[sage-devel] Re: Allow for --user pip installs for sage?

[Nils Bruin]

On Monday, 27 September 2021 at 14:45:44 UTC-7 Matthias Koeppe wrote:

> Much better than any of these options is to use user-defined venvs. 
> This has been available since Sage 9.2:
>
> https://wiki.sagemath.org/ReleaseTours/sage-9.2#Reusable_wheels_for_the_Python_packages_built_by_the_Sage_distribution
>
> Cool! Whatever "make build" does will be what's relevant for most people 
(and that includes me). I assume that's also how binary distributions will 
package it.

-- 
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/f5e71b1c-5aec-4d2a-9dcd-1c6d2684a311n%40googlegroups.com.

[sage-devel] Sympow build warning

[John H Palmieri]

On an OS X machine, the sympow build prints a warning at the start. Well, 
it calls it an error and then proceeds to build anyway:

-

CFLAGS for SYMPOW:  -std=gnu17 -fno-fast-math -mfpmath=sse -ffp-contract=on 
-Dx86  -O0
The double precision of your FPU is 105 bits.
Error: the Quad Double library used by SYMPOW assumes IEEE-754 double
precision numbers with exactly 53 bits in the mantissa (64 bits in
total).  Unfortunately, this is not the case on your system and we
currently have no workaround for your system.  Running SYMPOW will
almost certainly fail on some inputs.

Please report this problem to sage-devel
(http://groups.google.com/group/sage-devel), mentioning in particular
your operating system, processor type and compiler version
(run gcc --version).

-

So I'm reporting to sage-devel. This is with OS X 11.6. For processor type:

% sysctl -n machdep.cpu.brand_string
Intel(R) Xeon(R) W-2140B CPU @ 3.20GHz

And gcc:

% gcc --version
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr 
--with-gxx-include-dir=/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/usr/include/c++/4.2.1
Apple clang version 13.0.0 (clang-1300.0.29.3)
Target: x86_64-apple-darwin20.6.0
Thread model: posix
InstalledDir: 
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin

-- 
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/7cf8fac7-e81b-4d6c-aeb3-275260b21d4en%40googlegroups.com.

[sage-devel] Re: Allow for --user pip installs for sage?

[Matthias Koeppe]

On Monday, September 27, 2021 at 1:24:47 PM UTC-7 wst...@gmail.com wrote:

> On Monday, September 27, 2021 at 1:06:58 PM UTC-7 Matthias Koeppe wrote:
>
>> On Monday, September 27, 2021 at 12:46:30 PM UTC-7 Nils Bruin wrote:
>>
>>>
>>> Presently,
>>>
>>> $ sage -pip install --user .
>>>
>>> fails (at least for me, with a vanilla sage install on FC33) with
>>>
>>> ERROR: Can not perform a '--user' install. User site-packages are not 
>>> visible in this virtualenv.
>>>
>>> Perhaps it's worth having this?
>>>
>> This is deliberately disabled (look for PYTHONUSERBASE in 
>> src/bin/sage-env) because users tend to have random trash installed in the 
>> user scheme. Because these packages would take precedence over 
>> Sage-provided packages, this would lead to endless bug reports on the 
>> mailing lists.
>>
>
> However, as I mentioned in my other email, we explicitly enable it back in 
> CoCalc by patching the 
> Sage install. [...]
> I think the best solution for the official Sage distribution is to provide 
> an *option* for systems admins installing
> Sage to all or not allow user pip installs.  Probably the default should 
> be "no", but there should be a documented
> supported way of changing the behavior to "yes". 
>

Much better than any of these options is to use user-defined venvs. 
This has been available since Sage 9.2:
https://wiki.sagemath.org/ReleaseTours/sage-9.2#Reusable_wheels_for_the_Python_packages_built_by_the_Sage_distribution


-- 
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/3743857f-6b17-48e6-8496-6527882b7f3fn%40googlegroups.com.

Re: [sage-devel] Re: #32532 - removing gcc and gfortran spkgs

[Volker Braun]

On Monday, September 27, 2021 at 1:29:23 AM UTC+2 dim...@gmail.com wrote:

> Homebrew need not be installed into /usr/local, and doesn't need root if 
> one chooses not to install there.
>

You can theoretically untar homebrew in a different directory, but at least 
when I tried that a couple of years ago I found myself up in a shit creek 
without a paddle. Have you actually tried? The installation document says:

However do yourself a favour and use the installer to install to the 
default prefix. Some things may not build when installed elsewhere. One of 
the reasons Homebrew just works relative to the competition is because we 
recommend installing here. Pick another prefix at your peril!

