cocalc and sagemath webpage

[Frédéric Chapoton]

Dear all,

Beware that this message may be controversial.

If you look at https://www.sagemath.org/

you can see four links to cocalc, one of them hidden behind the icon "sage".

So, taking great care to acknowledge first all that William did and the fact that Harald is currently maintaining the website, should we just accept this without any kind of discussion ?

Maybe cocalc could become one of our sponsors, as a display of explicit and permanent friendship ?

Frédéric

[Kwankyu Lee]

Acknowledging first all that William did and the fact that Harald is currently maintaining the website,

I think that the Sage website should remove the links to Cocalc from the front page. Perhaps they can be collected in a displayed box dedicated for sponsors.   

[Emmanuel Charpentier]

I’m not sanguine about this question. My beef with these links and icons has more to do with the fact that Cocalc isn’t “defined. Someone discovering this page can’t deduce that Cocalc is a commercial service built around the free (as in GPL-3) software Sagemath. Hence some confusion.

Furthermore, Cocalc is not a a fork of Sagemath ; as far as I can tell, Cocalc-created patches to Sagemath are brought back to Sagemath. ISTR that, at some point in time, William proposed the sources to his server as an alternative front end.

A couple suggestions :

• A one-sentence presentation of Cocalc. Rough proposal : “Cocalc [link] is a company (founded by one of the Sagemath creators) offering commercial online computing and collaborative services centered on (but not limited to) Sagemath.”

• Possibly a one-sentence presentation of Sagemath Cell. Rough proposal : “Sagemath Cell [link] is a free Sagemath server offering a Sagemath computation facility embeddable in (X)HTML pages ; it can also be used as an online calculator.”

• Replace the icon linking to Cocalc by the Cocalc icon [ necessary if any link to Cocalc subsists on the page ].

• Offer a link to Sagemath cell (possibly with a specific icon, to be created).

BTW, my real beef with Cocalc is … Cocalc itself :

• William has justified his decision to spin a commercial company off Sagemath by the difficulty (to say the least) of financing Sagemath development in areas not already a research subject for Sagemath developers, resulting in a necessity of financing the development of Sagemath by commercial means.

• I can agree with the diagnosis : financing “non-sexy” development through public projects has become increasingly difficult ; this commercial drift of public research policy and management can be observed in all disciplines and all countries.

• I may or may not agree with the proposed therapeutics : IMNSHO, the commercial funding of research and development is but a band aid ; an etiologic treatment of the problem is a much larger problem, probably out of our scope (efficiently debunking the “magic of markets” fallacy is hard…).

• I am not aware of an assessment of the importance of Cocalc-created enhancements in Sage evolution, so don’t have an opinion about the success/failure of this treatment.

Hence my “moderate” attitude : be absolutely clear on the nature of Cocalc, but don’t hinder it “gratuitously”, unless and until proof of its “toxicity” to Sagemath can be made.

HTH,

​

[Dima Pasechnik]

On Sun, Nov 15, 2020 at 1:41 PM Frédéric Chapoton <fchap...@gmail.com> wrote:
> If you look at https://www.sagemath.org/
> you can see four links to cocalc, one of them hidden behind the icon "sage".

This icon should be replaced by CoCalc's icon:
https://cocalc.com/img/cocalc-icon.svg
And a SageMathCell getting a similarly sized icon.
Or perhaps making this icon a link to SageCell, and placing CoCalc
icon next to it.

>
> So, taking great care to acknowledge first all that William did and the fact that Harald is currently maintaining the website, should we just accept this without any kind of discussion ?
>
> Maybe cocalc could become one of our sponsors, as a display of explicit and permanent friendship ?
>
> Frédéric
>

> --
> 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/91cd2053-daeb-40b5-986e-20ae547ff181n%40googlegroups.com.

[Karl-Dieter Crisman]

I can't figure out how to reply in Google groups web interface all of a sudden ... hope this finds the right thread.

There are indeed various places online where the distinction has been difficult - for years I have had to try to correct people saying "Sage is now CoCalc" when it's more complex, for one thing the codebases, while both open, are completely different - but at this point we only have control over the ones on sagemath.org.  It seems fine to me to point people to places where they can use SageMath online, which could include services beyond CoCalc, as long as the distinction is clear but not blaring.  So Dima's solution seems pretty reasonable.

[Dima Pasechnik]

On Monday, November 16, 2020 at 4:39:02 PM UTC kcrisman wrote:

I can't figure out how to reply in Google groups web interface all of a sudden ... hope this finds the right thread.

in the top right corner of the message window there are two curved arrows - this is the "Reply" button.

HTH

Dima

[kcrisman]

On Wednesday, November 18, 2020 at 5:24:12 AM UTC-5 dim...@gmail.com wrote:

On Monday, November 16, 2020 at 4:39:02 PM UTC kcrisman wrote:

I can't figure out how to reply in Google groups web interface all of a sudden ... hope this finds the right thread.

in the top right corner of the message window there are two curved arrows - this is the "Reply" button.

