Opened 5 years ago

Closed 4 years ago

#19740 closed enhancement (fixed)

Notebook: run ExportSageNB by default

Reported by: vbraun Owned by:
Priority: major Milestone: sage-8.0
Component: notebook Keywords:
Cc: novoselt, iandrus Merged in:
Authors: Volker Braun, Jeroen Demeyer Reviewers: Volker Braun
Report Upstream: N/A Work issues:
Branch: 4d838d0 (Commits, GitHub, GitLab) Commit: 4d838d042236b62c9e997ecbc04e4606e0500fd1
Dependencies: #22458 Stopgaps:

Status badges

Description (last modified by jdemeyer)

Change the default notebook application in Sage to be sagenb_export. This small web application, built on top of Jupyter, acts as migration wizard which can convert legacy Sage notebooks to Jupyter. It also has buttons to run either the legacy Sage notebook server or the new Jupyter notebook server.

Tarball: http://sage.ugent.be/www/jdemeyer/sage/sagenb_export-3.1.tar.gz

Attachments (1)

Screen Shot 2017-03-17 at 2.34.12 PM.png (73.9 KB) - added by kcrisman 4 years ago.
widget javascript not detected?

Download all attachments as: .zip

Change History (77)

comment:1 Changed 5 years ago by vbraun

  • Branch set to u/vbraun/use_jupyter_notebook_by_default

comment:2 Changed 5 years ago by vbraun

  • Authors set to Volker Braun
  • Commit set to 580482ce9b2fc7ea45a9198bbf072d20adf723c3
  • Component changed from PLEASE CHANGE to notebook
  • Status changed from new to needs_review
  • Type changed from PLEASE CHANGE to defect

New commits:

580482cUse Jupyter notebook by default

comment:3 Changed 5 years ago by jdemeyer

Shouldn't the DEPRECATION_NOTICE be shown only when the "default" notebook is run?

And I certainly think this ticket needs a discussion on sage-devel (I have no opinion on the matter).

comment:4 Changed 5 years ago by jdemeyer

  • Status changed from needs_review to needs_info

I would also suggest to upgrade the relevant IPython/Jupyter packages and to get #19571 in too.

comment:5 Changed 5 years ago by git

  • Commit changed from 580482ce9b2fc7ea45a9198bbf072d20adf723c3 to a119f6e4f32347df2ad6f9ed3a2a1d45c1bbb820

Branch pushed to git repo; I updated commit sha1. New commits:

a119f6eShow deprecation notice only when launched for default

comment:6 Changed 5 years ago by vbraun

  • Status changed from needs_info to needs_review

comment:7 Changed 5 years ago by novoselt

  • Cc novoselt added

comment:8 Changed 5 years ago by tmonteil

  • Description modified (diff)

Modfiy description to point to the discussion threads.

comment:9 Changed 5 years ago by jhpalmieri

Small wording suggestion: change "The default notebook implementation changed" to "The default notebook implementation has changed".

Maybe it should say "The default notebook implementation has changed to Jupyter notebooks". I would prefer it if there were a method to convert old-style notebooks to new ones and that the deprecation notice described this method. I'm not sure if this is important enough to hold up a positive review for this ticket.

What happens to the Sage app on OS X? It starts a SageNB notebook server by default. Shouldn't that change, too? And print some sort of deprecation warning? I don't even see a way directly from the app to start a Jupyter server.

comment:10 Changed 5 years ago by kcrisman

  • Cc iandrus added

Maybe it should say "The default notebook implementation has changed to Jupyter notebooks". I would prefer it if there were a method to convert old-style notebooks to new ones and that the deprecation notice described this method. I'm not sure if this is important enough to hold up a positive review for this ticket.

I would more than prefer it.

What happens to the Sage app on OS X? It starts a SageNB notebook server by default. Shouldn't that change, too? And print some sort of deprecation warning? I don't even see a way directly from the app to start a Jupyter server.