-- 
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/67bd3615-2c52-4c19-b68a-70f2187e092cn%40googlegroups.com.

[sage-devel] Re: Allow for --user pip installs for sage?

[Nils Bruin]

On Monday, 27 September 2021 at 13:06:58 UTC-7 Matthias Koeppe wrote:

> On Monday, September 27, 2021 at 12:46:30 PM UTC-7 Nils Bruin wrote:
>
>>
>> $ sage -pip install --user .
>>
> This is deliberately disabled (look for PYTHONUSERBASE in 
> src/bin/sage-env) because users tend to have random trash installed in the 
> user scheme. Because these packages would take precedence over 
> Sage-provided packages, this would lead to endless bug reports on the 
> mailing lists.
>
> If we'd mangle the version of the sage-installed python (make it 
python3.9sage or something like that), it would make it very unlikely we'd 
stumble into random stuff the user didn't explicitly install in sage's 
python. Interestingly enough, this would also automatically prevent system 
libraries from taking precedence over the sage-provided ones. I don't know 
how doable this mangling is?

Of course, all of this goes out the window if one is running sage using the 
system python (modulo virtualenv's, which I don't know anything about). But 
I think it should.

-- 
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/d09da424-9895-44fe-b93c-f5f1e002529dn%40googlegroups.com.

Re: [sage-devel] Allow for --user pip installs for sage?

[Thierry]

Hi,

for what it worth, i am +1 for letting the random user to pip-install
her own packages, either on ~/.local as William suggested, and/or in
~/.sage/local as it used to be the case (this location ensures that 
Python packages unrelated to some Sage activity will not add noise).

Ciao,
Thierry




On Mon, Sep 27, 2021 at 01:16:16PM -0700, William Stein wrote:
> On Mon, Sep 27, 2021 at 12:46 PM Nils Bruin  wrote:
> >
> >
> > Presently,
> >
> > $ sage -pip install --user .
> >
> > fails (at least for me, with a vanilla sage install on FC33) with
> >
> > ERROR: Can not perform a '--user' install. User site-packages are not 
> > visible in this virtualenv.
> >
> > Perhaps it's worth having this?
> >
> > Also, I noticed:
> >
> > $ sage --python3 -c "import sys; print(sys.path)" ['', 
> > '/usr/lib64/python39.zip', '/usr/lib64/python3.9', 
> > '/usr/lib64/python3.9/lib-dynload', 
> > '/usr/local/sage/sage-git/local/lib64/python3.9/site-packages', 
> > '/usr/local/sage/sage-git/local/lib/python3.9/site-packages']
> >
> > note that the first enstries in this list actually refer to the system 
> > python, not the sage python! I think that means that any patched package 
> > installs that sage uses will be shadowed by the system installations. 
> > That's probably not what is meant.
> 
> On CoCalc I think we explicitly patch the Sage we provide to include ~/.local:
> 
> ~$ sage --python3 -c "import sys; print(sys.path)"
> ['', '/ext/sage/sage-9.3/local/lib/python39.zip',
> '/ext/sage/sage-9.3/local/lib/python3.9',
> '/ext/sage/sage-9.3/local/lib/python3.9/lib-dynload',
> '/home/user/.local/lib/python3.9/site-packages',
> '/ext/sage/sage-9.3/local/lib/python3.9/site-packages',
> '/ext/sage/sage-9.3/local/lib/python3.9/site-packages/awalipy-0.4-py3.9-linux-x86_64.egg']
> 
> Also note that all the other path entries are our sage install (not
> the system python, which is crazy, as you point out).
> 
> For us, it's very very important that "pip install --user" just works.
> Also, for us, usually people are working in a very clean well defined
> environment (a cocalc project), which they are using for a  very
> specific "project" (e.g., teaching or taking a specific class, or
> writing a paper), so using ~/.local is very reasonable.I see at
> [1] that our page about installing packages "into sage is not
> supported", and it should be updated to say that "sage --pip install"
> does work.   We also make pip default to --user so actually "pip
> install" does just work.
> 
> I just found a "random" package on pypi:
> 
> ~$ sage --pip install pytype
> Defaulting to user installation because normal site-packages is not writeable
> Collecting pytype
> ...
> ~$ sage
> ┌┐
> │ SageMath version 9.3, Release Date: 2021-05-09 │
> │ Using Python 3.9.2. Type "help()" for help.│
> └┘
> sage: import pytype
> sage:  # yeah
> 
> IMHO, this is pretty good user experience.
> 
> Sometimes people do break their sage installs via installing stuff
> into ~/.local, but it's also easy to fix by just deleting ~/.local
> and starting over, and it's usually very clear this is the problem
> based on the traceback.
> 
>  -- William
> 
> -- 
> 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/CACLE5GD2n47_2QHpPNa82NqypuEqC6iV2Djg_0PTsy_%2BdqLwpA%40mail.gmail.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/20210927203342.GA4053%40metelu.net.

