:

1. keeping track of the size of the io messages sent from any specific
kernel execution
2. When the total size of io reaches some specific size
(user-configurable), transmitting a special "throwing away output, but
here's how to save the output to a file if you want in the future, or how
to increase the limit" message
3. keep a running buffer of the last bit of output attempted to be sent,
and send it when the execution finishes (so basically a ring buffer that
overwrites the oldest message)

This:

* allows small output through
* provides an explanatory message
* provides the last bit of output as well

One thing to figure out: a limit on size of output that is text may not be
appropriate for output that is images, etc.

Thanks,

Jason


On Tue, Jan 5, 2016 at 12:11 PM, Jason Grout <ja...@jasongrout.org> wrote:

>
> -- Forwarded message --
> From: Jonathan Frederic <jon.fre...@gmail.com>
> Date: Tue, Jan 5, 2016 at 11:42 AM
> Subject: Re: [sage-devel] Re: Jupyter notebook by default?
> To: Jason Grout <grout.ja...@gmail.com>
> Cc: sage-devel <sage-devel@googlegroups.com>
>
>
> Jason,
>
> Thanks for pulling me in on this.
>
> William,
>
> I agree, getting a bunch of people to agree on stuff can seem impossible.
> However, you mention Sage offers a couple options to mitigate output
> overflows, can you point me to those options?  The Jupyter Notebook should
> provide multiple options too - this will also make it easier for everyone
> to agree.
>
> Also, in you experience, which of these options work the best?
>
> I was thinking initially of doing something simple, like hard limiting
> data/time, then printing an error if that's exceeded.  In the Jupyter
> Notebook, we have to worry about
> - Too many messages sent on the websocket
> - The notebook json file growing too large and consequently becoming
> unopenable
> - Too much data being appended to the DOM, crashing the browser
>
>
> Thanks!
> -Jon
>
> On Tue, Jan 5, 2016 at 10:19 AM, Jason Grout <grout.ja...@gmail.com>
> wrote:
>
>>
>>
>> On Tuesday, January 5, 2016 at 8:17:45 AM UTC-7, William wrote:
>>>
>>>
>>> One example of a subtle feature in Sage (notebook and worksheets) not
>>> in Jupyter, which I was just reminded of, is output limiting.  In Sage
>>> there are numerous rules/options to deal with people doing stuff like:
>>>
>>> while True:
>>>print "hi!"
>>>
>>> ... which is exactly what students will tend to do by accident...
>>> Jupyter doesn't deal with this, but it might not be too hard to
>>> implement in theory.  One of the main problems is figuring out what
>>> the arbitrary rate limiting defaults "should" be; it's arbitrary, and
>>> depends a lot on whether everything is local, over the web, etc. so
>>> getting a bunch of people to agree is hard, which might mean they will
>>> never implement anything.
>>>
>>
>>
>> William,
>>
>> Jon Frederic in the Jupyter dev meeting happening right now said that he
>> will be working on output limiting as one of his next things.
>>
>> Jason
>>
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jason Grout]

FYI, you can set the notebook directory with a command line switch:

jupyter notebook --NotebookApp.notebook_dir=~/.sage/jupyter

Thanks,

Jason

On Sunday, December 20, 2015 at 5:25:52 PM UTC-7, Volker Braun wrote:
>
> If you start Jupyter in ~/foo then you cannot open notebooks in 
> ~/.sage/jupyter. Thats just how it is.
>
> On Monday, December 21, 2015 at 1:12:57 AM UTC+1, William wrote:
>>
>> I clearly don't agree with that, since that's never been the case 
>> before, as I mentioned.  I'll not press this any further. 
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jonathan Frederic]

Jason,

Thanks for pulling me in on this.

William,

I agree, getting a bunch of people to agree about stuff can be impossible.
You mention Sage offers a couple options to mitigate output overflows, can
you point me to those options?  The Jupyter Notebook should provide
multiple options for this too.  Also, in your experience, which of the Sage
options work the best?

I was thinking initially of doing something simple, like hard limiting
data/time, then printing an error if that's exceeded.  In the Jupyter
Notebook, we have to worry about:
- Too many messages sent on the websocket
- The notebook json file growing too large and consequently becoming
unopenable
- Too much data being appended to the DOM, crashing the browser


Thanks!
-Jon

On Tue, Jan 5, 2016 at 10:19 AM, Jason Grout  wrote:

>
>
> On Tuesday, January 5, 2016 at 8:17:45 AM UTC-7, William wrote:
>>
>>
>> One example of a subtle feature in Sage (notebook and worksheets) not
>> in Jupyter, which I was just reminded of, is output limiting.  In Sage
>> there are numerous rules/options to deal with people doing stuff like:
>>
>> while True:
>>print "hi!"
>>
>> ... which is exactly what students will tend to do by accident...
>> Jupyter doesn't deal with this, but it might not be too hard to
>> implement in theory.  One of the main problems is figuring out what
>> the arbitrary rate limiting defaults "should" be; it's arbitrary, and
>> depends a lot on whether everything is local, over the web, etc. so
>> getting a bunch of people to agree is hard, which might mean they will
>> never implement anything.
>>
>
>
> William,
>
> Jon Frederic in the Jupyter dev meeting happening right now said that he
> will be working on output limiting as one of his next things.
>
> Jason
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Tue, Jan 5, 2016 at 5:57 AM, kcrisman  wrote:
> The question wasn't for you, but for all those who early in this thread said
> how awesome Jupyter was.  But thank you for confirming.

Some people tend to use/develop either Jupyter or SageMath notebooks
exclusively, and remain a little ignorant of the other one.  There are
notable exceptions though, like Jason Grout who works a huge amount on
Jupyter development now (for pay!), but is also very knowledgable
about Sage.Personally, I'm looking at Jupyter-related stuff today,
since I'm rewriting (again) how synchronized editing works, and trying
to address various issue...

One example of a subtle feature in Sage (notebook and worksheets) not
in Jupyter, which I was just reminded of, is output limiting.  In Sage
there are numerous rules/options to deal with people doing stuff like:

while True:
   print "hi!"

... which is exactly what students will tend to do by accident...
Jupyter doesn't deal with this, but it might not be too hard to
implement in theory.  One of the main problems is figuring out what
the arbitrary rate limiting defaults "should" be; it's arbitrary, and
depends a lot on whether everything is local, over the web, etc. so
getting a bunch of people to agree is hard, which might mean they will
never implement anything.

Another basic -- and much harder to implement(!) -- subtle feature of
the sage notebook (and SMC) that Jupyter doesn't have is the
following.  Try typing

import time
for i in range(10):
 time.sleep(1)
 print i

and closing your browser half way through.   In Sagenb (and sagews)
it'll compute all the output and put it in the browser, where you'll
find it later when you visit the page.  In Jupyter, all the output
that appears when you aren't observing the computation is lost.   I
remember in maybe 2006 or 2007 implementing this and that it was very
important to researchers -- you can just start:

for n in range(100):
print n, important_function_of(n)

and come back tomorrow and see the result -- researchers *love* to be
able to do  that without having to worry.  With Jupyter, you have no
choice but to create a file, and output each result to that file, then
look in the file later; this is a bigger cognitive load.

Implementing the above (recording all output without the browser
client open) requires adding a slightly nontrivial  idea to how
Jupyter is implemented, so I don't think it's likely to be really
easy.

Don't worry -- I've repeatedly mentioned the above differences to many
Jupyter developers, and I'm sure they will get addressed, since there
are a ton of people working on Jupyter.

Anyway, there are many subtle differences...  Everything can be worked
around, of course.

>
>>>
>>> And what of the long-term in Sage itself - would an eventual "SMC
>>> personal edition" become the default?  [Not rhetorical but probably too far
>>> in the future to speculate]
>>
>>
>> No clue.  There's no legal or technical reason it couldn't happen... This
>> year it likely won't as SMC is still changing way too much.
>
>
> So technical reason in the sense that even if someone did it, it would
> require a lot of maintenance to keep up with official SMC.
>
> +++
>
> On a less sarcastic note (and my apologies for that) I'm wondering what the
> status of Jupyterhub (the multi-user Jupyter, right?) is right now.  Active
> development, but so is HURD... not that I am expecting it to take 30 years
> to produce!  Just curious if there are any inside scoops.  Proper migration
> of entire servers being possible would be a much bigger reason to change the
> default.
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.



-- 
William (http://wstein.org)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jason Grout]

Brian, Sylvain, and Jon talked and came to an agreement.

Jason


On Tue, Jan 5, 2016 at 2:42 PM, Jeroen Demeyer 
wrote:

> On 2016-01-05 19:27, Jason Grout wrote:
>
>> FYI, Sylvain Corlay is making some changes to ipywidgets to bring them
>> more into line with the Sage syntax.  He said in the Jupyter dev meeting
>> just now that he'll be making a PR today.
>>
>
> Cool, however it seems from
> https://github.com/ipython/ipywidgets/issues/238 that @ellisonbg was the
> main person who needed to be convinced.
>
>
> Jeroen.
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

I know this is now hijacking the thread... but if we are on those lines, 
when it comes to "too much output", making sure that this is something that 
can allow people to easily see and compare an "old" "too much output" to a 
"new" "too much output" after evaluating a cell/command a second/several 
time/s would be very helpful - we've had numerous requests along those 
lines.  Doubtless this is already in the works though - nice to hear from 
you, Jason! :)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jason Grout]

I'm not sure exactly what you mean here.  Can you give an example?

Thanks,

Jason

On Tue, Jan 5, 2016 at 3:16 PM, kcrisman  wrote:

> I know this is now hijacking the thread... but if we are on those lines,
> when it comes to "too much output", making sure that this is something that
> can allow people to easily see and compare an "old" "too much output" to a
> "new" "too much output" after evaluating a cell/command a second/several
> time/s would be very helpful - we've had numerous requests along those
> lines.  Doubtless this is already in the works though - nice to hear from
> you, Jason! :)
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jeroen Demeyer]

On 2016-01-05 19:27, Jason Grout wrote:

FYI, Sylvain Corlay is making some changes to ipywidgets to bring them
more into line with the Sage syntax.  He said in the Jupyter dev meeting
just now that he'll be making a PR today.