Yes, ordinarily they are there, like they are right now :-) But for some reason they weren't there on Monday.  And it's not liked I logged in or out of anything ... sigh.

[Dima Pasechnik]

use email as interface, I guess many people gave up on Groups interface long ago. 

--
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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/8354514d-ca36-42dd-8716-37efcc17f4f7n%40googlegroups.com.

[Travis Scrimshaw]

First, let me acknowledge my respect and admiration to William for his contributions both to Sage and for having such a great service such as CoCalc. However, I must also disagree with having this being linked with the Sage logo as these are separate entities both in their mission and content. I agree with Dima that it should be with the CoCalc icon and SageMathCell should also get a similar sized icon. We will then have a prime number of icons, and a little more care will be needed for their layout. However, that is a minor issue IMO.

Best,

Travis

[Dima Pasechnik]

On Thu, Nov 19, 2020 at 11:34 PM 'Travis Scrimshaw' via sage-devel
<sage-...@googlegroups.com> wrote:
>
> First, let me acknowledge my respect and admiration to William for his contributions both to Sage and for having such a great service such as CoCalc. However, I must also disagree with having this being linked with the Sage logo as these are separate entities both in their mission and content. I agree with Dima that it should be with the CoCalc icon and SageMathCell should also get a similar sized icon. We will then have a prime number of icons, and a little more care will be needed for their layout. However, that is a minor issue IMO.

there are potentially more ways to run sagemath online, e.g. it used
to be possible at some point to run it in mybinder, IIRC.
Now it's broken, however: https://github.com/sagemath/sage-binder-env/issues/9

Anyone knows how to fix the latter, by the way?

Dima

>
> Best,
> Travis
>
>
> On Monday, November 16, 2020 at 8:22:00 PM UTC+10 dim...@gmail.com wrote:
>>
>> On Sun, Nov 15, 2020 at 1:41 PM Frédéric Chapoton <fchap...@gmail.com> wrote:
>> > If you look at https://www.sagemath.org/
>> > you can see four links to cocalc, one of them hidden behind the icon "sage".
>> This icon should be replaced by CoCalc's icon:
>> https://cocalc.com/img/cocalc-icon.svg
>> And a SageMathCell getting a similarly sized icon.
>> Or perhaps making this icon a link to SageCell, and placing CoCalc
>> icon next to it.
>>
>>
>> >
>> > So, taking great care to acknowledge first all that William did and the fact that Harald is currently maintaining the website, should we just accept this without any kind of discussion ?
>> >
>> > Maybe cocalc could become one of our sponsors, as a display of explicit and permanent friendship ?
>> >
>> > Frédéric
>> >
>> > --
>> > 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/91cd2053-daeb-40b5-986e-20ae547ff181n%40googlegroups.com.
>
> --
> 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/05ed12c5-bfd0-4af5-9db2-6f5ae8a7118cn%40googlegroups.com.

[Eric Gourgoulhon]

Le vendredi 20 novembre 2020 à 10:40:07 UTC+1, dim...@gmail.com a écrit :

there are potentially more ways to run sagemath online, e.g. it used
to be possible at some point to run it in mybinder, IIRC.
Now it's broken, however: https://github.com/sagemath/sage-binder-env/issues/9

Anyone knows how to fix the latter, by the way?

Well, this is more an issue with that particular repository (https://github.com/sagemath/more-sagemath-tutorials) than with Binder and Sage. For instance,

https://mybinder.org/v2/gh/egourgoulhon/SageMathTour/master?filepath=Notebooks/demo_pseudo_Riemannian_Schwarzschild.ipynb

works well in Binder running Sage 9.1.

One may notice that Dockerfile is outdated in the repository more-sagemath-tutorials:

the line

FROM sagemath/sagemath:9.0-py3

should be replaced by

FROM sagemath/sagemath:9.1

(until the Docker image of 9.2 is ready).

Eric.

[Eric Gourgoulhon]

PS: btw, I am also +1 with Dima proposal regarding the web page.

[Eric Gourgoulhon]

Le vendredi 20 novembre 2020 à 11:58:52 UTC+1, Eric Gourgoulhon a écrit :

PS: btw, I am also +1 with Dima proposal regarding the web page.

PPS: I forgot to mention (this is so obvious to me), that CoCalc is a great tool; I am using it for my research, to share notebooks with collaborators, adding public links to them in the published papers.

[Doris Behrendt]

Hi all,

since I’m not one of the >>old boys<< in sagemath, I read some background when this mailthread began.
Then I went to CoCalc and searched for something like >>imprint<< oder >>legal notice<<, since I live in the EU and wanted to know If I could recommend using this with students. I found NOTHING by direct search. Or more exactly, different contradicting information, like being based in Delaware and at the same time in Seattle.