[sage-devel] Re: Allow for --user pip installs for sage?

[William Stein]

On Monday, September 27, 2021 at 1:06:58 PM UTC-7 Matthias Koeppe wrote:

> On Monday, September 27, 2021 at 12:46:30 PM UTC-7 Nils Bruin wrote:
>
>>
>> Presently,
>>
>> $ sage -pip install --user .
>>
>> fails (at least for me, with a vanilla sage install on FC33) with
>>
>> ERROR: Can not perform a '--user' install. User site-packages are not 
>> visible in this virtualenv.
>>
>> Perhaps it's worth having this?
>>
>>
> This is deliberately disabled (look for PYTHONUSERBASE in 
> src/bin/sage-env) because users tend to have random trash installed in the 
> user scheme. Because these packages would take precedence over 
> Sage-provided packages, this would lead to endless bug reports on the 
> mailing lists.
>

This is a very good point.

However, as I mentioned in my other email, we explicitly enable it back in 
CoCalc by patching the 
Sage install.  This leads to probably one bug report a month, if that (so 
yes, it's endless, but very easy to
manage), and I'm not sure those are even Sage related, since the same 
problem can happen with any
Python install.  We make bug reports very, very easy for users to create 
(it's 2 clicks), and there is a lot [1] of usage 
of Sage on CoCalc.   The main thing that  makes explicitly enabling this 
pip functionality a good idea for cocalc
is that cocalc projects are clean Docker containers, and people use them 
for specific purposes.  Also support
is stupid easy due to it being a web application.   So the problem you 
mention with random trash goes away
*for us*.

I think the best solution for the official Sage distribution is to provide 
an *option* for systems admins installing
Sage to all or not allow user pip installs.  Probably the default should be 
"no", but there should be a documented
supported way of changing the behavior to "yes".  I don't like having to 
patch sage just because of this, but I do,
and *definitely* will continue to do so, because the reasoning for removing 
this functionality does not apply
to my use case. 

William

 [1] https://cocalc.com/info/status

 
>

-- 
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/7a0a19f0-efe1-4301-8837-85fa6c73c5aan%40googlegroups.com.

Re: [sage-devel] Allow for --user pip installs for sage?

[William Stein]

On Mon, Sep 27, 2021 at 12:46 PM Nils Bruin  wrote:
>
>
> Presently,
>
> $ sage -pip install --user .
>
> fails (at least for me, with a vanilla sage install on FC33) with
>
> ERROR: Can not perform a '--user' install. User site-packages are not visible 
> in this virtualenv.
>
> Perhaps it's worth having this?
>
> Also, I noticed:
>
> $ sage --python3 -c "import sys; print(sys.path)" ['', 
> '/usr/lib64/python39.zip', '/usr/lib64/python3.9', 
> '/usr/lib64/python3.9/lib-dynload', 
> '/usr/local/sage/sage-git/local/lib64/python3.9/site-packages', 
> '/usr/local/sage/sage-git/local/lib/python3.9/site-packages']
>
> note that the first enstries in this list actually refer to the system 
> python, not the sage python! I think that means that any patched package 
> installs that sage uses will be shadowed by the system installations. That's 
> probably not what is meant.

On CoCalc I think we explicitly patch the Sage we provide to include ~/.local:

~$ sage --python3 -c "import sys; print(sys.path)"
['', '/ext/sage/sage-9.3/local/lib/python39.zip',
'/ext/sage/sage-9.3/local/lib/python3.9',
'/ext/sage/sage-9.3/local/lib/python3.9/lib-dynload',
'/home/user/.local/lib/python3.9/site-packages',
'/ext/sage/sage-9.3/local/lib/python3.9/site-packages',
'/ext/sage/sage-9.3/local/lib/python3.9/site-packages/awalipy-0.4-py3.9-linux-x86_64.egg']

Also note that all the other path entries are our sage install (not
the system python, which is crazy, as you point out).