Cool, however it seems from 
https://github.com/ipython/ipywidgets/issues/238 that @ellisonbg was the 
main person who needed to be convinced.


Jeroen.

--
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

On Tuesday, January 5, 2016 at 5:18:27 PM UTC-5, Jason Grout wrote:
>
> I'm not sure exactly what you mean here.  Can you give an example?
>
>
See https://github.com/sagemath/sagenb/issues/224 - it would be nice in 
Jupyter to have a way to allow two different behaviors as desired or 
relevant, one where full output (wherever it's located) is always current, 
and one where you can see different versions of full output for comparison 
or something.  Assuming that "full output" has meaning, though from the 
discussion above it seems like there are situations where the full output 
of a cell would not all be displayed by default.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

Thank you Jonathan for clarifying what I thought was obvious in what I was 
saying.
 

> > As stated, that would break the compatibility with SMC magics that 
> William 
> > just declared intangible (and always "cell magic"). 
>
> Just to be clear:  I  don't care whether or not whatever design you 
> guys are coming up with in this thread is compatible with SMC Sage 
> worksheets.  SMC provides both Sage worksheets and Jupyter notebooks 
> on an equal footing; each has its pros and cons, and will continue to 
> be fully supported (likely with a converter between them, that at 
> least converts what can be converted). 
>
>  
I would be interested in why SMC provides Sage worksheets if Jupyter is so 
*clearly* the most superior option.  [Rhetorical question]

And what of the long-term in Sage itself - would an eventual "SMC personal 
edition" become the default?  [Not rhetorical but probably too far in the 
future to speculate]

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Mon, Jan 4, 2016 at 10:21 AM, kcrisman  wrote:

> Thank you Jonathan for clarifying what I thought was obvious in what I was
> saying.
>
>
>> > As stated, that would break the compatibility with SMC magics that
>> William
>> > just declared intangible (and always "cell magic").
>>
>> Just to be clear:  I  don't care whether or not whatever design you
>> guys are coming up with in this thread is compatible with SMC Sage
>> worksheets.  SMC provides both Sage worksheets and Jupyter notebooks
>> on an equal footing; each has its pros and cons, and will continue to
>> be fully supported (likely with a converter between them, that at
>> least converts what can be converted).
>>
>>
> I would be interested in why SMC provides Sage worksheets if Jupyter is so
> *clearly* the most superior option.  [Rhetorical question]
>

Weird question, given that I literally just wrote "Sage worksheets and
Jupyter notebooks on an equal footing; each has its pros and cons, and will
continue to be fully supported". There are many ways in which Sage
worksheets are superior to Jupyter notebooks, and some very nontrivial
subtle new things I've been working on the last month that will lead to
more...


>
> And what of the long-term in Sage itself - would an eventual "SMC personal
> edition" become the default?  [Not rhetorical but probably too far in the
> future to speculate]
>

No clue.  There's no legal or technical reason it couldn't happen... This
year it likely won't as SMC is still changing way too much.

 -- William


> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.
>



-- 
William (http://wstein.org)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Sun, Jan 3, 2016 at 1:49 PM, Emmanuel Charpentier
 wrote:
>
>
> Le dimanche 3 janvier 2016 21:23:56 UTC+1, Volker Braun a écrit :
>>
>> On Sunday, January 3, 2016 at 5:36:30 PM UTC+1, Emmanuel Charpentier
>> wrote:
>>>
>>> [ BTW : that's not really a bug, but rather a design conflict : the
>>> original %mode functions were designed to switch (for an indefinite scope)
>>> the behaviour of the REPL.This was transposed in the Sage notebook as what
>>> amounts to the equivalent of Jupyter's cell magics (scope defined as the
>>> current cell). Whereas line magics are, as far as I understand,
>>> Jupyter-specific... We can't be consistent across notebooks without
>>> redefining our "mode switch" magics as "cell magics", and rename them with
>>> "%%"...  ]
>>
>>
>> IMHO we should just switch the interface line magics to work like other
>> line magics in Jupyter/IPython, and also SMC: Take the rest of the line as
>> string and evaluate
>
>
> As stated, that would break the compatibility with SMC magics that William
> just declared intangible (and always "cell magic").

Just to be clear:  I  don't care whether or not whatever design you
guys are coming up with in this thread is compatible with SMC Sage
worksheets.  SMC provides both Sage worksheets and Jupyter notebooks
on an equal footing; each has its pros and cons, and will continue to
be fully supported (likely with a converter between them, that at
least converts what can be converted).

>  This could be (more or
> less) restored  if the magic function had access to the current interface.
> Pseudo-Python ahead :
>
> args, input=(RestOfTheFirstLline(), RestOfTheCell())
> if(SMC):
> ReturnAndPrint(magicfun(input))
> if(commandline): ## In that case, by definition, input==None
> if(restofthefirstline=""):
> switchREPLuntilEOFOrOtherMagic(magicfun())
> else:
> ReturnAndPrint(magicfun(args))
> else: ## Jupyter
> ProcessOptions(args)
> ReturnAndPrint(magicfun(input, ProcessedOptions))
>
> The difficulty is, of course, that the same part of the command structure
> (the rest of the first line) would mean two different things :
> - source code to execute in a command line
> - possible options to be  passed to a cell magic in Jupyter (some interface,
> already supported (rpy2, the IRkernel for R) already use this convention !).
>
> Furthermore, I supposed that SMC ignored the rest of the first line. This is
> what the %r magic does
> %r print(2+2)
> print(3+2)
>
> ## prints [1] 5, even if the pasted result is a bit more complicated :
> ︡3c5a5418-14e0-46fe-a34d-3fa7a0eb1b7d︡︡{"stdout":"[1]
> 4","done":false}︡{"stdout":" 5\n","done":false}︡{"done":true}
> ︠68da460a-f070-4758-9fbb-667f85c4819es︠
>
> but not the %maxima magic :
>
>
> %maxima diff(log(x),x)
> integrate(arctan(x),x)
>
> ## prints 1/x on one line, then x*arctan(x) - 1/2*log(x^2 + 1) on the next
> line, but again the reality is a bit more complicated :
>
> ︡410cbb81-a181-4e01-92a0-0c3e28da800a︡︡{"stdout":"1/x","done":false}︡{"stdout":"\n","done":false}︡{"stdout":"x*arctan(x)
> - 1/2*log(x^2 + 1)\n","done":false}︡{"done":true}
> ︠e820fb8a-cd58-4173-9d99-95dd022e4a6f︠
>
> Therefore, we have a specification/consistency problem, to be solved BEFORE
> the coding problem(s).
>
> We should also specify what happens with the computed results : are they
> just displayed in the output cell and kept in the target interpreterv
> instance ? (In that case, we should be damn sure that we vcan get the
> results back by other means, such as r() after an %r cell, or maxima() after
> a %maxima computation). Or should we try to return something ? If so, how ?
>
> We should also specify how to pass data to the called interpreter, and back
> to Sage. At least in the case of r(), it is hard to get anything else than
> string representations. One point of using %R is that this dat interchange
> problem is at least treated consistently.
>
> Note that the "coding problems" are not that small : ensuring that r(...),
> %r, %R and %%R "speak" to the same instance of R does not seem trivial (at
> least for my midget capacities)...
>>
>>
>>
>> Example for IPython default magic:
>>
>> sage: %system ls Makefile
>> ['Makefile']
>>
>>
>> Example for how we currently handle line magics on the commandline (BAD:
>> we just forget about the argument)
>>
>> sage: %gap 1+1;
>>
>>   --> Switching to Gap <--
>>
>> gap: ^C
>>
>>   --> Exiting back to Sage <--
>>
>>
>> Expected behavoir of %gap 1+1; is to be like
>>
>> sage: gap('1+1;')
>> 2
>>
>>
>> On the commandline ONLY line magics without arguments could still switch
>> the interface
>
>
> Should we open a ticket for this, or should we continue to discuss the
> specificatin here, opening tickets for the actual coding ?
>
> HTH (but doubting it...),
>
> --
> Emmanuel Charpentier
>
>>
>>
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop 

[sage-devel] Re: Jupyter notebook by default?

[Emmanuel Charpentier]

Le dimanche 3 janvier 2016 21:23:56 UTC+1, Volker Braun a écrit :
>
> On Sunday, January 3, 2016 at 5:36:30 PM UTC+1, Emmanuel Charpentier wrote:
>>
>> [ BTW : that's not really a bug, but rather a design conflict : the 
>> original %mode functions were designed to switch (for an indefinite scope) 
>> the behaviour of the REPL.This was transposed in the Sage notebook as what 
>> amounts to the equivalent of Jupyter's cell magics (scope defined as the 
>> current cell). Whereas line magics are, as far as I understand, 
>> Jupyter-specific... We can't be consistent across notebooks without 
>> redefining our "mode switch" magics as "cell magics", and rename them with 
>> "%%"...  ]
>>
>
> IMHO we should just switch the interface line magics to work like other 
> line magics in Jupyter/IPython, and also SMC: Take the rest of the line as 
> string and evaluate
>

As stated, that would break the compatibility with SMC magics that William 
just declared intangible (and always "cell magic"). This could be (more or 
less) restored  if the magic function had access to the current interface. 
Pseudo-Python ahead :

args, input=(RestOfTheFirstLline(), RestOfTheCell())
if(SMC):
ReturnAndPrint(magicfun(input))
if(commandline): ## In that case, by definition, input==None
if(restofthefirstline=""):
switchREPLuntilEOFOrOtherMagic(magicfun())
else:
ReturnAndPrint(magicfun(args))
else: ## Jupyter
ProcessOptions(args)
ReturnAndPrint(magicfun(input, ProcessedOptions))

The difficulty is, of course, that the same part of the command structure 
(the rest of the first line) would mean two different things :
- source code to execute in a command line
- possible options to be  passed to a cell magic in Jupyter (some 
interface, already supported (rpy2, the IRkernel for R) already use this 
convention !).

