[sage-devel] Re: Downgrade R to optional? See #31409.

[Emmanuel Charpentier]

I'm torn :


   - On one hand, R provides a lot (understatement...) of non-trivial 
   numerical and data-processing-related algorithms. A large part of this 
   "lot"  comes from user-written packages (17273 as of today), covering an 
   unparalleled range of use cases ; a lot of them have no equivalent in 
   Python's 291406 (as of today) packages...
   - On the other hand, Windows is the dominant platform among our 
   non-scholar users (especially students) ; losing support for even a part of 
   Sage's ability on this platform  should be a big no-no.
   - On the gripping hand (;-)), our current solution for Windows platform  
   rests squarely on the shoulders of E. Madison Bray, which accomplishes 
   alone the astonishing feat of maintaining what amounts to a port (to an 
   hostile platfotm) almost alone. 

Would it be possible to keep the R *interface* standard while relying on 
the target platform(s) to provide the R interpreter itself (in 
Cygwin-over-Window's case, the Cygwin "port"...). However, this would 
create a dependence on Cygwin's version of R, not necessarily synchronized 
with the one supported by Sage on other platforms.
An alternative would be to create an alternative Windows port relying on 
WSL2 (which essentially runs a Linux kernel and a Linux distribution on top 
of Windows, in native mode and with few performance impact), possibly 
presenting less maintenance problems. This would, however, exclude support 
of any Windows version earlier than recent Windows 10. Is that a problem ? 
(This is not a rhetorical question, but a real one : I am aware that 
upgrading Windows is problematic in various cases, for various reasons...). 
Furthermore, to be realistic, we should be able to commit ourselves to 
maintain a binary distribution for at least one WSL2-supported Linux 
platform.

I wait for further discussion before voting one way or another...

HTH,

Le lundi 8 mars 2021 à 05:22:12 UTC+1, John H Palmieri a écrit :

> Dear all,
>
> You should be aware that ticket #31409 (
> https://trac.sagemath.org/ticket/31409) intends to downgrade R to an 
> optional package because of difficulties building it on Cygwin. Just 
> letting you know in case you care about R being part of Sage and/or you 
> have ideas about how to fix the Cygwin build.
>
> (The branch there to downgrade R already has a positive review, by the 
> way. I have no position on this, but I thought that more Sage developers 
> should be aware.)
>
> -- 
> 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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/17c8f3b8-d308-4a76-8498-f0a165947179n%40googlegroups.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[Michael Orlitzky]

On Mon, 2021-03-08 at 21:04 -0800, William Stein wrote:
> Hi,
> 
> One related question: we tend to have a 1-year deprecation policy with
> Sage, and some could argue that removing R will break use of Sage that uses
> the R interface.  Should removing the R package from standard be subject to
> this deprecation policy or at least a shorter one (6 months)?
> 

If you happen to be on one of the platforms where R still works, you
can still install the optional package (or better yet, install R using
your package manager) to retain all of the old functionality.

If you're on one of the platforms where R is currently broken, nothing
is lost with respect to the new release.

We're also still stuck on an old version of R that has a security
vulnerability in CVE-2020-27637. It's a silly one, but distributions
are generally going to be working to eliminate the vulnerable versions
in favor or newer ones -- and this happens constantly. Keeping an old
version as a standard package in cases like that can force people to
install an insecure version of the package in addition to the secure
version they already have installed.


-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/682ebfafbc98f15a9cc0935e44266048268fcbac.camel%40orlitzky.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[Dima Pasechnik]

On Tue, 9 Mar 2021, 04:11 kcrisman,  wrote:

> The question is whether the R interface will remain even marginally usable
> once downgraded to optional.  It's fine to have optional packages, as long
> as there is a clear way to install them and that this is tested.  Will this
> happen?  R seems like an awfully big part of "viable competitor" to let
> that happen to.
>

as long as r2py is alive and well, it can use the, hopefully up to date, R
from the system.

Maintaining R as a Sage package, given wide availability of R on systems
Sage can run, is a burden. I would argue we ought to drop it, along with
gcc/gfortran, patch, etc.





