SageCell Doctor Page

40 views
Skip to first unread message

Ingo Dahn

unread,
Feb 15, 2020, 4:16:39 AM2/15/20
to sage-cell
Would it be possible to establish and maintain on https://sagecell.sagemath.org/ a SageCell Doctor Page that helps curing broken SageCell worksheets? It might be a simple web page with a table giving
  • Affected part (broken construct
  • Symptom (Error message)
  • Prescription (Workaround, Replacement)

Andrey Novoseltsev

unread,
Feb 16, 2020, 12:46:30 PM2/16/20
to sage-cell
 Well, it is certainly possible, but who is going to establish and maintain it? It has to be somebody actively working on this problem. Do you volunteer?-)

Implementation-wise the best approach is probably to create a wiki page like https://wiki.sagemath.org/Python3-user or perhaps just use that one as nothing is really SageMathCell-specific. I can link to it above the "About" header on https://sagecell.sagemath.org/help.html which is visible when someone click "Help" under results frame for embedded cells and on the root page https://sagecell.sagemath.org/ it is visible right away. Do you think this will be helpful?

Ingo Dahn

unread,
Feb 17, 2020, 12:29:29 AM2/17/20
to sage-cell
Yes and No:

Skip the next two paragraphs if you're short of time, they just give reasons :-).

Try to get into the mind of peple like the colleagues in my network. You are busy with your daily duties as a working and teaching scientist or engineer.You are not overly interested in programming. A year ago or so you discovered that SageCell enables you to do useful things that you couldn't do before and you invested the time to find and learn the 10 SageMath commands that you need to solve a few specific problems. and built your worksheets from these (well, maybe not even that, perhaps you just got your worksheets from a friendly colleague.) and you use them regularly.You're happy with these and you don't have a reason to invest any further time into tooling up - your sheets work reliably over a year.
One morning you come to your desk or to your class and - bang - find your code broken; frustration. You are under time pressure, your Sage knowledge is limited and rusty. Suppose you find the python-3-user page, you'll not find it too helpful (you had never been working with Python and you have just a vague idea what an iterator might be, you are reluctant to set up another account to report your problem - you just want the damn' code to work again as it used to and as effortlessly as possible)..

Summary: SageCell is so mature that it is used by people who have no deep interest in it.

Short answers:

No: The Python-3-user-page is  not useful for all SageCell users.
Yes: This page is useful as an augmenting information why certain solutions work.
Yes: I volunteer to set up something simple for collecting problems, solution recipes and links to augmenting information but I cannot fill those in.

Andrey Novoseltsev

unread,
Feb 17, 2020, 10:56:57 AM2/17/20
to sage-cell
I've been there, to the point that something works on my computer in the morning and a few hours later things are strangely broken during the class. The only solution to this is running your own SageCellServer (or using CoCalc where for your particular project you can keep using the same version of Sage all day long). Having a collection of quick recipes on how to fix certain things would not be useful, I think, as in class there is no time for that anyway. For fixing things back in your office during "free time" there is some potential, of course. It still seems to me that something where people can easily add problems and solutions like wiki is the best approach. Python3-user page may not address your needs but we can link it together with something else.

So: can you please let me know the links and words you'd like to see on the top of https://sagecell.sagemath.org/help.html that would be most helpful for your situation? This is relatively static, as in no changes will be made without my manual participation, but the linked pages of course can and should be editable by many others.

Ingo Dahn

unread,
Feb 17, 2020, 11:57:41 AM2/17/20
to Andrey Novoseltsev, sage-cell
Hi,
a collection of recipes WOULD BE useful in our case here, because I already issued a warning to my NetMath network to re-check SageCell worksheets before using them in class. Moreover we are approaching the semester break which leaves time for corrections - provided one knows how to correct.

Something where people can easily add and find problems and solutions is exactly what I'd like to have. If you prefer to setup and maintain a Wiki for that, it's perfectly fine. If you consider using my  ad-hoc solution you may link to https://dahn-research.eu/sagecelldoctor/ (I don't have a Wiki installed on my server and that link is maintenance-free for me :-)). That allows for adding problems. If you think editing the solutions page should also be allowed for anyone I can enable that. If you chose to go for some other solution, I'll de-activate that link.