Furthermore, I supposed that SMC ignored the rest of the first line. This 
is what the %r magic does
%r print(2+2)
print(3+2)

## prints [1] 5, even if the pasted result is a bit more complicated :
︡3c5a5418-14e0-46fe-a34d-3fa7a0eb1b7d︡︡{"stdout":"[1] 
4","done":false}︡{"stdout":" 5\n","done":false}︡{"done":true}
︠68da460a-f070-4758-9fbb-667f85c4819es︠

but not the %maxima magic :


%maxima diff(log(x),x)
integrate(arctan(x),x)

## prints 1/x on one line, then x*arctan(x) - 1/2*log(x^2 + 1) on the next 
line, but again the reality is a bit more complicated :

︡410cbb81-a181-4e01-92a0-0c3e28da800a︡︡{"stdout":"1/x","done":false}︡{"stdout":"\n","done":false}︡{"stdout":"x*arctan(x)
 
- 1/2*log(x^2 + 1)\n","done":false}︡{"done":true}
︠e820fb8a-cd58-4173-9d99-95dd022e4a6f︠

Therefore, we have a specification/consistency problem, to be solved BEFORE 
the coding problem(s).

We should also specify what happens with the computed results : are they 
just displayed in the output cell and kept in the target interpreterv 
instance ? (In that case, we should be damn sure that we vcan get the 
results back by other means, such as r() after an %r cell, or maxima() 
after a %maxima computation). Or should we try to return something ? If so, 
how ?

We should also specify how to pass data to the called interpreter, and back 
to Sage. At least in the case of r(), it is hard to get anything else than 
string representations. One point of using %R is that this dat interchange 
problem is at least treated consistently.

Note that the "coding problems" are not that small : ensuring that r(...), 
%r, %R and %%R "speak" to the same instance of R does not seem trivial (at 
least for my midget capacities)...

>
>
> Example for IPython default magic:
>
> sage: %system ls Makefile 
> ['Makefile']
>
>
> Example for how we currently handle line magics on the commandline (BAD: 
> we just forget about the argument)
>
> sage: %gap 1+1;
>
>   --> Switching to Gap <--
>
> gap: ^C
>
>   --> Exiting back to Sage <--
>
>
> Expected behavoir of %gap 1+1; is to be like
>
> sage: gap('1+1;')
> 2
>
>
> On the commandline ONLY line magics without arguments could still switch 
> the interface
>

Should we open a ticket for this, or should we continue to discuss the 
specificatin here, opening tickets for the actual coding ?

HTH (but doubting it...),

--
Emmanuel Charpentier
 

>  
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

On Sunday, January 3, 2016 at 5:36:30 PM UTC+1, Emmanuel Charpentier wrote:
>
> [ BTW : that's not really a bug, but rather a design conflict : the 
> original %mode functions were designed to switch (for an indefinite scope) 
> the behaviour of the REPL.This was transposed in the Sage notebook as what 
> amounts to the equivalent of Jupyter's cell magics (scope defined as the 
> current cell). Whereas line magics are, as far as I understand, 
> Jupyter-specific... We can't be consistent across notebooks without 
> redefining our "mode switch" magics as "cell magics", and rename them with 
> "%%"...  ]
>

IMHO we should just switch the interface line magics to work like other 
line magics in Jupyter/IPython, and also SMC: Take the rest of the line as 
string and evaluate


Example for IPython default magic:

sage: %system ls Makefile 
['Makefile']


Example for how we currently handle line magics on the commandline (BAD: we 
just forget about the argument)

sage: %gap 1+1;

  --> Switching to Gap <--

gap: ^C

  --> Exiting back to Sage <--


Expected behavoir of %gap 1+1; is to be like

sage: gap('1+1;')
2


On the commandline ONLY line magics without arguments could still switch 
the interface
 

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

We can't (and shouldn't) change how Jupyter extracts arguments. Its either 
line magic

%magic one line of argument

or cell magic

%%magic
argument
more argument

Note that the "coding problems" are not that small : ensuring that r(...), 
> %r, %R and %%R "speak" to the same instance of R does not seem trivial
>
 
Thats the easy part, just call r(...) with the magic argument (as a string).

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

On Sunday, January 3, 2016 at 2:41:51 PM UTC+1, Emmanuel Charpentier wrote:
>
>
>- "other language" cells : %maxima and %r (possibly other interpreters 
>such as %octave) do not work as expected. Both %maxima and %r open new 
>instances of their respective interpreters and enter an REPL that cannot 
> be 
>exited (even with an explicit "quit();" or "q('no')". In other words, they 
>never return
>
> Interactive line magics obviously can't work in the browser like on the 
command line

The lack of appropriate cell magics (%%maxima)  is just an existing bug on 
the commandline.

Of course if you care about it then send in a patch.

On the plus side, the jupyter notebook comes with an official R kernel 
available as Sage optional package: sage -r r_jupyter.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Emmanuel Charpentier]

Le dimanche 3 janvier 2016 03:52:35 UTC+1, Jonathan a écrit :
>
> This looks like you are on the correct track.  I would just urge you to 
> indicate what features of Sagenb are not available yet.
>

A few things pop to my mind. I already signaled some f them (some might say 
I ranted about them), but I think that they are stumbling blocks.


   - "other language" cells : %maxima and %r (possibly other interpreters 
   such as %octave) do not work as expected. Both %maxima and %r open new 
   instances of their respective interpreters and enter an REPL that cannot be 
   exited (even with an explicit "quit();" or "q('no')". In other words, they 
   never return.
   - r("") doesn't display anything in the Jupyter 
   notebook, even when bracketed by r.png() and r.dev_off(). You have to save 
   (r.dev_copy()) the graph and display it in the notebook explicitely. 
   Annoying...

Rpy2 (part of Sage) offers the rpy2.ipython extension, which offer %R loine 
magic an %%R celle magic that work well (including graphics) ... in an R 
instance different of the one used by sage "r.xxx" functions. The data 
interchange  format is different of the one used by the current r interface.


   - The latex (mathjax) interface seems to have a few Jupyter-specific 
   quirks. For example, I have been unable to use a "$\frac{a}{b}$" tick 
   marker in a Sage jupyter worksheet. I suspect that the Jupyter notebook 
   doesn't use the (user-configurable) Latex configuration of Sage n the same 
   way as the Sage notebook does.
   - I noted that plot3d seems to call a Web-based  utility for some plots 
   (one sees messages related to loading data flashing in the region where the 
   jsmol (?) result will be displayed). This might implu that these finctions 
   might be unusable on an unconnected system (I didn't check that yet...).
   - In the current implementation, one cannot use some of the features the 
   default 3D renderer (jsmol ?). For example, when one attempts animation, a 
   right-click displays the relevant options, but one cannot activate them.

 

> I think things are close to the point that the only significant thing will 
> be interacts.  If plotly is installed, I think good interactive 2D and 3D 
> graphs will be available.
>


   - Interacts : In a Python worksheet, ipywidgets offers a nice 
   interactive framework, ... which turns out no be non obvious to use from a 
   Sage worksheet. A replacement and/or a compatibility layer are needed. 
   (Even if we offer a replacement, a compatibility layer might be useful to 
   ease transition/conversion of old sagenb worksheets).


Yet another interact interface ? That might be interesting if and only if 
the "new" interface has significantl enhancements over the "old" one...

HTH,

--
Emmanuel Charpentier
 

>
> Jonathan
>
> On Saturday, January 2, 2016 at 6:15:01 PM UTC-6, Volker Braun wrote:
>>
>> I'm adding a Jupyter notebook extension to the sagenb exporter (
>> https://github.com/vbraun/ExportSageNB) to hook into the GUI. So my 
>> proposal would be that the initial window will be something like this:
>>
>>
>>
>>
>>
>> 
>>
>>
>> On Friday, December 18, 2015 at 11:35:53 AM UTC+1, Jeroen Demeyer wrote:
>>>
>>> Should the Jupyter notebook be the default notebook for the next Sage 
>>> 7.0 release? I don't really have an opinion on the matter. 
>>>
>>>
>>> Pros: 
>>>
>>> * Nice tracebacks! 
>>> * The Jupyter notebook is a mature well-maintained project, unlike 
>>> SageNB. It is widely used in the "scientific Python" community. 
>>> * Availability of other Jupyter kernels besides Sage. 
>>>
>>>
>>> Cons: 
>>>
>>> * Less compatible with Sage: Sage interacts don't work, some graphics 
>>> don't work. 
>>> * Certain features of SageNB are missing: live documentation, 
>>> sharing/publishing of worksheets. 
>>> * It clutters the file system with .ipynb files. 
>>>
>>>
>>> Don't cares: 
>>>
>>> * It's just a default choice, both options remain available. 
>>>
>>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jeroen Demeyer]

On 2016-01-03 14:41, Emmanuel Charpentier wrote:

  * Interacts : In a Python worksheet, ipywidgets offers a nice
interactive framework, ... which turns out no be non obvious to use
from a Sage worksheet. A replacement and/or a compatibility layer
are needed. (Even if we offer a replacement, a compatibility layer
might be useful to ease transition/conversion of old sagenb worksheets).


During Sage Days 70, some work was done to make ipywidgets interacts 
usable within Sage. This now works using the git master for ipywidgets 
(which is not in Sage yet).


Unfortunately, the API of @interact has some significant differences 
between SageNB and Jupyter. IMHO, that of SageNB makes more sense but I 
have not been able to convince upstream Jupyter of this fact. Jupyter 
might have more fancy stuff (like a game controller widget), but fewer 
basic "just works" stuff.


I think that a compatibility layer might be possible, but I don't know 
if it can cover all features of SageNB interacts.


Jeroen.