> On Monday, March 8, 2021 at 3:53:33 AM UTC-5 Volker Braun wrote:
>
>> There are way better distributions of R than ours, just install one of
>> these and the R interface will still work. In fact, if you rely on R then
>> you shouldn't be using the outdated version in Sage...
>>
>>
>> On Monday, March 8, 2021 at 5:22:12 AM UTC+1 John H Palmieri wrote:
>>
>>> Dear all,
>>>
>>> You should be aware that ticket #31409 (
>>> https://trac.sagemath.org/ticket/31409) intends to downgrade R to an
>>> optional package because of difficulties building it on Cygwin. Just
>>> letting you know in case you care about R being part of Sage and/or you
>>> have ideas about how to fix the Cygwin build.
>>>
>>> (The branch there to downgrade R already has a positive review, by the
>>> way. I have no position on this, but I thought that more Sage developers
>>> should be aware.)
>>>
>>> --
>>> 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 view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-devel/0f71e2d0-8834-4688-b110-55cd308d2ddbn%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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAAWYfq3y2LoPY07787LX71uf178RQ5J9a8nQfYyy3D285jjNZw%40mail.gmail.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[Nathan Dunfield]

To what extent does installing optional packages "just work" with the 
current binary distributions of Sage?  I'm thinking of both those posted on 
sagemath.org as well as things not directly under our control such as the 
sage packages for conda, debian, gentoo, etc.  My past experience has been 
that "sage -i foo" works only if I had built Sage from source, though I 
haven't tried any of the binaries recently.

I bring this up because the user impact of moving R or any other package to 
optional depends tremendously on whether "sage -i R" just works.  If 
switching R to optional is tantamount to requiring users of R to build all 
of Sage from source, that would be very disruptive for those users of R in 
Sage.  Building Sage from source  is a huge hurdle for 95% users and a 
nontrivial hassle for the rest.

Best,

Nathan

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/e2bd0b88-baea-4c7d-94b8-9f519addd2d3n%40googlegroups.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[E. Madison Bray]

On Tue, Mar 9, 2021 at 2:17 PM Emmanuel Charpentier
 wrote:
>
> I'm torn :
>
> On one hand, R provides a lot (understatement...) of non-trivial numerical 
> and data-processing-related algorithms. A large part of this "lot"  comes 
> from user-written packages (17273 as of today), covering an unparalleled 
> range of use cases ; a lot of them have no equivalent in Python's 291406 (as 
> of today) packages...
> On the other hand, Windows is the dominant platform among our non-scholar 
> users (especially students) ; losing support for even a part of Sage's 
> ability on this platform  should be a big no-no.
> On the gripping hand (;-)), our current solution for Windows platform  rests 
> squarely on the shoulders of E. Madison Bray, which accomplishes alone the 
> astonishing feat of maintaining what amounts to a port (to an hostile 
> platfotm) almost alone.

FWIW, although I haven't kept up with the current problem(s?) with R
on Cygwin, I'm willing to look into solving those problems.  I have
patched R for Cygwin before.  There may also be patches from the
Cygwin package for R that we can use.

The R developers have been hostile in the past to accepting fixed for
Cygwin, but I have had no problems in the past getting patches into
the Cygwin packages.

I have no strong opinion otherwise on keeping R as a standard package
in Sage, though I have anecdotal experience that most serious R users
have no use for the R provided by Sage, and that our focus should be
on integrating with external R distributions.  This does not
necessarily mean we have to let R support in Sage become dilapidated.
We *could* make make R a "semi-standard" package in that it is not
installed by default, but our buildbots, etc. either install it by
default or install and test against an external R on the relevant
platforms...

> Le lundi 8 mars 2021 à 05:22:12 UTC+1, John H Palmieri a écrit :
>>
>> Dear all,
>>
>> You should be aware that ticket #31409 
>> (https://trac.sagemath.org/ticket/31409) intends to downgrade R to an 
>> optional package because of difficulties building it on Cygwin. Just letting 
>> you know in case you care about R being part of Sage and/or you have ideas 
>> about how to fix the Cygwin build.
>>
>> (The branch there to downgrade R already has a positive review, by the way. 
>> I have no position on this, but I thought that more Sage developers should 
>> be aware.)
>>
>> --
>> 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 view this discussion on the web visit 
> https://groups.google.com/d/msgid/sage-devel/17c8f3b8-d308-4a76-8498-f0a165947179n%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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAOTD34ZeXsjMMUO2OMPHM-XOkFtfP9A4ADrjvtwhxCU46rv_4A%40mail.gmail.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[Dima Pasechnik]

On Tue, 9 Mar 2021, 15:00 Nathan Dunfield,  wrote:

> To what extent does installing optional packages "just work" with the
> current binary distributions of Sage?  I'm thinking of both those posted on
> sagemath.org as well as things not directly under our control such as the
> sage packages for conda, debian, gentoo, etc.  My past experience has been
> that "sage -i foo" works only if I had built Sage from source, though I
> haven't tried any of the binaries recently.
>
> I bring this up because the user impact of moving R or any other package
> to optional depends tremendously on whether "sage -i R" just works.  If
> switching R to optional is tantamount to requiring users of R to build all
> of Sage from source, that would be very disruptive for those users of R in
> Sage.  Building Sage from source  is a huge hurdle for 95% users and a
> nontrivial hassle for the rest.
>