Well, I guess you can launch any terminal command from the app, but I don't think it's been updated to have a script dedicated to launching ipynb (though you can get some terminals). Maybe at least adding this should be part of this as well as changing its launch page to note the new situation; I think it's unlikely to be good to change behavior for this audience without a clear "abort" situation and very clear instructions on how to migrate at a later time.

comment:11 Changed 5 years ago by vbraun

The SageNB migration is at #19877

comment:12 Changed 5 years ago by iandrus

I created #19881 to track the Mac app changes.

comment:13 Changed 5 years ago by nthiery

  • Type changed from defect to enhancement

comment:14 Changed 4 years ago by jdemeyer

  • Authors changed from Volker Braun to Volker Braun, Jeroen Demeyer
  • Milestone changed from sage-7.0 to sage-7.6
  • Status changed from needs_review to needs_work
  • Summary changed from Use Jupyter notebook by default to Use ExportSageNB by default

comment:15 Changed 4 years ago by jdemeyer

  • Summary changed from Use ExportSageNB by default to Notebook: run ExportSageNB by default

comment:16 Changed 4 years ago by jdemeyer

  • Milestone changed from sage-7.6 to sage-8.0

comment:17 Changed 4 years ago by jdemeyer

  • Branch changed from u/vbraun/use_jupyter_notebook_by_default to u/jdemeyer/use_jupyter_notebook_by_default

comment:18 Changed 4 years ago by jdemeyer

  • Commit changed from a119f6e4f32347df2ad6f9ed3a2a1d45c1bbb820 to 3be65f901ff6fe33d3d0126842f92f9fd8a8c219

New commits:

20d6394Run sagenb_export as default notebook
3be65f9Update to sagenb_export 3.0

comment:19 Changed 4 years ago by jdemeyer

  • Description modified (diff)

comment:20 Changed 4 years ago by jdemeyer

  • Status changed from needs_work to needs_review

comment:21 Changed 4 years ago by vbraun

  • Reviewers set to Volker Braun
  • Status changed from needs_review to positive_review

comment:22 Changed 4 years ago by jdemeyer

  • Dependencies set to #21267

comment:23 Changed 4 years ago by kcrisman

What's the timeline for Sage 8.0, then?

comment:24 follow-up: Changed 4 years ago by vbraun

The release management script does not consider milestones apart from invalid/wontfix, for the record....

Last edited 4 years ago by vbraun (previous) (diff)

comment:25 Changed 4 years ago by kcrisman

True, but one could presumably manually decide not to merge certain things until a certain release, eh?

comment:26 in reply to: ↑ 24 Changed 4 years ago by jdemeyer

Replying to vbraun:

The release management script does not consider milestones apart from invalid/wontfix, for the record....

Does the script decide which tickets to merge? I thought that was still done by a human, or am I wrong?

In any case, I added #21267 as dependency, which is not merged yet.

comment:27 follow-ups: Changed 4 years ago by nthiery

Hi! It's not quite clear from the summary and description what this ticket is exactly about. Could this be briefly clarified? Thanks!

comment:28 in reply to: ↑ 27 Changed 4 years ago by jdemeyer

Replying to nthiery:

Hi! It's not quite clear from the summary and description what this ticket is exactly about. Could this be briefly clarified? Thanks!

In all honesty, I think the best way to know what it's about is just to checkout this branch and run sage -n. One interactive web-app says so much more than 1000 words :-)

comment:29 in reply to: ↑ 27 Changed 4 years ago by jdemeyer

Replying to nthiery:

Hi! It's not quite clear from the summary and description what this ticket is exactly about.