--
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Emmanuel Charpentier]

Omitted from the previous ran^Kpost (and thus is probably still quite 
incomplete...) :


   - (Very ?) serious : no %cython cells, and no "obvious" shortcut to get 
   them. The "semi-obvious" replacement (%%writefile + load()) is not *that* 
   obvious to realize "cleanly" (no left-over files, etc...). a cell magic 
   would be (more than) nice
   - Annoying more than serious : no obvious way to get rid of the "sage:" 
   and "...  " prompts when cutting-and-pasting from examples. Not a problem 
   for Sage function, but this pasted medley of code and prompts won't load() 
   cleanly from a file ('indentation error').
   

HTH,

--
Emmanuel Charpentier

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jeroen Demeyer]

On 2016-01-03 17:36, Emmanuel Charpentier wrote:

how to define the function in the global namespace
as in the command line or Sage notebook... The relevant
compile_and_load() function returns a module (from which one can of
course import *, which would mimic the current Sage notebook behavior) ;
the point is to do this automatically from the magic function).


No, the point is to fix cython() such that this works *without* special 
support in the magic function.


This is part of http://trac.sagemath.org/ticket/18083

--
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Emmanuel Charpentier]

Dear Volker,

Le dimanche 3 janvier 2016 16:33:36 UTC+1, Volker Braun a écrit :
>
> On Sunday, January 3, 2016 at 2:41:51 PM UTC+1, Emmanuel Charpentier wrote:
>>
>>
>>- "other language" cells : %maxima and %r (possibly other 
>>interpreters such as %octave) do not work as expected. Both %maxima and 
>> %r 
>>open new instances of their respective interpreters and enter an REPL 
>> that 
>>cannot be exited (even with an explicit "quit();" or "q('no')". In other 
>>words, they never return
>>
>> Interactive line magics obviously can't work in the browser like on the 
> command line 
>

> The lack of appropriate cell magics (%%maxima)  is just an existing bug on 
> the commandline.
>

Indeed. We should have %%maxima, %%r , %%pari, etc...

[ BTW : that's not really a bug, but rather a design conflict : the 
original %mode functions were designed to switch (for an indefinite scope) 
the behaviour of the REPL.This was transposed in the Sage notebook as what 
amounts to the equivalent of Jupyter's cell magics (scope defined as the 
current cell). Whereas line magics are, as far as I understand, 
Jupyter-specific... We can't be consistent across notebooks without 
redefining our "mode switch" magics as "cell magics", and rename them with 
"%%"...  ]

>
> Of course if you care about it then send in a patch.
>

I do and I plan to. Therefore, I'm trying to understand how to add such 
magics to Jupyter. I'm currently swimming (drowning...) on my source tree 
in order to understand how this should be done. The difficulty is not the 
code itself, but how to package it as a magic.

Same question about %cython (which should become %%cython ?), with another 
difficulty : how to define the function in the global namespace as in the 
command line or Sage notebook... The relevant compile_and_load() function 
returns a module (from which one can of course import *, which would mimic 
the current Sage notebook behavior) ; the point is to do this automatically 
from the magic function). 

>
> On the plus side, the jupyter notebook comes with an official R kernel 
> available as Sage optional package: sage -r r_jupyter.
>

Indeed ! And that's appreciable (and apprecied). The same could be done for 
various  Sage-inclu(ded-dable) languages (pari, octave) for which a Jupyter 
kernel exists (see this list 
). 
Maxima is a special case : the existing Maxima-Jupyter kernel wont't adapt 
to our ECL-compiled implementation (I asked)...

BTW, your r_jupyter package doesn't appear in the output of sage -installed, 
nor in the output of sage -optional...

--
Emmanuel Charpentier

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Sun, Jan 3, 2016 at 8:36 AM, Emmanuel Charpentier
 wrote:
> Dear Volker,
>
> Le dimanche 3 janvier 2016 16:33:36 UTC+1, Volker Braun a écrit :
>>
>> On Sunday, January 3, 2016 at 2:41:51 PM UTC+1, Emmanuel Charpentier
>> wrote:
>>>
>>> "other language" cells : %maxima and %r (possibly other interpreters such
>>> as %octave) do not work as expected. Both %maxima and %r open new instances
>>> of their respective interpreters and enter an REPL that cannot be exited
>>> (even with an explicit "quit();" or "q('no')". In other words, they never
>>> return
>>
>> Interactive line magics obviously can't work in the browser like on the
>> command line
>>
>>
>> The lack of appropriate cell magics (%%maxima)  is just an existing bug on
>> the commandline.
>
>
> Indeed. We should have %%maxima, %%r , %%pari, etc...
>
> [ BTW : that's not really a bug, but rather a design conflict : the original
> %mode functions were designed to switch (for an indefinite scope) the
> behaviour of the REPL.This was transposed in the Sage notebook as what
> amounts to the equivalent of Jupyter's cell magics (scope defined as the
> current cell). Whereas line magics are, as far as I understand,
> Jupyter-specific... We can't be consistent across notebooks without
> redefining our "mode switch" magics as "cell magics", and rename them with
> "%%"...  ]

In case you are curious, in SageMathCloud worksheets a line mode is

%foo 

and a cell mode is

%foo(optional, arguments)


In SMC %foo is just foo('... rest of line or cell...'), with some
optional hooks (e.g., support foo.eval for backwards compat with
sage).  So a "magic" is much less magic -- just define any function
that takes a string as input, and you've written one.

I don't use %% anywhere.

I don't care what you do with Jupyter, but I'm sticking with this
design for cell modes, and definitely won't change.

 -- William

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

On Sunday, January 3, 2016 at 6:19:53 PM UTC+1, William wrote:
>
> In case you are curious, in SageMathCloud worksheets a line mode is 
>
> %foo  
>
> and a cell mode is 
>
> %foo(optional, arguments) 
>  


Thats almost the same except for the second percent sign in the cell magic

%%foo


FWIW the line magic in the command line (and by extension jupyter) is such 
that 

%foo
1+1

runs %foo (whatever that does) and then "1+1" as the subsequent Sage 
command. E.g.

sage: %cpaste
Pasting code; enter '--' alone on the line to stop or use Ctrl-D.
:%ls
:1+1
:--
aclocal.m4   bootstrap*  config/  config.status*  configure.ac 
 dist/  local/  m4/   README.txt  src/  upstream/
autom4te.cache/  build/  config.log@  configure*  COPYING.txt   
home/  logs/   Makefile  sage*   tmp/  VERSION.txt
2

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

You still haven't understood that this is about how to create new 
notebooks. We have the legacy support covered by shipping the old notebook. 
Also, I don't recall Microsoft bundling a free copy of word 2007 with word 
2010, you are comparing apples and oranges.

On Friday, January 1, 2016 at 4:01:18 PM UTC+1, kcrisman wrote:
>
> How many of those zillions of easy-to-search sample worksheets are about 
> something like calculating elliptic curve invariants or a nice demo for a 
> graph theory course?
>

And maybe we have the market for EC invariants saturated, and the reason 
for not growing any more is that we don't fit into the larger scientific 
Python ecosystem?

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

> You still haven't understood that this is about how to create new 
> notebooks. We have the legacy support covered by shipping the old notebook. 
> Also, I don't recall Microsoft bundling a free copy of word 2007 with word 
> 2010, you are comparing apples and oranges.
>
>
??

MS Word still supports pretty old .doc files - and cheerfully asks to 
upgrade them to .docx or whatever.  LibreOffice is even better - I can open 
ClarisWorks documents well over a decade old with it.  And I bet (though I 
haven't used Maple in many years) that when they switched to their 
Java-based "Clickable Calculus" thing (that was how it worked, right?) that 
they upgraded worksheets for use in the new system.  Ditto for Mma?

That's what I'm talking about.  "Legacy support" for us cannot be about 
treating people with a lot of sagenb worksheets like a company with COBOL 
programs that need continued support but will not have new stuff written 
from scratch in.  If Jupyter really is the future, and Sage really wants to 
be professional about a transition to it, it needs to provide an *easy* and 
*effective* upgrade path for worksheets created in sagenb (and, presumably, 
in sagews format, I don't know if they convert?) to that.  (See the rst2sws 
and sws2rst and tex2sws and sws2tex for some interesting conversions in 
addition to the several successive conversions William has written over the 
years.)

Think of all of the people we have not been able to gain to Sage over the 
years because we don't have a simple Maple -> Sage or Mma -> Sage 
converter.  (I'm not complaining about that! I don't think such people 
recognize the technical hurdle there.  But if there were such things, we 
would have a lot more users, I think it's fairly clear.)  Similarly, not 
having a sagenb -> ipynb (or whatever the future brings) converter that is 
reliable may cause people to just go back to whatever they were using 
beforehand, or to say the heck with it.

(By the way, as to location - one could let the user decide, with a default 
of ~/MyOldSageWorksheets or some such nonsense, if there was an interactive 
GUI or script or web thing on the sagenb or sage or jupyter side.  No need 
to have a specific default because not everyone will upgrade.)

Otherwise, I don't want to hear any complaints about people sticking with 
proprietary solutions that do cover such upgrades.  The reality is that 
most people stick with the devil they know.  This isn't about technical 
superiority; it's about treating our users as if they have exactly enough 
time to use Sage, and very little to do anything else regarding upgrades or 
whatever.   Because who will guarantee that in five years there won't be 
some brand-new toy instead of Jupyter?  We already see Julia and Nemo as 
the hot new thing (in a different direction than the notebooks, of course); 
what can we predict for the next five years in terms of anything?  (Perhaps 
OpenDreamKit will help in that regard.)