We can distribute Sage with R provided (on systems where it works at all)


>
> Best,
>
> Nathan
>
> --
> 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 view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-devel/e2bd0b88-baea-4c7d-94b8-9f519addd2d3n%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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAAWYfq0iw0-vqonfuU3f55hKKuq3KxHwUj0%3Dk-b%3DarxjgGqDPg%40mail.gmail.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[Nathan Dunfield]

An update: I just tried three different binary versions on Linux: The 
Ubuntu 20.04 binary posted at SageMath.org, the sagemath/sagemath:develop 
Docker image, and conda on RHEL 7.  None of them "just worked" with "sage 
-i foo".  The Docker image and conda failed completely with

make: *** No rule to make target 'all-toolchain'.  Stop.

I got farther with the Ubuntu binary.  Choosing "sage -i pyflakes", it 
successfully pip-installed pyflakes and then started rebuilding all of Sage 
from scratch, starting with libpng, pkgconf, etc.  So in some sense this 
worked --- I was able to abort the build and import pyflakes --- but in the 
end was equivalent to building Sage from source if I hadn't stopped it.  

Best,

Nathan




On Tuesday, March 9, 2021 at 9:00:22 AM UTC-6 Nathan Dunfield wrote:

> To what extent does installing optional packages "just work" with the 
> current binary distributions of Sage?  I'm thinking of both those posted on 
> sagemath.org as well as things not directly under our control such as the 
> sage packages for conda, debian, gentoo, etc.  My past experience has been 
> that "sage -i foo" works only if I had built Sage from source, though I 
> haven't tried any of the binaries recently.
>
> I bring this up because the user impact of moving R or any other package 
> to optional depends tremendously on whether "sage -i R" just works.  If 
> switching R to optional is tantamount to requiring users of R to build all 
> of Sage from source, that would be very disruptive for those users of R in 
> Sage.  Building Sage from source  is a huge hurdle for 95% users and a 
> nontrivial hassle for the rest.
>
> Best,
>
> Nathan
>

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/4c6b267c-b29d-4aae-8bbd-f52f7f9ae820n%40googlegroups.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[Matthias Koeppe]

Yes, there is a big problem with the current binary distributions for all 
platforms. 
What seems to be missing is a procedure to test them before releasing them. 
For those interested in helping with this, see 
https://trac.sagemath.org/ticket/31133 (Meta-ticket: Making and testing 
binary distributions)


On Tuesday, March 9, 2021 at 9:13:53 AM UTC-8 Nathan Dunfield wrote:

> An update: I just tried three different binary versions on Linux: The 
> Ubuntu 20.04 binary posted at SageMath.org, the sagemath/sagemath:develop 
> Docker image, and conda on RHEL 7.  None of them "just worked" with "sage 
> -i foo".  The Docker image and conda failed completely with
>
> make: *** No rule to make target 'all-toolchain'.  Stop.
>
> I got farther with the Ubuntu binary.  Choosing "sage -i pyflakes", it 
> successfully pip-installed pyflakes and then started rebuilding all of Sage 
> from scratch, starting with libpng, pkgconf, etc.  So in some sense this 
> worked --- I was able to abort the build and import pyflakes --- but in the 
> end was equivalent to building Sage from source if I hadn't stopped it.  
>
> Best,
>
> Nathan
>
>
>
>
> On Tuesday, March 9, 2021 at 9:00:22 AM UTC-6 Nathan Dunfield wrote:
>
>> To what extent does installing optional packages "just work" with the 
>> current binary distributions of Sage?  I'm thinking of both those posted on 
>> sagemath.org as well as things not directly under our control such as 
>> the sage packages for conda, debian, gentoo, etc.  My past experience has 
>> been that "sage -i foo" works only if I had built Sage from source, though 
>> I haven't tried any of the binaries recently.
>>
>> I bring this up because the user impact of moving R or any other package 
>> to optional depends tremendously on whether "sage -i R" just works.  If 
>> switching R to optional is tantamount to requiring users of R to build all 
>> of Sage from source, that would be very disruptive for those users of R in 
>> Sage.  Building Sage from source  is a huge hurdle for 95% users and a 
>> nontrivial hassle for the rest.
>>
>> Best,
>>
>> Nathan
>>
>

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/c8c09854-c92c-4371-a0d7-8716db7bc994n%40googlegroups.com.