See https://github.com/OpenDreamKit/OpenDreamKit/raw/master/WP4/D4.5/report-final.pdf Figure 4 :-)

comment:30 Changed 4 years ago by jdemeyer

  • Description modified (diff)

comment:31 Changed 4 years ago by git

  • Commit changed from 3be65f901ff6fe33d3d0126842f92f9fd8a8c219 to f2a35d62de7930e286ab40b634bd5295e305781f
  • Status changed from positive_review to needs_review

Branch pushed to git repo; I updated commit sha1 and set ticket back to needs_review. New commits:

f2a35d6Update to sagenb_export 3.1

comment:32 Changed 4 years ago by jdemeyer

  • Description modified (diff)

comment:33 Changed 4 years ago by jdemeyer

  • Status changed from needs_review to needs_work

comment:34 Changed 4 years ago by jdemeyer

  • Dependencies changed from #21267 to #22458

comment:35 Changed 4 years ago by git

  • Commit changed from f2a35d62de7930e286ab40b634bd5295e305781f to 4d838d042236b62c9e997ecbc04e4606e0500fd1

Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:

1253416Run sagenb_export as default notebook
4d838d0Update to sagenb_export 3.1

comment:36 Changed 4 years ago by jdemeyer

  • Status changed from needs_work to needs_review

Rebased to 7.6.rc1

comment:37 follow-ups: Changed 4 years ago by kcrisman

A couple initial comments:

Please wait while the SageNB export server starts...
[W 14:11:04.595 NotebookApp] server_extensions is deprecated, use nbserver_extensions

Is this a problem?

Also, there doesn't seem to be a "batch" way to convert notebooks - one could imagine someone who used worksheets for daily class and now has to convert dozens at once. I know that the command-line interface to this should, or at least I thought I tried that out once. Could the converter support this? (I realize this is asking a lot.)

comment:38 follow-up: Changed 4 years ago by kcrisman

There is no information in the launch message as to where these notebooks are saved. Turns out it is in the directory you launch from - dangerous for newbies to the Jupyter, especially if they have only launched with the Mac app bundle. Also would cause that directory to get quite cluttered.

Is there a way to get output to typeset in Jupyter? I don't know the answer.

Obviously, I get

NameError: name 'DATA' is not defined

when I rely on that. Any solutions in mind for what to do with this folder? Where do people using Jupyter keep data (in this case, an image file)? I imagine there is a canonical answer to that but I'm not familiar with it.

comment:39 Changed 4 years ago by kcrisman

Can Sage in Jupyter plot in R mode? E.g.

%%r
x <- rnorm(100)
y <- rnorm(100, mean=5)*x+rnorm(100, mean=0.1)
png()
plot(y ~ x)
abline(lm(y ~ x), lwd=2)
.silent.me. <- dev.off()

not sure how to/whether possible to fix this; it was hard enough to get working in the sagenb lo these many years ago.

comment:40 follow-ups: Changed 4 years ago by kcrisman

Interacts library fails.

interacts.calculus.simpson_integration()

(To be fair, I'm not sure whether this works right in the sagenb any more either.)

comment:41 in reply to: ↑ 37 ; follow-up: Changed 4 years ago by tmonteil

Replying to kcrisman: [...]

Also, there doesn't seem to be a "batch" way to convert notebooks - one could imagine someone who used worksheets for daily class and now has to convert dozens at once. I know that the command-line interface to this should, or at least I thought I tried that out once. Could the converter support this? (I realize this is asking a lot.)

For worksheets that are not in the ~/.sage directory but are .sws files, you can loop over them and combine sage -sws2rst and sage -rst2ipynb commands.

comment:42 follow-up: Changed 4 years ago by kcrisman

Layout parameters for interacts do not seem to work properly. E.g.

val1(v1,v23,v24,v34,v234,a)=v1-a*(v23+v24+v34)-v234
val2(v2,v13,v14,v34,v134,a)=v2-a*(v13+v14+v34)-v134
val3(v3,v12,v14,v24,v124,a)=v3-a*(v12+v14+v24)-v124
val4(v4,v12,v13,v23,v123,a)=v4-a*(v12+v13+v23)-v123