By the way, 'whatever' means not just 'upgrades to notebooks' but also 
'hard documentation to search' and 'I still don't know if Sage works in El 
Capitan' and lots of other things I get personal message about fairly 
regularly that I don't necessarily pass on here.  This is just a symptom of 
the usual F/LOSS problem of being written largely for the developers - 
which is way, way less fair to say about Sage than many other projects, due 
largely to some of the very people I'm arguing with in this thread!  But we 
still have the same symptoms, even if it's a relatively mild infection. 
 (Sorry, I just got over a pretty bad case of the flu so I've been using 
that kind of analogy too much.)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Sat, Jan 2, 2016 at 10:41 AM, kcrisman  wrote:
>
>> You still haven't understood that this is about how to create new
>> notebooks. We have the legacy support covered by shipping the old notebook.
>> Also, I don't recall Microsoft bundling a free copy of word 2007 with word
>> 2010, you are comparing apples and oranges.
>>
>
> ??
>
> MS Word still supports pretty old .doc files - and cheerfully asks to
> upgrade them to .docx or whatever.  LibreOffice is even better - I can open
> ClarisWorks documents well over a decade old with it.  And I bet (though I
> haven't used Maple in many years) that when they switched to their
> Java-based "Clickable Calculus" thing (that was how it worked, right?) that
> they upgraded worksheets for use in the new system.  Ditto for Mma?
>
> That's what I'm talking about.  "Legacy support" for us cannot be about
> treating people with a lot of sagenb worksheets like a company with COBOL
> programs that need continued support but will not have new stuff written
> from scratch in.  If Jupyter really is the future, and Sage really wants to
> be professional about a transition to it, it needs to provide an *easy* and
> *effective* upgrade path for worksheets created in sagenb (and, presumably,
> in sagews format, I don't know if they convert?) to that.  (See the rst2sws
> and sws2rst and tex2sws and sws2tex for some interesting conversions in
> addition to the several successive conversions William has written over the
> years.)
>
> Think of all of the people we have not been able to gain to Sage over the
> years because we don't have a simple Maple -> Sage or Mma -> Sage converter.
> (I'm not complaining about that! I don't think such people recognize the
> technical hurdle there.  But if there were such things, we would have a lot
> more users, I think it's fairly clear.)  Similarly, not having a sagenb ->
> ipynb (or whatever the future brings) converter that is reliable may cause
> people to just go back to whatever they were using beforehand, or to say the
> heck with it.
>
> (By the way, as to location - one could let the user decide, with a default
> of ~/MyOldSageWorksheets or some such nonsense, if there was an interactive
> GUI or script or web thing on the sagenb or sage or jupyter side.  No need
> to have a specific default because not everyone will upgrade.)
>
> Otherwise, I don't want to hear any complaints about people sticking with
> proprietary solutions that do cover such upgrades.  The reality is that most
> people stick with the devil they know.  This isn't about technical
> superiority; it's about treating our users as if they have exactly enough
> time to use Sage, and very little to do anything else regarding upgrades or
> whatever.   Because who will guarantee that in five years there won't be
> some brand-new toy instead of Jupyter?  We already see Julia and Nemo as the
> hot new thing (in a different direction than the notebooks, of course); what
> can we predict for the next five years in terms of anything?  (Perhaps
> OpenDreamKit will help in that regard.)
>
> By the way, 'whatever' means not just 'upgrades to notebooks' but also 'hard
> documentation to search' and 'I still don't know if Sage works in El
> Capitan' and lots of other things I get personal message about fairly
> regularly that I don't necessarily pass on here.  This is just a symptom of
> the usual F/LOSS problem of being written largely for the developers - which
> is way, way less fair to say about Sage than many other projects,

Karl,

It's not "the usual F/LOSS problem", it's the usual problem of a
volunteer project with almost no funding.  There is a ton of high
quality well funded FLOSS software out there that doesn't have this
problem (e.g., Firefox, Chromium, parts of OS X, the swift language,
...)

For better or worse, I think we just have to acknowledge the reality of the
situation, and move on and do a very bad job in the above
regards compared to the well funded alternatives until/if we do
eventually get funding.  Right now, it is best of the Sage project
focuses squarely on its core competence, which is research level pure
mathematics, (and I hope also undergrad math teaching).

I think there is no question that if some group popped up and
implemented all the automated backwards compatibility, and other nice
things that you describe above, all very good quality, and with proper
review on trac, then Volker would smile and merge it in seconds.
Implementing this stuff we want is precisely what is nearly
impossibly hard for any sane research mathematician to justify
spending serious time on.

For now, let's just move on, make Jupyter (which just got $6,000,000
in funding recently) the default, and deal with the pain and continued
failure to grow.   Maybe someday things will be different.

 -- William




-- 
William (http://wstein.org)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from 

Re: [sage-devel] Re: Jupyter notebook by default?

[Jonathan]

William, Volker, et al...

I think you misunderstand Karl.  He has stated what is ideal, but I think 
from his messages that he understands we do not have the manpower within 
the group of people who can commit time to Sage to make his ideal happen.

What he is asking is that you avoid doing this transition in a manner that 
blindsides many people who are using Sage (especially for teaching using 
servers they maintain).  If ipynb is going to be the default it needs to 
launch the first time with messages on the command line and in the browser 
that explain how to use the Old notebook if necessary and with some advice 
on conversion, if any exists.  It would also be nice to have a brief list 
of key notebook features that do not yet work in ipynb.  I suggest language 
like:

WARNING!!  If you have been using Sagemath using the Sagenotebook (default 
prior to ) for interactive Sagemath sessions in a browser, you should 
be aware that Sagenotebook is no longer the default.  The default is ipynb 
(web link..).  You can launch the older Sagenotebook as described below.  
However, unless you need the following features which do not yet exist in 
ipynb we recommend you use the new ipynb, as it has better support: ___list 
of missing features___.  To use the the old default ...

Regards,
Jonathan
On Saturday, January 2, 2016 at 1:13:58 PM UTC-6, William wrote:
>
>
> Karl, 
>
> It's not "the usual F/LOSS problem", it's the usual problem of a 
> volunteer project with almost no funding.  There is a ton of high 
> quality well funded FLOSS software out there that doesn't have this 
> problem (e.g., Firefox, Chromium, parts of OS X, the swift language, 
> ...) 
>
> For better or worse, I think we just have to acknowledge the reality of 
> the 
> situation, and move on and do a very bad job in the above 
> regards compared to the well funded alternatives until/if we do 
> eventually get funding.  Right now, it is best of the Sage project 
> focuses squarely on its core competence, which is research level pure 
> mathematics, (and I hope also undergrad math teaching). 
>
> I think there is no question that if some group popped up and 
> implemented all the automated backwards compatibility, and other nice 
> things that you describe above, all very good quality, and with proper 
> review on trac, then Volker would smile and merge it in seconds. 
> Implementing this stuff we want is precisely what is nearly 
> impossibly hard for any sane research mathematician to justify 
> spending serious time on. 
>
> For now, let's just move on, make Jupyter (which just got $6,000,000 
> in funding recently) the default, and deal with the pain and continued 
> failure to grow.   Maybe someday things will be different. 
>
>  -- William 
>
>
>
>
> -- 
> William (http://wstein.org) 
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

I'm adding a Jupyter notebook extension to the sagenb exporter 
(https://github.com/vbraun/ExportSageNB) to hook into the GUI. So my 
proposal would be that the initial window will be something like this:







On Friday, December 18, 2015 at 11:35:53 AM UTC+1, Jeroen Demeyer wrote:
>
> Should the Jupyter notebook be the default notebook for the next Sage 
> 7.0 release? I don't really have an opinion on the matter. 
>
>
> Pros: 
>
> * Nice tracebacks! 
> * The Jupyter notebook is a mature well-maintained project, unlike 
> SageNB. It is widely used in the "scientific Python" community. 
> * Availability of other Jupyter kernels besides Sage. 
>
>
> Cons: 
>
> * Less compatible with Sage: Sage interacts don't work, some graphics 
> don't work. 
> * Certain features of SageNB are missing: live documentation, 
> sharing/publishing of worksheets. 
> * It clutters the file system with .ipynb files. 
>
>
> Don't cares: 
>
> * It's just a default choice, both options remain available. 
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Jonathan]

This looks like you are on the correct track.  I would just urge you to 
indicate what features of Sagenb are not available yet.  I think things are 
close to the point that the only significant thing will be interacts.  If 
plotly is installed, I think good interactive 2D and 3D graphs will be 
available.

Jonathan

On Saturday, January 2, 2016 at 6:15:01 PM UTC-6, Volker Braun wrote:
>
> I'm adding a Jupyter notebook extension to the sagenb exporter (
> https://github.com/vbraun/ExportSageNB) to hook into the GUI. So my 
> proposal would be that the initial window will be something like this:
>
>
>
>
>
> 
>
>
> On Friday, December 18, 2015 at 11:35:53 AM UTC+1, Jeroen Demeyer wrote:
>>
>> Should the Jupyter notebook be the default notebook for the next Sage 
>> 7.0 release? I don't really have an opinion on the matter. 
>>
>>
>> Pros: 
>>
>> * Nice tracebacks! 
>> * The Jupyter notebook is a mature well-maintained project, unlike 
>> SageNB. It is widely used in the "scientific Python" community. 
>> * Availability of other Jupyter kernels besides Sage. 
>>
>>
>> Cons: 
>>
>> * Less compatible with Sage: Sage interacts don't work, some graphics 
>> don't work. 
>> * Certain features of SageNB are missing: live documentation, 
>> sharing/publishing of worksheets. 
>> * It clutters the file system with .ipynb files. 
>>
>>
>> Don't cares: 
>>
>> * It's just a default choice, both options remain available. 
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[David Roe]

On Fri, Jan 1, 2016 at 7:01 AM, kcrisman  wrote:

>
>
> And we are not removing it for all the legacy reasons that you listed.
>> This thread is about what format new notebooks are supposed to be. Anybody
>> who prefers SageNB can still use SageNB
>>
>>
> But how easy will that be?  It already sounds like sagenb notebooks are
> not really portable to Jupyter.
>
>
>> But the fact is: Its unmaintained/-able, and the longer we wait the more
>> painful it'll be to switch. How is waiting another year going to help?
>> Because by then there will be another book explaining how to use SageNB? Do
>> you think the book author is going to be happy that we never made it clear
>> that SageNB is on the way out? SageNB is an evolutionary dead end, and the
>> only question is how deep into that hole do we want to go.
>>
>
> As William said, the answer is to either find a way to make them port, or
> to find a way to gracefully detect that someone has existing sagenb
> worksheets (perhaps slightly more gracefully than just checking DOT_SAGENB
> since that doesn't mean they actually did anything) and either ask to
> upgrade to Jupyter worksheets, if that's possible, or make it very easy to
> change a default setting.
>