[sage-devel] Re: Downgrade R to optional? See #31409.

[Matthias Koeppe]

On Tuesday, March 9, 2021 at 5:17:37 AM UTC-8 emanuel.c...@gmail.com wrote:

> An alternative would be to create an alternative Windows port relying on 
> WSL2 (which essentially runs a Linux kernel and a Linux distribution on top 
> of Windows, in native mode and with few performance impact), possibly 
> presenting less maintenance problems. This would, however, exclude support 
> of any Windows version earlier than recent Windows 10. Is that a problem ? 
> [...] Furthermore, to be realistic, we should be able to commit ourselves 
> to maintain a binary distribution for at least one WSL2-supported Linux 
> platform.
>

This is an important point. In fact, Ubuntu and perhaps other distributions 
that run on WSL (1 or 2) already package recent versions of Sage. 
So perhaps all we need to do is test that these packages work well; and 
then update our documentation to recommend one or the other to potential 
users on Windows. (For the issue of testing of downstream packaging of Sage 
- see https://trac.sagemath.org/ticket/29060)

Also other aspects of supporting Windows can be much improved by what 
basically amounts to writing documentation.

See 
- https://trac.sagemath.org/ticket/31156 (Doc: Add instructions how to run 
Sage + Jupyter notebook in WSL, browser in Windows)
- https://trac.sagemath.org/ticket/31157 (Doc: Add instructions on how to 
run the SageMath jupyter kernel in WSL, add as a kernel to Jupyter running 
natively in Windows)



 

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/d7c4fdae-e659-4534-8f30-bc32f1106f1an%40googlegroups.com.

[sage-devel] Re: Downgrade R to optional? See #31409.

[Matthias Koeppe]

On Tuesday, March 9, 2021 at 5:17:37 AM UTC-8 emanuel.c...@gmail.com wrote:

> Would it be possible to keep the R *interface* standard while relying on 
> the target platform(s) to provide the R interpreter itself (in 
> Cygwin-over-Window's case, the Cygwin "port"...). However, this would 
> create a dependence on Cygwin's version of R, not necessarily synchronized 
> with the one supported by Sage on other platforms.
>

No, we are currently unable to use system R on Cygwin - 
see https://trac.sagemath.org/ticket/30163
That's exactly what makes this issue a blocker for 9.3 -- we are neither 
able to use system R, nor build our own copy on this platform.




 

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/eb4ad84f-c9cb-44a3-999a-ef7ab8cf71ean%40googlegroups.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[Matthias Koeppe]

On Tuesday, March 9, 2021 at 6:13:38 AM UTC-8 Michael Orlitzky wrote:

> We're also still stuck on an old version of R that has a security 
> vulnerability in CVE-2020-27637. [...] Keeping an old 
> version as a standard package in cases like that can force people to 
> install an insecure version of the package in addition to the secure 
> version they already have installed. 
>

We already use system R via the spkg-configure mechanism, so our outdated 
SPKG is only used if users are on unsuitable distributions or ignore the 
system package advice that ./configure prints. Except... on Cygwin, where 
use of system R is disabled because it does not work -- 
see https://trac.sagemath.org/ticket/30163

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/23cae385-f702-4799-87a4-d6b8afb799can%40googlegroups.com.

[sage-devel] trac notification emails

[David Coudert]

I don't know why, but apparently I don't receive notification emails from 
the trac server since a couple of days. Did I miss some configuration 
change ?
Are others having the same issue ?

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/83672d87-f599-491e-a8c3-1718f707ec62n%40googlegroups.com.

Re: [sage-devel] trac notification emails

[Dima Pasechnik]

On Tue, 9 Mar 2021, 22:41 David Coudert,  wrote:

> I don't know why, but apparently I don't receive notification emails from
> the trac server since a couple of days. Did I miss some configuration
> change ?
> Are others having the same issue ?
>

I checked logs, everything is in order on the sender side, your emails are
being delivered.
e.g.

Mar  9 18:34:12 sagemath smtpd[93054]: 407e223ef83c2266 mta delivery
evpid=519550eedbe2e43f from= to=
rcpt=<-> source="46.23.93.76" relay="192.134.164.98 (
mail3-smtp-sop.national.inria.fr)" delay=2s result="Ok" stat="250 ok:
 Message 375249000 accepted"

I guess it's your side that has issues. A new spam filter in INRIA?

Dima

-- 
> 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 view this discussion on the web visit
> https://groups.google.com/d/msgid/sage-devel/83672d87-f599-491e-a8c3-1718f707ec62n%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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAAWYfq2VZmncfAJmbK%3DS_oE%2BHw%3DNnuu4r5BQdEL-SyKaDX3muQ%40mail.gmail.com.