@interact(layout=[['v1','v2','v3','v4'],['v12','v13','v14','v23','v24','v34'],['v123','v124','v134','v234']])
def _(v1=[0,1],v2=[0,1],v3=[0,1],v4=[0,1],v12=[0,1],v13=[0,1],v14=[0,1],v23=[0,1],v24=[0,1],v34=[0,1],v123=[0,1],v124=[0,1],v134=[0,1],v234=[0,1]):
    html("Banzhaf values are %s,%s,%s,%s"%(val1(v1,v23,v24,v34,v234,1),val2(v2,v13,v14,v34,v134,1),val3(v3,v12,v14,v24,v124,1),val4(v4,v12,v13,v23,v123,1)))
    html("Shapley-Shubik values are %s,%s,%s,%s"%(val1(v1,v23,v24,v34,v234,1/2),val2(v2,v13,v14,v34,v134,1/2),val3(v3,v12,v14,v24,v124,1/2),val4(v4,v12,v13,v23,v123,1/2)))

comment:43 in reply to: ↑ 41 ; follow-up: Changed 4 years ago by kcrisman

Also, there doesn't seem to be a "batch" way to convert notebooks - one could imagine someone who used worksheets for daily class and now has to convert dozens at once. I know that the command-line interface to this should, or at least I thought I tried that out once. Could the converter support this? (I realize this is asking a lot.)

For worksheets that are not in the ~/.sage directory but are .sws files, you can loop over them and combine sage -sws2rst and sage -rst2ipynb commands.

Sure! But most people's will be in that directory, and people who need this facility will not know shell script enough to loop through them. (Interesting that the sws2rst and rst2ipynb double conversion is robust!)

comment:44 Changed 4 years ago by kcrisman

JSmol menu doesn't seem to work quite right - I can get the menu with right-click, but then the second levels do not appear because of some barrier. I know this is perhaps moot if three.js is now ready, but wanted to point this out.

Changed 4 years ago by kcrisman

widget javascript not detected?

comment:45 Changed 4 years ago by kcrisman

https://trac.sagemath.org/attachment/ticket/19740/Screen%20Shot%202017-03-17%20at%202.34.12%20PM.png Shows something weird as a javascript widget error message - I don't see any problems in the image, though I did end up with two jsmols eventually.

comment:46 follow-up: Changed 4 years ago by kcrisman

(As a side note, the first time you log on the the Jupyter it asks for a token - but it doesn't want the URL for the token, it wants just the token that appears in the terminal message. Possibly really confusing?)

comment:47 follow-ups: Changed 4 years ago by kcrisman

Final comments.

  • Interacts and even 3d stuff in the Jupyter notebook (done this way, I don't know about in general) seems ... very slow. I think that lag explains the double images I occasionally see in interacts. And since the only indication a cell isn't done is the little asterisk on the side, it's very different in style from sagenb. That isn't a criticism but a comment. I still feel ipynb is made more for power users... but anyway maybe the lag is weird. Even 2+2 in a brand new notebook takes forever. Maybe the issue is that the sagenb didn't start until Sage was started, but ipynb can open before the Sage kernel is "ready", Sage is notorious for taking forever to start up.
  • ERROR:root:Cell magic %%fortran not found.
  • I have similar problems with %%cython though it completes running such a cell.

As to the exporter itself, looks very nice. Thoughts:

  • I suspect it isn't possible to click on the "use Sage notebook" button once one has started converting, correct? A message to that effect would be nice.
  • A message for using the converter that says "your stuff will be saved in the following location" would be good - probably could be gotten on the fly from pwd, right?
  • Ideally you'd get to click a button that says "remember my decision" for the default, but that might be too much extra work.
  • More importantly, the button to the Jupyter just gives you a file listing - again, maybe not so helpful. Ideas about what the default should be - maybe opening a new notebook with Sage kernel?

comment:48 in reply to: ↑ 46 Changed 4 years ago by jdemeyer

Replying to kcrisman:

(As a side note, the first time you log on the the Jupyter it asks for a token - but it doesn't want the URL for the token, it wants just the token that appears in the terminal message. Possibly really confusing?)