If we change the default to Jupyter, detecting existing Sage notebooks and
presenting a choice to the user to upgrade their existing notebooks is a
great idea.
David


>
> Could there be an environment variable DEFAULT_SAGE_NOTEBOOK that one
> could have incredibly explicit instructions for how to use?  (Would making
> an alias work in this context?  Not in the sage: notebook() situation, I
> guess.)
>
> But it has to be idiot-proof.  Imagine the following conversation at the
> Sage table:
>
> Customer (yes they are customers!): Will I ever have any problems with
> upgrading?
> Sage rep: Yep, we will make it so all your old stuff is never upgradeable
> with essentially no warning other than some cryptic command line thing you
> don't even see because you only look at your browser!
> Customer: I didn't hear you right, I think? You mean, "we'll support
> legacy stuff so long it will make Microsoft look good."  I mean, I don't
> want to still be using .doc in 2020.
> Sage rep: Nope, we're sort of the opposite; we won't have any
> 'compatibility mode' nor any incentive to upgrade your old documents; you
> either go to the new one - oh, by the way, it doesn't do everything the old
> one does yet - or you just stick with the old kind forever.
>
> Yup, I see that as a real winner.  And of course this conversation can
> happen again whenever the next new hit comes out in five years, which it no
> doubt will.
>
> Writing up a very accessible (perhaps even linked to in the warning)
> translation guide to getting started in Jupyter would be good too.  As much
> as there might be a bigger ecosystem for them, that is not a very
> compelling argument to me; otherwise we should all be using RStudio and
> shiny, and well basically just using R for everything.  How many of those
> zillions of easy-to-search sample worksheets are about something like
> calculating elliptic curve invariants or a nice demo for a graph theory
> course?  Probably just a tiny proportion of the ones about data science or
> scientific computing.
>
> On a separate note, I'm wondering how changing the default will impact
> those running servers.Which is a higher knowledge crowd but also one
> who might not be reading sage-devel (in fact likely isn't).  Basically, I'm
> asking about any worst-case scenarios when someone upgrades from (say) 6.7
> to 7.2 at the end of a school year and either doesn't read the 'release
> notes', doesn't even know there are release notes, or our release notes are
> so minimal as to be useless, or something.  (I assume that commercial
> software has that problem of non-reading too.)
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

And we are not removing it for all the legacy reasons that you listed. This 
> thread is about what format new notebooks are supposed to be. Anybody who 
> prefers SageNB can still use SageNB
>
>
But how easy will that be?  It already sounds like sagenb notebooks are not 
really portable to Jupyter.
 

> But the fact is: Its unmaintained/-able, and the longer we wait the more 
> painful it'll be to switch. How is waiting another year going to help? 
> Because by then there will be another book explaining how to use SageNB? Do 
> you think the book author is going to be happy that we never made it clear 
> that SageNB is on the way out? SageNB is an evolutionary dead end, and the 
> only question is how deep into that hole do we want to go.
>

As William said, the answer is to either find a way to make them port, or 
to find a way to gracefully detect that someone has existing sagenb 
worksheets (perhaps slightly more gracefully than just checking DOT_SAGENB 
since that doesn't mean they actually did anything) and either ask to 
upgrade to Jupyter worksheets, if that's possible, or make it very easy to 
change a default setting.

Could there be an environment variable DEFAULT_SAGE_NOTEBOOK that one could 
have incredibly explicit instructions for how to use?  (Would making an 
alias work in this context?  Not in the sage: notebook() situation, I 
guess.)

But it has to be idiot-proof.  Imagine the following conversation at the 
Sage table:

Customer (yes they are customers!): Will I ever have any problems with 
upgrading?
Sage rep: Yep, we will make it so all your old stuff is never upgradeable 
with essentially no warning other than some cryptic command line thing you 
don't even see because you only look at your browser!
Customer: I didn't hear you right, I think? You mean, "we'll support legacy 
stuff so long it will make Microsoft look good."  I mean, I don't want to 
still be using .doc in 2020.
Sage rep: Nope, we're sort of the opposite; we won't have any 
'compatibility mode' nor any incentive to upgrade your old documents; you 
either go to the new one - oh, by the way, it doesn't do everything the old 
one does yet - or you just stick with the old kind forever.

Yup, I see that as a real winner.  And of course this conversation can 
happen again whenever the next new hit comes out in five years, which it no 
doubt will.

Writing up a very accessible (perhaps even linked to in the warning) 
translation guide to getting started in Jupyter would be good too.  As much 
as there might be a bigger ecosystem for them, that is not a very 
compelling argument to me; otherwise we should all be using RStudio and 
shiny, and well basically just using R for everything.  How many of those 
zillions of easy-to-search sample worksheets are about something like 
calculating elliptic curve invariants or a nice demo for a graph theory 
course?  Probably just a tiny proportion of the ones about data science or 
scientific computing.

On a separate note, I'm wondering how changing the default will impact 
those running servers.Which is a higher knowledge crowd but also one 
who might not be reading sage-devel (in fact likely isn't).  Basically, I'm 
asking about any worst-case scenarios when someone upgrades from (say) 6.7 
to 7.2 at the end of a school year and either doesn't read the 'release 
notes', doesn't even know there are release notes, or our release notes are 
so minimal as to be useless, or something.  (I assume that commercial 
software has that problem of non-reading too.)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

On Thursday, December 31, 2015 at 2:43:56 AM UTC+1, kcrisman wrote:
>
> Regardless of how "useful" or awesome the Jupyter notebook is and how many 
> people in scientific computing are using it, there is a substantial 
> ecosystem now designed around the Sage notebook proper
>

And we are not removing it for all the legacy reasons that you listed. This 
thread is about what format new notebooks are supposed to be. Anybody who 
prefers SageNB can still use SageNB

But the fact is: Its unmaintained/-able, and the longer we wait the more 
painful it'll be to switch. How is waiting another year going to help? 
Because by then there will be another book explaining how to use SageNB? Do 
you think the book author is going to be happy that we never made it clear 
that SageNB is on the way out? SageNB is an evolutionary dead end, and the 
only question is how deep into that hole do we want to go.


-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[kcrisman]

Regardless of how "useful" or awesome the Jupyter notebook is and how many 
people in scientific computing are using it, there is a substantial 
ecosystem now designed around the Sage notebook proper - including perhaps 
even the not-yet-one-year-old "Sage for Undergraduates" book? - and a lot 
of people "know" how to use it to do various things. Some of the below were 
mentioned above.
 * Access to documentation
 * Access to live documentation
 * things like %auto, %hide - presumably some of these are in Jupyter
 * difference in @interact API
 * It can't handle attached files properly, really? I was surprised to read 
that
 * Different way to create math markup/'text' cells
I'll also point out that William apparently felt it was useful to create a 
.sagews format rather than just have ipynb be the default in SMC, which 
presumably has different features?  Or was that only because Jupyter was 
not as good when he started SMC?

Another issue would be how to deal with the Mac app, what its default 
should be, etc.  Similarly for the virtual Windows solution - will people 
even be able to find their stuff?

> I clearly don't agree with that, since that's never been the case before, 
as I mentioned.  I'll not press this any further. 

I'll press it further: not doing this is basically flipping the bird to a 
lot of long-time users.  I don't mean this sarcastically.

> >  And thats not going to get easier by delaying when we change the 
defaults. 

> It will be much easier for our users to actually *use* a converter if we 
delay changing the defaults until after such a converter is written. 

Uh, duh.

All these comments are predicated upon a "strong" change of default.  In a 
normal user interface (to me, as a user) there would be a really easy way 
to decide *not* to change the default when first downloading/upgrading.  I 
think someone mentioned this before as well.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Nils Bruin]

On Sunday, December 20, 2015 at 10:40:23 PM UTC-10, Volker Braun wrote:
>
> On Monday, December 21, 2015 at 5:16:44 AM UTC+1, Nils Bruin wrote:
>>
>> different notebook. We could possibly alleviate the problem of lack of 
>> information by somehow guessing if this is the first time this user starts 
>> up sage ipython notebook
>>
>
> Thats inconsistent UX, now you open different notebooks on different 
> computers / accounts. Unless you mean: print help to the terminal. The 
> ticket does that already.
>

The reason why I expect that  printing on the terminal is insufficient, is 
because most notebook users will not look there at all: they're just 
looking at the browser that pops up. There are quite some precedents for 
informational pop-ups: firefox and chrome are usually quite insistent on 
start-up to ask you about migrating information to themselves and/or 
selecting them as default browser.

We could make the UX more consistent by "preinstalling" a migration ipy 
notebook. In fact, that notebook could include code to open and read the 
sagenb worksheets (jupyter kernels don't run chrooted by default, right? so 
their code can still access HOME/.sagenb) and help with translating the 
worksheets.

The main scenario is that of single-user notebooks. multi-user setups don't 
translate easily to jupyter anyway, so a more manual translation approach 
will be required anyway.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

On Tuesday, December 22, 2015 at 9:16:55 PM UTC+1, Nils Bruin wrote:
>
> We could make the UX more consistent by "preinstalling" a migration ipy 
> notebook. 
>

Into which directory are we going to put that MigrateFromSagenb.ipynb file?
 

> In fact, that notebook could include code to open and read the sagenb 
> worksheets (jupyter kernels don't run chrooted by default, right? so their 
> code can still access HOME/.sagenb)
>

The kernel and the file browser ui are two separate components, potentially 
running on separate computers. The kernel can execute arbitrary code (and 
access ~/.sage if the OS allows), the file browser is restricted to the 
directory tree below the startup directory. 