For us, it's very very important that "pip install --user" just works.
Also, for us, usually people are working in a very clean well defined
environment (a cocalc project), which they are using for a  very
specific "project" (e.g., teaching or taking a specific class, or
writing a paper), so using ~/.local is very reasonable.I see at
[1] that our page about installing packages "into sage is not
supported", and it should be updated to say that "sage --pip install"
does work.   We also make pip default to --user so actually "pip
install" does just work.

I just found a "random" package on pypi:

~$ sage --pip install pytype
Defaulting to user installation because normal site-packages is not writeable
Collecting pytype
...
~$ sage
┌┐
│ SageMath version 9.3, Release Date: 2021-05-09 │
│ Using Python 3.9.2. Type "help()" for help.│
└┘
sage: import pytype
sage:  # yeah

IMHO, this is pretty good user experience.

Sometimes people do break their sage installs via installing stuff
into ~/.local, but it's also easy to fix by just deleting ~/.local
and starting over, and it's usually very clear this is the problem
based on the traceback.

 -- William

-- 
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/CACLE5GD2n47_2QHpPNa82NqypuEqC6iV2Djg_0PTsy_%2BdqLwpA%40mail.gmail.com.

[sage-devel] Re: Allow for --user pip installs for sage?

[Matthias Koeppe]

On Monday, September 27, 2021 at 12:46:30 PM UTC-7 Nils Bruin wrote:

>
> Presently,
>
> $ sage -pip install --user .
>
> fails (at least for me, with a vanilla sage install on FC33) with
>
> ERROR: Can not perform a '--user' install. User site-packages are not 
> visible in this virtualenv.
>
> Perhaps it's worth having this?
>
>
This is deliberately disabled (look for PYTHONUSERBASE in src/bin/sage-env) 
because users tend to have random trash installed in the user scheme. 
Because these packages would take precedence over Sage-provided packages, 
this would lead to endless bug reports on the mailing lists.


 

-- 
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/1a260f60-8574-4402-91c4-384508ecaa60n%40googlegroups.com.

[sage-devel] Re: Allow for --user pip installs for sage?

[Matthias Koeppe]

On Monday, September 27, 2021 at 12:46:30 PM UTC-7 Nils Bruin wrote:

> I noticed:
>
> $ sage --python3 -c "import sys; print(sys.path)" ['', 
> '/usr/lib64/python39.zip', '/usr/lib64/python3.9', 
> '/usr/lib64/python3.9/lib-dynload', 
> '/usr/local/sage/sage-git/local/lib64/python3.9/site-packages', 
> '/usr/local/sage/sage-git/local/lib/python3.9/site-packages'] 
>
> note that the first enstries in this list actually refer to the system 
> python, not the sage python! I think that means that any patched package 
> installs that sage uses will be shadowed by the system installations.
>

No, that's not how it works. Sage uses a venv over the system python. The 
first items in the list provide access to python's standard library. 
System-installed packages would be found in a directory like 
/usr/lib64/python3.9/site-packages. Our venv is configured so that it does 
not provide access to system packages, which is why you don't see this 
directory in the path.


 

-- 
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/2b0e4dc9-2664-484d-8131-1e270657cd35n%40googlegroups.com.

[sage-devel] Allow for --user pip installs for sage?

[Nils Bruin]

Presently,

$ sage -pip install --user .

fails (at least for me, with a vanilla sage install on FC33) with

ERROR: Can not perform a '--user' install. User site-packages are not 
visible in this virtualenv.

Perhaps it's worth having this?

Also, I noticed:

$ sage --python3 -c "import sys; print(sys.path)" ['', 
'/usr/lib64/python39.zip', '/usr/lib64/python3.9', 
'/usr/lib64/python3.9/lib-dynload', 
'/usr/local/sage/sage-git/local/lib64/python3.9/site-packages', 
'/usr/local/sage/sage-git/local/lib/python3.9/site-packages'] 

note that the first enstries in this list actually refer to the system 
python, not the sage python! I think that means that any patched package 
installs that sage uses will be shadowed by the system installations. 
That's probably not what is meant.

-- 
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/59cc60b7-b4ec-4644-a40b-852e4a617357n%40googlegroups.com.