For finding the problems/solutions page I'd suggest having an h1 headline like "Help with Broken Code" or "Help with Broken SageCell Scripts" on the SageCell Help page, linked to wherever one can put in problems and solutions. Some additional text explaining the current SageCell transition and giving also a link to the Python-3-help page would certainly be welcome too.

Am Mo., 17. Feb. 2020 um 16:56 Uhr schrieb Andrey Novoseltsev <novo...@gmail.com>:
I've been there, to the point that something works on my computer in the morning and a few hours later things are strangely broken during the class. The only solution to this is running your own SageCellServer (or using CoCalc where for your particular project you can keep using the same version of Sage all day long). Having a collection of quick recipes on how to fix certain things would not be useful, I think, as in class there is no time for that anyway. For fixing things back in your office during "free time" there is some potential, of course. It still seems to me that something where people can easily add problems and solutions like wiki is the best approach. Python3-user page may not address your needs but we can link it together with something else.

So: can you please let me know the links and words you'd like to see on the top of https://sagecell.sagemath.org/help.html that would be most helpful for your situation? This is relatively static, as in no changes will be made without my manual participation, but the linked pages of course can and should be editable by many others.

--
You received this message because you are subscribed to a topic in the Google Groups "sage-cell" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/sage-cell/M5AuiBslUeo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sage-cell+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-cell/3776c2b9-a541-447f-a519-a7ff1dbd29e7%40googlegroups.com.

Andrey Novoseltsev

unread,
Feb 17, 2020, 11:08:04 PM2/17/20
to sage-cell
OK, I've added the links, let me know if something has to be changed.

By the way - for how long do you think we need to keep this message there?

Ingo Dahn

unread,
Feb 18, 2020, 12:04:27 PM2/18/20
to Andrey Novoseltsev, sage-cell
Thanks, Andrey. I'd suggest re-phrasing the sentence on  https://sagecell.sagemath.org/help.html to make the distinction between both links apparent:
For more details please consult with Python3-Switch wiki page. For suggestions for corrections and for submitting broken code go to SageCellDoctor spreadsheet.
How long this page will be useful depends on the transition of the community and on the efforts to make SageCell backwards compatible. Perhaps until the end of 2020?
I'll launch a new thread "Repairing Broken Code" to make it as easy as possible to submit solutions.

Am Di., 18. Feb. 2020 um 05:08 Uhr schrieb Andrey Novoseltsev <novo...@gmail.com>:
OK, I've added the links, let me know if something has to be changed.

By the way - for how long do you think we need to keep this message there?

--
You received this message because you are subscribed to a topic in the Google Groups "sage-cell" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/sage-cell/M5AuiBslUeo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sage-cell+...@googlegroups.com.

Andrey Novoseltsev

unread,
Feb 18, 2020, 12:59:40 PM2/18/20
to sage-cell
On Tuesday, 18 February 2020 10:04:27 UTC-7, Ingo Dahn wrote:
Thanks, Andrey. I'd suggest re-phrasing the sentence on  https://sagecell.sagemath.org/help.html to make the distinction between both links apparent:
For more details please consult with Python3-Switch wiki page. For suggestions for corrections and for submitting broken code go to SageCellDoctor spreadsheet.
How long this page will be useful depends on the transition of the community and on the efforts to make SageCell backwards compatible. Perhaps until the end of 2020?
I'll launch a new thread "Repairing Broken Code" to make it as easy as possible to submit solutions.

I have changed the page per your suggestion.

As for efforts to make SageMathCell backwards compatible - what you mean is making SageMath backwards compatible, since SageMathCell "just runs it". The only way to make it "more compatible" is to allow people who embed cells to specify an explicit version and maintain servers that run these version "forever". Apart from implementing this new feature it will require more hosting power to run multiple version and extra security considerations with running old versions which are no longer supported. One may also question how good it will be to learn some code, see that it works, and then use it as a base of your own code to discover that it actually does not work anymore. So from a practical point of view, taking into account history, I don't think there will be any change in this direction. For SageMath itself there are efforts to keep it backward compatible over a year and break something only with deprecation warnings, but of course there are cases when this intention was not implemented.
Reply all
Reply to author
Forward
0 new messages