-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Jeroen Demeyer]

On 2015-12-21 09:40, Volker Braun wrote:

Interacts DO work. Its just Jupyter interacts and not SageNB interacts.
I think that the SageNB @interact API is much better designed and more 
consistent than the Jupyter @interact API. Unfortunately, I am not able 
to convince upstream Jupyter (really: ipywidgets) that this is the case. 
See https://github.com/ipython/ipywidgets/issues/238


It's true that ipywidgets might be able to do more fancy stuff (like 
support for game controllers), but the basic stuff works better in SageNB.



Jeroen.

--
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

On Monday, December 21, 2015 at 5:16:44 AM UTC+1, Nils Bruin wrote:
>
> different notebook. We could possibly alleviate the problem of lack of 
> information by somehow guessing if this is the first time this user starts 
> up sage ipython notebook
>

Thats inconsistent UX, now you open different notebooks on different 
computers / accounts. Unless you mean: print help to the terminal. The 
ticket does that already.

If the default changes, then most users will assume they are supposed to 
> change with it (and we should only do this once we think users should 
> change, e.g., when interacts work) 
>

Interacts DO work. Its just Jupyter interacts and not SageNB interacts. 
There are many more published Jupyter interacts, and lots more 
documentation about them.
 
Not changing the default also sends a message to our users, namely that its 
a good idea to use SageNB for new projects and that it is the UI that we 
are actively maintaining.

At the very least there should be a method to:
>  - start the classic notebook
>

This whole thread is ONLY about changing the default, nobody was talking 
about removing SageNB

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[John H Palmieri]

On Sunday, December 20, 2015 at 8:16:44 PM UTC-8, Nils Bruin wrote:
>
> On Sunday, December 20, 2015 at 2:12:57 PM UTC-10, William wrote:
>>
>> It will be much easier for our users to actually *use* a converter if 
>> we delay changing the defaults until after such a converter is 
>> written. 
>>
>
> ++1 (OK, I guess that's an error in most language). 
>
> It's going to be a relatively unpleasant surprise for people if they start 
> up the "sage notebook" and find a different notebook. We could possibly 
> alleviate the problem of lack of information by somehow guessing if this is 
> the first time this user starts up sage ipython notebook (and perhaps only 
> if we detect there is evince of a sagenb) and startup with a document that 
> explains how to do migration.
>
> If the default changes, then most users will assume they are supposed to 
> change with it (and we should only do this once we think users should 
> change, e.g., when interacts work) 
>
> At the very least there should be a method to:
>  - start the classic notebook
>  - save a collection of worksheets (sagenb)
>  - convert them to ipynb, at least in a way that works for "simple" 
> worksheets. Attached files etc. are obviously going to be an issue.
>
> If we don't have a tool that makes it straightforward to do this migration 
> then we're not ready to change defaults, in my opinion.
>
>

I think we need a migration tool and a deprecation warning: "The old Sage 
notebook is deprecated in favor of Jupyter notebooks. For information about 
converting old-style notebooks to Jupyter notebooks, see ... To use Jupyter 
notebooks now, run 'sage -notebook=jupyter'." Then the next version of Sage 
can change the default behavior.

-- 
John

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Sun, Dec 20, 2015 at 4:06 PM, Volker Braun  wrote:
> On Sunday, December 20, 2015 at 10:27:39 PM UTC+1, William wrote:
>>
>> Notebooks in sagenb are basically individual files; it's just that
>> they are stored as a directory tree in the filesystem.
>
>
> Yes, what I meant was: SageNB doesn't show a filesystem view so nobody is
> going to know *where* they are stored in the file system.
>
> Whereas Jupyter notebooks are always displayed as files. And the average
> user isn't going to find ~/.sage/jupyter with the jupyter file browser. In
> fact, if that is not a subdirectory of the directory where Jupyter is
> launched then you cannot open it. And you don't want a ton of converted
> notebooks dumped in the current directory either. Realistically, any
> notebook converter will have to be manually invoked.

I clearly don't agree with that, since that's never been the case
before, as I mentioned.  I'll not press this any further.

>  And thats not going to
> get easier by delaying when we change the defaults.

It will be much easier for our users to actually *use* a converter if
we delay changing the defaults until after such a converter is
written.

>
>>
>> https://github.com/sagemathinc/smc/blob/master/src/smc_pyutil/smc_pyutil/sws2sagews.py
>
>
> And its easy to programmatically generate new IPython notebooks:
> http://nbviewer.ipython.org/gist/fperez/9716279
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.



-- 
William (http://wstein.org)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Nils Bruin]

On Sunday, December 20, 2015 at 2:12:57 PM UTC-10, William wrote:
>
> It will be much easier for our users to actually *use* a converter if 
> we delay changing the defaults until after such a converter is 
> written. 
>

++1 (OK, I guess that's an error in most language). 

It's going to be a relatively unpleasant surprise for people if they start 
up the "sage notebook" and find a different notebook. We could possibly 
alleviate the problem of lack of information by somehow guessing if this is 
the first time this user starts up sage ipython notebook (and perhaps only 
if we detect there is evince of a sagenb) and startup with a document that 
explains how to do migration.

If the default changes, then most users will assume they are supposed to 
change with it (and we should only do this once we think users should 
change, e.g., when interacts work) 

At the very least there should be a method to:
 - start the classic notebook
 - save a collection of worksheets (sagenb)
 - convert them to ipynb, at least in a way that works for "simple" 
worksheets. Attached files etc. are obviously going to be an issue.

If we don't have a tool that makes it straightforward to do this migration 
then we're not ready to change defaults, in my opinion.

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

On Sunday, December 20, 2015 at 10:27:39 PM UTC+1, William wrote:
>
> Notebooks in sagenb are basically individual files; it's just that 
> they are stored as a directory tree in the filesystem. 


Yes, what I meant was: SageNB doesn't show a filesystem view so nobody is 
going to know *where* they are stored in the file system.

Whereas Jupyter notebooks are always displayed as files. And the average 
user isn't going to find ~/.sage/jupyter with the jupyter file browser. In 
fact, if that is not a subdirectory of the directory where Jupyter is 
launched then you cannot open it. And you don't want a ton of converted 
notebooks dumped in the current directory either. Realistically, any 
notebook converter will have to be manually invoked. And thats not going to 
get easier by delaying when we change the defaults.


 
> https://github.com/sagemathinc/smc/blob/master/src/smc_pyutil/smc_pyutil/sws2sagews.py
>  


And its easy to programmatically generate new IPython notebooks: 
 http://nbviewer.ipython.org/gist/fperez/9716279

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

If you start Jupyter in ~/foo then you cannot open notebooks in 
~/.sage/jupyter. Thats just how it is.

On Monday, December 21, 2015 at 1:12:57 AM UTC+1, William wrote:
>
> I clearly don't agree with that, since that's never been the case 
> before, as I mentioned.  I'll not press this any further. 
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Sun, Dec 20, 2015 at 12:47 PM, Volker Braun  wrote:
> The ticket already has a notice printed that tells you how to run the old
> SageNB
>
> Realistically, since we are switching to a model where notebooks are
> individual files there is always going to be somebody who won't read the
> hints. We'll have to do *that* switch anyways and it isn't going to be any
> easier if we wait.

Notebooks in sagenb are basically individual files; it's just that
they are stored as a directory tree in the filesystem.  The .sws file
that you get when you click "File --> Download" in sagenb is a file.
It represents almost exactly the same information as an ipynb file.
It is feasible to run a script that you automatically convert the
~/.sage/sagenb* (or whatever) directory into a collection of Jupyter
notebooks, say ~/.sage/jupyter.   This would make life much easier for
users who have collections of notebooks.  The hard part is
functionality that doesn't map from sagenb to jupyter, e.g.,
interacts, % modes, etc., etc.,

I wrote an sws parser, which is here in case anybody takes up
automating the above conversion (I won't):

 
https://github.com/sagemathinc/smc/blob/master/src/smc_pyutil/smc_pyutil/sws2sagews.py

The


>
>
> On Sunday, December 20, 2015 at 7:52:46 PM UTC+1, Dima Pasechnik wrote:
>>
>>
>>
>> On Friday, 18 December 2015 15:27:29 UTC, Volker Braun wrote:
>>>
>>> IMHO the Jupyter/IPython notebook is a clear winner, no real contest.
>>>
>>> Nobody prevents you form using SageNB for existing worksheets or the live
>>> documentation; But if you start a new worksheet today then you really should
>>> be using Jupyter.
>>>
>>>
>>> On Friday, December 18, 2015 at 11:35:53 AM UTC+1, Jeroen Demeyer wrote:

 * Certain features of SageNB are missing: live documentation,
 sharing/publishing of worksheets.
>>>
>>>
>>> IMHO sharing is already better, you can just upload the ipynb file
>>> anywhere (including http://gist.github.com or any other pastebin) and then
>>> share it using the http://nbviewer.ipython.org service. For example:
>>>
>>> http://nbviewer.ipython.org/url/sagepad.org/pub/sage-test.ipynb
>>
>>
>> Some care will need to be taken, so that we don't have to deal with
>> complaints of kind
>>
>>   " I run 'sage -n', and all my worksheets are gone!!! Help!!! "
>>
>>
>>>
>>>
>>>
 * It clutters the file system with .ipynb files.
>>>
>>>
>>> and .ipynb_checkpoints hidden directories. See also:
>>> http://trac.sagemath.org/19746
>>>
>>>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.