Re: [sage-devel] incomprehensible merge failure on trac

[Dima Pasechnik]

for the record:
to get your branch on: (you probably don't need these 2 lines)

git fetch trac public/rings/lazy_dirichlet_series-32309
git checkout -b public/rings/lazy_dirichlet_series-32309
trac/public/rings/lazy_dirichlet_series-32309

merge the upstream devel branch (i.e. 9.5.beta2):

git pull trac/develop
git push trac HEAD



On Mon, Sep 27, 2021 at 12:44 PM Dima Pasechnik  wrote:

> I just merged the latest devel branch into your ticket branch
>
> On Mon, Sep 27, 2021 at 12:34 PM 'Martin R' via sage-devel <
> sage-devel@googlegroups.com> wrote:
>
>>
>> The branch field on https://trac.sagemath.org/ticket/32309 is currently
>> red, but "git trac try 32309" does not report a merge failure.
>>
>> What can I do?
>>
>> Martin
>>
>> --
>> 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/3de0f90f-05d6-4d50-8555-98a9ba7d3aebn%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/CAAWYfq3OvtTzFqDr6-h6JxOePUoCBvwhPd49mMKMMBudwx2Y%3DA%40mail.gmail.com.

Re: [sage-devel] incomprehensible merge failure on trac

[Dima Pasechnik]

I just merged the latest devel branch into your ticket branch

On Mon, Sep 27, 2021 at 12:34 PM 'Martin R' via sage-devel <
sage-devel@googlegroups.com> wrote:

>
> The branch field on https://trac.sagemath.org/ticket/32309 is currently
> red, but "git trac try 32309" does not report a merge failure.
>
> What can I do?
>
> Martin
>
> --
> 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/3de0f90f-05d6-4d50-8555-98a9ba7d3aebn%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/CAAWYfq3vh4%3Dy-9BFp7JpVAsG2JKbhv7BNvQ5qk30MF7uk%2B_7Rw%40mail.gmail.com.

[sage-devel] incomprehensible merge failure on trac

['Martin R' via sage-devel]

The branch field on https://trac.sagemath.org/ticket/32309 is currently 
red, but "git trac try 32309" does not report a merge failure.

What can I do?

Martin

-- 
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/3de0f90f-05d6-4d50-8555-98a9ba7d3aebn%40googlegroups.com.

Re: [sage-devel] Re: #32532 - removing gcc and gfortran spkgs

[Samuel Lelièvre]

2021-09-27 à 08:32 UTC, Samuel Lelievre:
>
> Nathan,
>
> I believe William is talking about "sage-forge" or "sagemath-forge"
> as discussed recently on sage-devel in these threads:
>
> https://groups.google.com/g/sage-devel/c/JYwHrmcqNhc
> https://groups.google.com/g/sage-devel/c/QeYle_D8Otc
> https://groups.google.com/g/sage-devel/c/NfUKjAaTcUg
>
> I'm hoping to test it this week.  --Samuel

Scratch the first link (from 12 years 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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sage-devel/CAEcArF3Vr2pWPM_Jfz6k_55Fh4Qz_tNFKEq2dcapQyzufWyq8A%40mail.gmail.com.

Re: [sage-devel] Re: #32532 - removing gcc and gfortran spkgs

[Samuel Lelievre]

2021-09-27 02:03:51 UTC, Nathan Dunfield:

> On Friday, September 24, 2021 at 1:12:38 PM UTC-5 William Stein wrote:
>
>> I assume you are talking about the official binaries that are distributed
>>
> on Sagemath.org.  Fortunately, the Sage binaries on MacOS that are
>>
> produced by the conda-forge devs are not total crap.  
>>
>
> William,
>
> There are two different Sage binaries for macOS referred to on 
> SageMath.org. 
>
The first, which is the recommended one is:
>
> https://github.com/3-manifolds/Sage_macOS/releases
>
> created by Marc Culler over the past year and is very far from total crap.
>
While it does not yet support the M1 natively, it does offer a code-signed
>
and notarized app that is just as easy to install as Mathematica et al.
>
(Note that the JupyterLab App is *not* notarized so one gets the
>
"macOS cannot verify that this app is free from malware" message
>
and have to go through System Preferences to run the installer.)
>
> The second is the old:
>
> http://mirrors.mit.edu/sage/osx/intel/index.html
>
> Best,
>
> Nathan
>

Nathan,

I believe William is talking about "sage-forge" or "sagemath-forge"
as discussed recently on sage-devel in these threads:

https://groups.google.com/g/sage-devel/c/JYwHrmcqNhc
https://groups.google.com/g/sage-devel/c/QeYle_D8Otc
https://groups.google.com/g/sage-devel/c/NfUKjAaTcUg

I'm hoping to test it this week.  --Samuel

-- 
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/25b4851d-a476-4c3d-8e3c-a8a31d2a1a40n%40googlegroups.com.