So, also with no offence of course, I must say that I find this at least problematic and intransparent and that it leaves me with open questions.

Then, I also had a look at the sagemath page, and no imprint there also …
at least some info about how the project is financed is there, but some GDPR stuff would perhaps be a good idea. And in the context of transparency it would also be good to know who is giving money to the project. Think of Chinese (nothing against Chinese Companys! I don’t care who of the 2, the Chinese or the NSA, or somebody else, try spying after me) companys financing chairs at US universities etc., but, to be honest, I didn’t look if this information was available, I had no time. Does anyone of you know?

And I agree that the icons should be unique and clear.

Doris

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/05ed12c5-bfd0-4af5-9db2-6f5ae8a7118cn%40googlegroups.com.

[Dima Pasechnik]

Hi Doris,

CoCalc is not something we have any control over here. I cc this to CoCalc webmaster, in case.

sagemath.org itself does not store any personal data (as far as I know - it ought to be double-checked though) and thus does not have to worry about GDPR (I am not a lawyer, though).

HTH

Dima

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/69B4554E-4136-40AE-8176-118D9049FB67%40me.com.

[Dima Pasechnik]

PS. as far as current funding for the sagemath (itself  - sagemath's dependencies are another story, some enjoy support by NumFocus, DFG, CZI, etc etc, some are doing well in this sense, e.g. Jupyter, R, some have nothing, some are orphaned/taken over by sagemath) project goes, very minimal funding (in form of web hosting support) is provided by CoCalc -which is a US LLC, and small donations (tens of euro p.m.) are made by a dozen private persons via github sponsorship and opencollective sponsorship. We are thankful to them, and they don't tell us what to do in any way.

HTH

Dima

[Dima Pasechnik]

On Fri, Nov 20, 2020 at 10:32 AM Eric Gourgoulhon
<egourg...@gmail.com> wrote:
>
> Le vendredi 20 novembre 2020 à 10:40:07 UTC+1, dim...@gmail.com a écrit :
>>
>>
>> there are potentially more ways to run sagemath online, e.g. it used
>> to be possible at some point to run it in mybinder, IIRC.
>> Now it's broken, however: https://github.com/sagemath/sage-binder-env/issues/9
>>
>> Anyone knows how to fix the latter, by the way?
>>
>
> Well, this is more an issue with that particular repository (https://github.com/sagemath/more-sagemath-tutorials) than with Binder and Sage. For instance,
> https://mybinder.org/v2/gh/egourgoulhon/SageMathTour/master?filepath=Notebooks/demo_pseudo_Riemannian_Schwarzschild.ipynb
> works well in Binder running Sage 9.1.

Is there a place where such a setup is described?

>
> One may notice that Dockerfile is outdated in the repository more-sagemath-tutorials:

could you please do a PR there? (cc to me)

> the line
> FROM sagemath/sagemath:9.0-py3
> should be replaced by
> FROM sagemath/sagemath:9.1
> (until the Docker image of 9.2 is ready).
>
> Eric.
>

> --
> 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/9ea313c0-ee57-4ee0-81f6-445322697528n%40googlegroups.com.

[Eric Gourgoulhon]

Le samedi 21 novembre 2020 à 10:58:17 UTC+1, dim...@gmail.com a écrit :

On Fri, Nov 20, 2020 at 10:32 AM Eric Gourgoulhon
>

> Well, this is more an issue with that particular repository (https://github.com/sagemath/more-sagemath-tutorials) than with Binder and Sage. For instance,
> https://mybinder.org/v2/gh/egourgoulhon/SageMathTour/master?filepath=Notebooks/demo_pseudo_Riemannian_Schwarzschild.ipynb
> works well in Binder running Sage 9.1.

Is there a place where such a setup is described?

No particular setup here. Basically, it suffices to endow the github repository containing the notebooks with a Dockerfile that has these two lines:

FROM sagemath/sagemath:9.1
COPY . ${HOME}

plus possibly other command if you need some specific package, see

https://github.com/egourgoulhon/SageMathTour/blob/master/Dockerfile

Then, when you open a notebook with nbviewer.jupyter.org, e.g.

https://nbviewer.jupyter.org/github/egourgoulhon/SageMathTour/blob/master/Notebooks/demo_pseudo_Riemannian_Schwarzschild.ipynb

clicking on "Execute on Binder" button (the three circles) in the upper right menu will do the job.

All the notebooks of manifold examples (https://sagemanifolds.obspm.fr/examples.html) can be run in Binder with Sage 9.1 in that manner.

I don't know a place in Sage documentation where this is described, but maybe there is one...

 

>
> One may notice that Dockerfile is outdated in the repository more-sagemath-tutorials:

could you please do a PR there? (cc to me)

OK I will.

Best regards,

Eric.