-- 
William (http://wstein.org)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Dima Pasechnik]

On Friday, 18 December 2015 15:27:29 UTC, Volker Braun wrote:
>
> IMHO the Jupyter/IPython notebook is a clear winner, no real contest.
>
> Nobody prevents you form using SageNB for existing worksheets or the live 
> documentation; But if you start a new worksheet today then you really 
> should be using Jupyter.
>
>
> On Friday, December 18, 2015 at 11:35:53 AM UTC+1, Jeroen Demeyer wrote:
>>
>> * Certain features of SageNB are missing: live documentation, 
>> sharing/publishing of worksheets. 
>>
>
> IMHO sharing is already better, you can just upload the ipynb file 
> anywhere (including http://gist.github.com or any other pastebin) and 
> then share it using the http://nbviewer.ipython.org service. For example:
>
> http://nbviewer.ipython.org/url/sagepad.org/pub/sage-test.ipynb
>

Some care will need to be taken, so that we don't have to deal with 
complaints of kind

  " I run 'sage -n', and all my worksheets are gone!!! Help!!! "



>
>
> * It clutters the file system with .ipynb files. 
>>
>
> and .ipynb_checkpoints hidden directories. See also: 
> http://trac.sagemath.org/19746
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Sun, Dec 20, 2015 at 10:52 AM, Dima Pasechnik  wrote:
>
>
> On Friday, 18 December 2015 15:27:29 UTC, Volker Braun wrote:
>>
>> IMHO the Jupyter/IPython notebook is a clear winner, no real contest.
>>
>> Nobody prevents you form using SageNB for existing worksheets or the live
>> documentation; But if you start a new worksheet today then you really should
>> be using Jupyter.
>>
>>
>> On Friday, December 18, 2015 at 11:35:53 AM UTC+1, Jeroen Demeyer wrote:
>>>
>>> * Certain features of SageNB are missing: live documentation,
>>> sharing/publishing of worksheets.
>>
>>
>> IMHO sharing is already better, you can just upload the ipynb file
>> anywhere (including http://gist.github.com or any other pastebin) and then
>> share it using the http://nbviewer.ipython.org service. For example:
>>
>> http://nbviewer.ipython.org/url/sagepad.org/pub/sage-test.ipynb
>
>
> Some care will need to be taken, so that we don't have to deal with
> complaints of kind
>
>   " I run 'sage -n', and all my worksheets are gone!!! Help!!! "

With every single upgrade of sagenb in the past, there was always an
automatic and safe migration step that would take care of this
problem, even if the notebook format changed completely (which
happened).   I always wrote that code to do the migration, even though
it was a lot of work.  Somebody should write something like that for
sagenb to jupyter.

 -- William

>
>
>>
>>
>>
>>> * It clutters the file system with .ipynb files.
>>
>>
>> and .ipynb_checkpoints hidden directories. See also:
>> http://trac.sagemath.org/19746
>>
>>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.



-- 
William (http://wstein.org)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

The ticket already has a notice printed that tells you how to run the old 
SageNB

Realistically, since we are switching to a model where notebooks are 
individual files there is always going to be somebody who won't read the 
hints. We'll have to do *that* switch anyways and it isn't going to be any 
easier if we wait.


On Sunday, December 20, 2015 at 7:52:46 PM UTC+1, Dima Pasechnik wrote:
>
>
>
> On Friday, 18 December 2015 15:27:29 UTC, Volker Braun wrote:
>>
>> IMHO the Jupyter/IPython notebook is a clear winner, no real contest.
>>
>> Nobody prevents you form using SageNB for existing worksheets or the live 
>> documentation; But if you start a new worksheet today then you really 
>> should be using Jupyter.
>>
>>
>> On Friday, December 18, 2015 at 11:35:53 AM UTC+1, Jeroen Demeyer wrote:
>>>
>>> * Certain features of SageNB are missing: live documentation, 
>>> sharing/publishing of worksheets. 
>>>
>>
>> IMHO sharing is already better, you can just upload the ipynb file 
>> anywhere (including http://gist.github.com or any other pastebin) and 
>> then share it using the http://nbviewer.ipython.org service. For example:
>>
>> http://nbviewer.ipython.org/url/sagepad.org/pub/sage-test.ipynb
>>
>
> Some care will need to be taken, so that we don't have to deal with 
> complaints of kind
>
>   " I run 'sage -n', and all my worksheets are gone!!! Help!!! "
>
>
>
>>
>>
>> * It clutters the file system with .ipynb files. 
>>>
>>
>> and .ipynb_checkpoints hidden directories. See also: 
>> http://trac.sagemath.org/19746
>>
>>
>>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Volker Braun]

IMHO the Jupyter/IPython notebook is a clear winner, no real contest.

Nobody prevents you form using SageNB for existing worksheets or the live 
documentation; But if you start a new worksheet today then you really 
should be using Jupyter.


On Friday, December 18, 2015 at 11:35:53 AM UTC+1, Jeroen Demeyer wrote:
>
> * Certain features of SageNB are missing: live documentation, 
> sharing/publishing of worksheets. 
>

IMHO sharing is already better, you can just upload the ipynb file anywhere 
(including http://gist.github.com or any other pastebin) and then share it 
using the http://nbviewer.ipython.org service. For example:

http://nbviewer.ipython.org/url/sagepad.org/pub/sage-test.ipynb


* It clutters the file system with .ipynb files. 
>

and .ipynb_checkpoints hidden directories. See 
also: http://trac.sagemath.org/19746


-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[sage-devel] Re: Jupyter notebook by default?

[Emmanuel Charpentier]

A few addenda :

Pros :
 * The python debugger works in the IPython notebook. Damn useful !
 * Some extensions developed for Python worksheets run in Sage worksheets.

Cons :
 * (Currently) no standard way to create a cell in another language, as 
allowed by the Sage notebook's magics such as %maxima : in the current 
implementation, they create separate instances of interpreters, which do 
not return to the worksheet, even when explicitely terminated. Exceptions :
  - Python (creates a separate subprocess)
  - R, if rpy2 is young enough (but the interpreter used in %%R cells is 
different of the one used by r functions. This MUST be fixed).
 * The lack of interacts is a sore point. But I hear there is work on this 
front...

I don't see what you mean when you say that live documentation is missing. 
You can access the online help by typing a function name and a question 
mark, and a pager will open on the right page. And the "help" menu contains 
a load of links to the manuals and tutorials (this should be extended, by 
the way...)

And I find that the "cluttering" of the filesystem (i.e. one worksheet, one 
file) is much saner that the Sage notebook "solution" (one worksheet, a 
$#!+load of anonymous files hidden in an invisble directory).

Hoping that my two cents may help,

--
Emmanuel Charpentier


Le vendredi 18 décembre 2015 11:35:53 UTC+1, Jeroen Demeyer a écrit :
>
> Should the Jupyter notebook be the default notebook for the next Sage 
> 7.0 release? I don't really have an opinion on the matter. 
>
>
> Pros: 
>
> * Nice tracebacks! 
> * The Jupyter notebook is a mature well-maintained project, unlike 
> SageNB. It is widely used in the "scientific Python" community. 
> * Availability of other Jupyter kernels besides Sage. 
>
>
> Cons: 
>
> * Less compatible with Sage: Sage interacts don't work, some graphics 
> don't work. 
> * Certain features of SageNB are missing: live documentation, 
> sharing/publishing of worksheets. 
> * It clutters the file system with .ipynb files. 
>
>
> Don't cares: 
>
> * It's just a default choice, both options remain available. 
>

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

Re: [sage-devel] Re: Jupyter notebook by default?

[William Stein]

On Fri, Dec 18, 2015 at 6:47 AM, Emmanuel Charpentier
 wrote:
> A few addenda :
>
> Pros :
>  * The python debugger works in the IPython notebook. Damn useful !
>  * Some extensions developed for Python worksheets run in Sage worksheets.
>
> Cons :
>  * (Currently) no standard way to create a cell in another language, as
> allowed by the Sage notebook's magics such as %maxima : in the current
> implementation, they create separate instances of interpreters, which do not
> return to the worksheet, even when explicitely terminated. Exceptions :

Wow, that's annoying; I didn't realize they had that limit.

>   - Python (creates a separate subprocess)
>   - R, if rpy2 is young enough (but the interpreter used in %%R cells is
> different of the one used by r functions. This MUST be fixed).
>  * The lack of interacts is a sore point. But I hear there is work on this
> front...
>
> I don't see what you mean when you say that live documentation is missing.
> You can access the online help by typing a function name and a question
> mark, and a pager will open on the right page. And the "help" menu contains
> a load of links to the manuals and tutorials (this should be extended, by
> the way...)
>

The Sage notebook lets you browse the entire 10K-page reference
manual, tutorial, etc., of Sage as live, auto-converted worksheets.
This is very useful for people learning Sage.  You don't have to copy
paste from the docs -- you directly play with them.   This was a very
cool feature Dorian Raymer added to the Sage notebook in maybe 2007 or
2008.   There was some work on trying to create something similar at
the Berkeley sage days.

> And I find that the "cluttering" of the filesystem (i.e. one worksheet, one
> file) is much saner that the Sage notebook "solution" (one worksheet, a
> $#!+load of anonymous files hidden in an invisble directory).

Agreed... (in defense, the current design was something we came up
with in 2006, when we were planning to have *everything* stored in a
database, which was the sort of standard way to make websites back
then...)

>
> Hoping that my two cents may help,
>
> --
> Emmanuel Charpentier
>
>
>
> Le vendredi 18 décembre 2015 11:35:53 UTC+1, Jeroen Demeyer a écrit :
>>
>> Should the Jupyter notebook be the default notebook for the next Sage
>> 7.0 release? I don't really have an opinion on the matter.
>>
>>
>> Pros:
>>
>> * Nice tracebacks!
>> * The Jupyter notebook is a mature well-maintained project, unlike
>> SageNB. It is widely used in the "scientific Python" community.
>> * Availability of other Jupyter kernels besides Sage.
>>
>>
>> Cons:
>>
>> * Less compatible with Sage: Sage interacts don't work, some graphics
>> don't work.
>> * Certain features of SageNB are missing: live documentation,
>> sharing/publishing of worksheets.
>> * It clutters the file system with .ipynb files.
>>
>>
>> Don't cares:
>>
>> * It's just a default choice, both options remain available.
>
> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+unsubscr...@googlegroups.com.
> To post to this group, send email to sage-devel@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.



-- 
William (http://wstein.org)

-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to sage-devel+unsubscr...@googlegroups.com.
To post to this group, send email to sage-devel@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.