[sage-devel] Sage has been accepted for GSoC 2021 as a mentor organization

['Travis Scrimshaw' via sage-devel]

Students will be able to apply for GSoC from March 30, 2021 - April 14, 
2021.

For those of you who have posted projects, I will be adding you as a mentor 
in the next few days.

Feel free to add additional projects to the GSoC projects page: 
https://wiki.sagemath.org/GSoC/2021

Best,
Travis

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/a85149b9-d701-493f-94dc-dc051b70a944n%40googlegroups.com.

[sage-devel] Re: Sage has been accepted for GSoC 2021 as a mentor organization

[Matthias Koeppe]

Wonderful news! Thanks, Travis, for serving as the project admin this year. 

On Tuesday, March 9, 2021 at 5:26:20 PM UTC-8 Travis Scrimshaw wrote:

> Students will be able to apply for GSoC from March 30, 2021 - April 14, 
> 2021.
>
> For those of you who have posted projects, I will be adding you as a 
> mentor in the next few days.
>
> Feel free to add additional projects to the GSoC projects page: 
> https://wiki.sagemath.org/GSoC/2021
>
> Best,
> Travis
>
>

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/6c1135fd-bef8-4626-bb61-9db89497b2b5n%40googlegroups.com.

[sage-devel] Re: Sage has been accepted for GSoC 2021 as a mentor organization

[kcrisman]

On Tuesday, March 9, 2021 at 9:23:37 PM UTC-5 Matthias Koeppe wrote:

> Wonderful news! Thanks, Travis, for serving as the project admin this 
> year. 
>
>>
>>
+1 

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/9d82e1c0-53ca-4c88-85a8-659a0a87ff47n%40googlegroups.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[kcrisman]

This has been a good conversation.  I think that, despite some emotional 
attachment to the R in Sage (and I've actually used it for research, 
because it was the only way to get a certain algorithm in Sage), having the 
R provided in Sage might not be necessary.

However, we definitely should have easy to read instructions that come up 
when anyone tries "R."  without having a system install that directs 
them how to do this.  Any more than when one has a brand new Mac OS and one 
types "git" it tells you what to do to get the command line tools.  I 
wonder if we already have that - for instance, I believe the small groups 
library for GAP has a useful error message (though it has been a while 
since I had to install it, so that might be outdated).

As for Cygwin, that seems to me to be a red herring.  If it doesn't work 
with the one in Sage AND with user-provided, maybe saying "well, it's 
supported so we should drop R" is disingenuous; is that not also the case 
for the current release of Sage?  Or did that only crop up in the current 
beta cycle?


-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/c687e67f-e904-4f70-9f06-ecea48a7cde0n%40googlegroups.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[kcrisman]

> Maintaining R as a Sage package, given wide availability of R on systems 
> Sage can run, is a burden. I would argue we ought to drop it, along with 
> gcc/gfortran, patch, etc.
>

So Numpy no longer needs gfortran? 
https://numpy.org/doc/stable/user/building.html

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/02b2fded-4e5b-4156-8627-b2a04937aeean%40googlegroups.com.

[sage-devel] Re: Downgrade R to optional? See #31409.

[kcrisman]

>
>
>
>- On the gripping hand (;-)), 
>
> I almost thought you were channeling Tevye: 
https://www.youtube.com/watch?v=bWGtjqv19ZA 

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/0386e916-4a9c-48ef-924c-9e6763ea5d92n%40googlegroups.com.

Re: [sage-devel] Re: Downgrade R to optional? See #31409.

[Matthias Koeppe]

On Tuesday, March 9, 2021 at 8:33:41 PM UTC-8 kcrisman wrote:

> If it doesn't work with the one in Sage AND with user-provided, maybe 
> saying "well, it's supported so we should drop R" is disingenuous; is that 
> not also the case for the current release of Sage?  Or did that only crop 
> up in the current beta cycle?
>

Sage 9.1 and 9.2 fully supported Cygwin at the times of their release; this 
included, of course, building R from source as an SPKG. (However, using 
system R did not work, and the same was true for system BLAS.)
Cygwin is a rolling-release platform: Packages are updated continuously. 
Even if we do not make any changes in Sage, something that used to work can 
suddenly stop working. A small number of Cygwin-specific tickets has been 
merged in the recent releases to fix what has stopped working. Well, what 
happened in the 9.3 cycle is that we can suddenly no longer build R from 
source. 

-- 
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 view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/6de8851f-4168-477e-b616-97f84ad1ce5en%40googlegroups.com.