Two bits here:

   1. payloads are more than the pager. I think we are only talking about
   the pager payload, not other payloads like set_next_input that allow
   creation of cells, etc.
   2. have implemented IPython’s ? and ?? twice in the message
   specification:
      - invoking %pinfo formats output and triggers display via the pager
      payload
      - inspect_request requests the same information at the *protocol
      level*, triggering the same code at the Python leel, but skipping the
      pager. This is how tooltips are produced (shift-enter some places,
      automatically others).

Summarizing some discussion from the dev meeting: I support deprecating the
pager in JupyterLab, and moving IPython’s default behavior of ? to produce
regular output (I’ve used extensions
<https://github.com/minrk/ipython_extensions/blob/master/extensions/print_page.py>
to do this since about the beginning of the notebook, and it sounds like
others have, too). The fact that JupyterLab can tear off output puts a big
dent in the main reason to keep the pager in the notebook: that it can be
independent of the document scrolling, and kept for reference.

Argument for keeping the pager: not all frontends have JupyterLab’s setup.
The console, for instance, would still benefit from sending this to the
pager for the same reason terminal IPython does.

-Min
​

On Mon, Sep 26, 2016 at 8:07 PM, Jason Grout <[email protected]> wrote:

> That's how the Sage notebook has worked for as long as I remember: A cell
> executed with ? and ?? render the relevant docs inline as output.
> SageMathCloud does this too: https://cloud.sagemath.
> com/projects/82a3ca9f-7a34-400b-86ee-98fb0b07f23c/files/
> 2016-05-05-133427.sagews. It is handy having it inline - the Sage
> tutorial uses this to explain ? and ??, for example, and many other
> tutorials will discuss a function and then do the function? to show the
> docs inline. HOWEVER, doing something like `math.cos?` and pressing *tab*
> (instead of shift-enter) will either pop up the completion (SMC) or render
> it temporarily in the output above existing output, with an option to pop
> the output out into a dialog (old sage notebook)
>
> This also would make the notebook more consistent with the terminal
> console, where docs are rendered as outputs.
>
> Thanks,
>
> Jason
>
>
> On Mon, Sep 26, 2016 at 10:06 AM Kyle Kelley <[email protected]> wrote:
>
>> Part of the reason we put the pager data within the output area was that
>> we didn't want that "popup" experience and people seemed to like having it
>> inline. We then faced, exactly as you put it, a lack of reproducibility of
>> the notebook - people wanted to see that "output" of documentation right
>> there.
>>
>> On Mon, Sep 26, 2016 at 8:01 AM, Brian Granger <[email protected]>
>> wrote:
>>
>>> Hi all,
>>>
>>> For a while, the ipython kernel has had a flag that lets you send
>>> payload message to regular output. I have run my ipython with this
>>> enabled for a long time. In talking with Kyle this weekend, nteract
>>> has decided to always render payloads as output.
>>>
>>> I would like to propose that we start to do the same in the classic
>>> notebook and jupyterlab.
>>>
>>> The reasons are these:
>>>
>>> * If you have a cell with a ? ("math.cos?") no-one can see the result
>>> of running that code in the notebook document (on github, nbviewer).
>>> This means that the notebook isn't really reproducible.
>>> * In JupyterLab, outputs can be dragged anywhere in the dock panel.
>>> Thus, if you want the separate pager like experience we can still
>>> offer it.
>>> * The existing inspector in juptyerlab is overly complicated. This
>>> would allow it to just focus on tooltips as the user types.
>>>
>>> Cheers,
>>>
>>> Brian
>>>
>>> --
>>> Brian E. Granger
>>> Associate Professor of Physics and Data Science
>>> Cal Poly State University, San Luis Obispo
>>> @ellisonbg on Twitter and GitHub
>>> [email protected] and [email protected]
>>>
>>> --
>>> You received this message because you are subscribed to the Google
>>> Groups "Project Jupyter" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to [email protected].
>>> To post to this group, send email to [email protected].
>>> To view this discussion on the web visit https://groups.google.com/d/
>>> msgid/jupyter/CAH4pYpTDrOyEP_rbj2zeydboVHy_2AapPD0xiyi_
>>> 4sqD1hDqxQ%40mail.gmail.com.
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>
>>
>>
>> --
>> Kyle Kelley (@rgbkrk <https://twitter.com/rgbkrk>; lambdaops.com)
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "Project Jupyter" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected].
>> To post to this group, send email to [email protected].
>> To view this discussion on the web visit https://groups.google.com/d/
>> msgid/jupyter/CA%2BtbMaW8KsXnCRRseKSCuRcVkX5aeT
>> Nm-4-9%2BbDjpSP3cBnOUw%40mail.gmail.com
>> <https://groups.google.com/d/msgid/jupyter/CA%2BtbMaW8KsXnCRRseKSCuRcVkX5aeTNm-4-9%2BbDjpSP3cBnOUw%40mail.gmail.com?utm_medium=email&utm_source=footer>
>> .
>> For more options, visit https://groups.google.com/d/optout.
>>
> --
> You received this message because you are subscribed to the Google Groups
> "Project Jupyter" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To post to this group, send email to [email protected].
> To view this discussion on the web visit https://groups.google.com/d/
> msgid/jupyter/CAPDWZHzRZb%3Dc2u5OpEzkq_ALw95vaU0SiZ4kEN_
> 8H1oC-iSCGg%40mail.gmail.com
> <https://groups.google.com/d/msgid/jupyter/CAPDWZHzRZb%3Dc2u5OpEzkq_ALw95vaU0SiZ4kEN_8H1oC-iSCGg%40mail.gmail.com?utm_medium=email&utm_source=footer>
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"Project Jupyter" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jupyter/CAHNn8BWX%2Bjr-dA-TkT7ii3F2yW_3fzMQeewh4vRDjhXnQbxR8w%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to