Yes this happens sometimes. This issue randomly appears for me, which I really do not understand. Yesterday I got this problem and started working on it and then the problem suddenly disappeared and now I cannot reproduce it anymore. Really strange...

comment:49 in reply to: ↑ 38 Changed 4 years ago by jdemeyer

Replying to kcrisman:

There is no information in the launch message as to where these notebooks are saved. Turns out it is in the directory you launch from

True, that's the usual Jupyter behaviour. I wouldn't change this, simply because that is how Jupyter works.

especially if they have only launched with the Mac app bundle

What's wrong with the Mac app bundle (I am not a Mac user)?

comment:50 in reply to: ↑ 40 Changed 4 years ago by jdemeyer

Replying to kcrisman:

Interacts library fails.

I was not even aware of the existence of this "interacts library". But I guess it could be fixed...

comment:51 in reply to: ↑ 47 Changed 4 years ago by jdemeyer

Replying to kcrisman:

  • I have similar problems with %%cython though it completes running such a cell.

What do you mean? Which problems do you have?

comment:52 in reply to: ↑ 47 Changed 4 years ago by jdemeyer

Replying to kcrisman:

  • More importantly, the button to the Jupyter just gives you a file listing - again, maybe not so helpful.

That's the way how Jupyter works and see no reason to change that.

comment:53 in reply to: ↑ 42 Changed 4 years ago by jdemeyer

Replying to kcrisman:

@interact(layout=v1','v2','v3','v4'],['v12','v13','v14','v23','v24','v34'],['v123','v124','v134','v234?)

I was not aware of the interact(layout) parameter (it does not appear on the many examples from the Sage interacts wiki page that I tried). I need to see if this can be ported to Jupyter.

comment:54 in reply to: ↑ 40 ; follow-ups: Changed 4 years ago by jdemeyer

Replying to kcrisman:

Interacts library fails.

interacts.calculus.simpson_integration()

(To be fair, I'm not sure whether this works right in the sagenb any more either.)

See #22636

comment:55 in reply to: ↑ 37 Changed 4 years ago by jdemeyer

Replying to kcrisman:

Also, there doesn't seem to be a "batch" way to convert notebooks - one could imagine someone who used worksheets for daily class and now has to convert dozens at once. I know that the command-line interface to this should, or at least I thought I tried that out once. Could the converter support this?

I am really a beginner when it comes to web development. If somebody else wants to implement the user interface for this, I will gladly implement the server part of that feature.

comment:56 in reply to: ↑ 43 Changed 4 years ago by nthiery

Replying to kcrisman:

Interesting that the sws2rst and rst2ipynb double conversion is robust!

Is it? That would be good news. And a surprise to me indeed, at least for the second part which I still consider as a quick hack.

comment:57 follow-up: Changed 4 years ago by kcrisman

What's wrong with the Mac app bundle (I am not a Mac user)?

Nothing is wrong with it (or maybe it is? I didn't test it yet), but people who had used it would be completely clueless as to this since they wouldn't have been exposed to command line at all, in principle, much less having to save it somewhere.

... %%cython

What do you mean? Which problems do you have?

The cell runs, but then I can't use the function defined in it. %%fortran is not recognized; %%r works but no graphics.

I was not aware of the interact(layout) parameter (it does not appear on the many examples from the Sage interacts wiki page that I tried). I need to see if this can be ported to Jupyter.

That is very understandable, as it is not used there much, more in later things. I do know a fair number of people who have used it, and if it can't be emulated it means multi-input interacts become far less useful (because too hard to see all the controls at once).

I am really a beginner when it comes to web development. If somebody else wants to implement the user interface for this, I will gladly implement the server part of that feature.

Hmm, I guess I am even more of a beginner there :) What I would say is that the UI at a very minimum could have a warning the top saying something like "Want to export several worksheets at once? Click here for precise instructions on how to do this." and then have said precise instructions, including giving them the custom path to use with the scripts in question.

Otherwise one would have to implement checkboxes, some sort of GET (?) operation, and then have an interface for where to save them (presumably not ALL at once in the usual directory, but a sub-directory) and now you're getting complicated. Plus then one wouldn't have the once-off "click on this worksheet to export it" feature, which is very nice, unless you separated that from the checkboxes ...

comment:58 in reply to: ↑ 54 Changed 4 years ago by kcrisman

(To be fair, I'm not sure whether this works right in the sagenb any more either.)

Update - works in sagenb still, though some stuff does not pretty print any more, as that was apparently not updated.

See #22636

Yup. See also #22644.

comment:59 in reply to: ↑ 57 Changed 4 years ago by jdemeyer

Replying to kcrisman:

... %%cython

What do you mean? Which problems do you have?

The cell runs, but then I can't use the function defined in it.

I regularly use %%cython myself and never found any problems with it in Jupyer.

Still, if it does not work for you, please open a ticket with more details (in particular an actual .ipynb file demonstrating the problem).

comment:60 Changed 4 years ago by jdemeyer

I created #22654 for %%fortran.

comment:61 follow-up: Changed 4 years ago by jdemeyer

kcrisman: what would be needed for you to give positive review to this ticket?

comment:62 follow-up: Changed 4 years ago by vbraun

For batch conversion, how about just a link to download all notebooks (converted and zipped up). That would be easy to implement (ideally not on this ticket).

I'll merge this ticket when 7.6 is out...

comment:63 in reply to: ↑ 61 ; follow-ups: Changed 4 years ago by kcrisman

kcrisman: what would be needed for you to give positive review to this ticket?

List (modulo merging this ticket and opening new 8.0 blockers for these):

  • Some kind of batch conversion option (perhaps as Volker suggests), including detailed instructions as to exactly where they end up. Opening a dialog as to where to save them would be really by far the best. Here's why: any user who has not used the command line except to start the notebook (in particular, Mac app users) will be completely lost trying to find them if they are sent to the pwd. In the Mac app that is in the midst of the bundle.
  • Probably more wording on the export main page about exactly what it's doing, where things end up. That wouldn't be too hard to get feedback on once one sends a beta out for testing. Maybe that would include instructions on what to do to start a new notebook in Jupyter, how to get "text cells", typesetting as in the recent sage-devel post, etc.
  • Testing whether this works properly with Mac app bundle at startup etc. I have no idea how this interfaces with it (haven't had time to test that), but if we send 8.0 and basically drop GUI Mac support, that is really bad marketing.
  • I would say that supporting remaining interact stuff (beyond what Jeroen has already fixed) would not be blocker, except that I think that Jupyter would be the only interface not supporting the layout parameter - is that correct?

A lot of this will be revealed in heavy testing. We should definitely recruit non-Jupyter users to test this out. Don't skimp!

comment:64 in reply to: ↑ 62 Changed 4 years ago by nthiery

Replying to vbraun:

For batch conversion, how about just a link to download all notebooks (converted and zipped up). That would be easy to implement

I like that; it falls in a usual workflow for users, with an explicit flexibility to choose where to store the notebooks.

comment:65 in reply to: ↑ 63 ; follow-up: Changed 4 years ago by jdemeyer

Replying to kcrisman:

  • Probably more wording on the export main page about exactly what it's doing, where things end up. That wouldn't be too hard to get feedback on once one sends a beta out for testing. Maybe that would include instructions on what to do to start a new notebook in Jupyter, how to get "text cells", typesetting as in the recent sage-devel post, etc.

I don't think all that should be on the export page. "more wording on the export main page about exactly what it's doing" sounds like a good idea, but explaining everything about Jupyter notebooks not.

comment:66 in reply to: ↑ 63 Changed 4 years ago by jdemeyer

I created #22663 for interact(layout=...).

comment:67 in reply to: ↑ 65 ; follow-ups: Changed 4 years ago by nthiery

Replying to jdemeyer:

I don't think all that should be on the export page. "more wording on the export main page about exactly what it's doing" sounds like a good idea, but explaining everything about Jupyter notebooks not.

Yup; better link to a good existing notebook tutorial rather than duplicating one there. One option would be to link to existing thematic tutorial after having updated it w.r.t. the Jupyter notebook.

With Sébastien, we will be running this tutorial at the upcoming Sage Days in Montréal at the end of April; this would be a good occasion to beta test it (and work on it beforehand).

comment:68 in reply to: ↑ 67 ; follow-up: Changed 4 years ago by jdemeyer

Replying to nthiery:

better link to a good existing notebook tutorial rather than duplicating one there. One option would be to link to existing thematic tutorial after having updated it w.r.t. the Jupyter notebook.

Well, to be honest, none of this can be done easily in Jupyter:

Making this help page into a worksheet

If you are browsing this document as a static web page, you can see all the examples; however you need to copy-paste them one by one to experiment with them. Use the Upload worksheet button of the notebook and copy-paste the URL of this page to obtain an editable copy in your notebook.

If you are browsing this document as part of Sage’s live documentation, you can play with the examples directly here; however your changes will be lost when you close this page. Use Copy worksheet from the File... menu at the top of this page to get an editable copy in your notebook.

comment:69 in reply to: ↑ 68 Changed 4 years ago by nthiery

Replying to jdemeyer:

Well, to be honest, none of this can be done easily in Jupyter

Indeed. That's why we need a round of beta test to see in practice what's the best way to present a tutorial with similar aims but using the Jupyter notebook.

comment:70 in reply to: ↑ 67 ; follow-up: Changed 4 years ago by kcrisman

I don't think all that should be on the export page. "more wording on the export main page about exactly what it's doing" sounds like a good idea, but explaining everything about Jupyter notebooks not.

Yup; better link to a good existing notebook tutorial rather than duplicating one there. One option would be to link to existing thematic tutorial after having updated it w.r.t. the Jupyter notebook.

Sure, that is fine, I didn't mean that you needed a ridiculously long first page, but making info easy to find is important.

Well, to be honest, none of this can be done easily in Jupyter

Indeed. That's why we need a round of beta test to see in practice what's the best way to present a tutorial with similar aims but using the Jupyter notebook.

Yeah, I would hope we could have a pretty lengthy beta period for this 8.0, even if that means a few tickets have to wait to get in as I guess 8.0 is only about this and a few attendant things. There are a few lists I'd like to spam about testing this out once a download of a beta becomes available (preferably also a Mac app version of the download, only for testing purposes).

comment:71 in reply to: ↑ 70 Changed 4 years ago by kcrisman

Yeah, I would hope we could have a pretty lengthy beta period for this 8.0, even if that means a few tickets have to wait to get in as I guess 8.0 is only about this and a few attendant things. There are a few lists I'd like to spam about testing this out once a download of a beta becomes available (preferably also a Mac app version of the download, only for testing purposes).

See #19881 which was merged quite a while ago - I have never tried that out but perhaps it would be easy to integrate with this.

comment:72 in reply to: ↑ 63 ; follow-up: Changed 4 years ago by vbraun

Replying to kcrisman:

Here's why: any user who has not used the command line except to start the notebook (in particular, Mac app users) will be completely lost trying to find them if they are sent to the pwd. In the Mac app that is in the midst of the bundle.

That should be fixable by chdir-ing to $HOME in the mac app...

comment:73 in reply to: ↑ 72 Changed 4 years ago by kcrisman

Here's why: any user who has not used the command line except to start the notebook (in particular, Mac app users) will be completely lost trying to find them if they are sent to the pwd. In the Mac app that is in the midst of the bundle.

That should be fixable by chdir-ing to $HOME in the mac app...

Absolutely, but would have to be implemented - I don't want to ask a user who thinks dollar signs are for money to do it :-)

comment:74 in reply to: ↑ 54 Changed 4 years ago by kcrisman

What happens with this in Jupyter? Is it a supported thing (whether single or double percent)?

%hide

Same question for #auto/%auto. These will be in many converted worksheets.

comment:75 Changed 4 years ago by vbraun

  • Status changed from needs_review to positive_review

comment:76 Changed 4 years ago by vbraun

  • Branch changed from u/jdemeyer/use_jupyter_notebook_by_default to 4d838d042236b62c9e997ecbc04e4606e0500fd1
  • Resolution set to fixed
  • Status changed from positive_review to closed
Note: See TracTickets for help on using tickets.