Sage 9.3.rc1 released

[Volker Braun]

As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html 

a89f81d778 (tag: 9.3.rc1, trac/develop) Updated SageMath version to 9.3.rc1

6e084f9dba Trac #31521: numpy build broken with SAGE_FAT_BINARY=yes

2c25f07cfd (tag: 9.3.rc0) Updated SageMath version to 9.3.rc0

[Matthias Köppe]

Thanks, Volker, for preparing 9.3.rc0 and .rc1. 

As the GH Actions run at https://github.com/sagemath/sage/actions/runs/699525589 shows, there are still various serious issues that need to be fixed, in particular so that we do not have regressions in platform support compared to the previous release. 

In addition, there are a number of quality-of-release issues that will be helpful to fix so that they do not become a burden to those developers who respond to user questions on sage-support/sage-devel. 

For many but not all of these issues, we already have positively reviewed tickets. Other tickets are in need of help or are waiting for review.

[Eric Gourgoulhon]

Thanks for this release. Would it be possible to merge the positively reviewed ticket

https://trac.sagemath.org/ticket/31536

in the next rc?

This ticket

- fixes the blocker https://trac.sagemath.org/ticket/31513

- fixes the %matplotlib notebook magic, which is broken in a %display latex context in Sage 9.2

- performs some code improvement regarding the handling of MathJax

- has been quite tested with Jupyter notebooks involving various types of Sage objects, Matplotlib objects and ipywidgets, so that it should be pretty safe to merge it  (although we cannot pretend to exhaustivity).

Thank you.

Eric.

[Justin C. Walker]

> On Mar 30, 2021, at 13:41 , Volker Braun <vbrau...@gmail.com> wrote:
>
> As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html

After some weird problems building Sage 9.3 from B-2 to B-8, on macOS 10.15.7, those problems went away with B9. Moreover, I was able to go back and build the back betas without problems.

The problems were always the same: running “make configure”, I see checking of RTT to various mirrors, and then the attempt to download from the chosen mirror fails. Multiple attempts, in different locations (cities) produced the same kind of failure (I don’t have log files from those attempts).

However, with rc0 and rc1, I have a real problem on all three macOS systems of mine (10.13.6, 10.14.6 and 10.15.7): the build proceeds a while, and then blows up on pynac. The log for this build, and the config.log, are attached (and I do not use any package managers).

One possibly odd thing is this line near the end of the pynac log:
ld: can't map file, errno=22 file '/Users/Sage/sage-9.3.rc1/local/lib' for architecture x86_64

Thoughts or pointers?

Thanks,

Justin

[Matthias Köppe]

On Wednesday, March 31, 2021 at 2:53:22 PM UTC-7 jus...@mac.com wrote:

with rc0 and rc1, I have a real problem on all three macOS systems of mine (10.13.6, 10.14.6 and 10.15.7): the build proceeds a while, and then blows up on pynac. The log for this build, and the config.log, are attached (and I do not use any package managers).

One possibly odd thing is this line near the end of the pynac log:
ld: can't map file, errno=22 file '/Users/Sage/sage-9.3.rc1/local/lib' for architecture x86_64

From pynac.log:

/bin/bash ../libtool  --tag=CXX   --mode=link g++ -std=gnu++11 -std=c++11  -O2 -g -O2 -g -march=native -version-info 21:5:3 -L/Users/Sage/sage-9.3.rc0/local/lib -Wl,-rpath,/Users/Sage/sage-9.3.rc0/local/lib  -o libpynac.la -rpath /Users/Sage/sage-9.3.rc0/local/lib libpynac_la-add.lo [...] libpynac_la-useries.lo -bundle -undefined dynamic_lookup -L/Users/Sage/sage-9.3.rc0/local/lib -Wl,-rpath,/Users/Sage/sage-9.3.rc0/local/lib -L. -L/Users/Sage/sage-9.3.rc0/local/lib -Wl,-rpath,/Users/Sage/sage-9.3.rc0/local/lib -L/Users/Sage/sage-9.3.rc0/local/lib -Wl,-rpath,/Users/Sage/sage-9.3.rc0/local/lib -O2 -g -march=native -lflint -lgmp  -lfactory -Wl,/Users/Sage/sage-9.3.rc0/local/lib -lflint -lmpfr -lntl -L/usr/local/lib -Wl,-rpath -Wl,/usr/local/lib -lgmp -lomalloc -L/Users/Sage/sage-9.3.rc0/local/lib -lsingular_resources    -lflint -lgmp 

Note the "-Wl,-rpath -Wl,/usr/local/lib" - which gets reordered by libtool in the next line to something that causes the error.

Correct would be "-Wl,-rpath,/usr/local/lib". It would be worth checking whether this is coming in from some environment variables that you may have set.

[Justin C. Walker]

Thanks for looking at this. See below:

Here is the environment in my normal shell:

$ env|grep lib
$

so nothing that mentions ‘lib’. I ran this in SAGE_ROOT for 9.3-rc1

Then, following the instructions near the end of the pynac log, I did the same:

(sage-buildsh) justin@Kronecker:pynac-0.7.27.p1$ env|grep lib
LDFLAGS=-L/Users/Sage/sage-9.3.rc1/local/lib -Wl,-rpath,/Users/Sage/sage-9.3.rc1/local/lib
R_MAKEVARS_SITE=/Users/Sage/sage-9.3.rc1/local/lib/R/share/Makevars.site
ECLDIR=/Users/Sage/sage-9.3.rc1/local/lib/ecl/
LIBRARY_PATH=/Users/Sage/sage-9.3.rc1/local/lib
PERL5LIB=/Users/Sage/sage-9.3.rc1/local/lib/perl5:
MPLCONFIGDIR=/Users/justin/.sage//matplotlib-1.5.1
SAGE_SPKG_INST=/Users/Sage/sage-9.3.rc1/local/var/lib/sage/installed
RANLIB=ranlib
SAGE_PARI_CFG=/Users/Sage/sage-9.3.rc1/local/lib/pari/pari.cfg
(sage-buildsh) justin@Kronecker:pynac-0.7.27.p1$

Doesn’t seem like “/usr/local/lib” is mentioned. Does ‘env’ catch all environment variables?

Justin

--
Justin C. Walker
Director
Institute for the Enhancement of the Director's Income
--
Fame is fleeting, but obscurity
just drags on and on. F&E

[Justin C. Walker]

A quick follow-up:

I see the same thing you pointed out in the rc0 log for pynac, but there is no mention at all of “/usr/local” in the pynac logs for earlier betas. This is on my 10.13.6 system, and have not updated this system’s tool chain in forever (a technical term meaning “I've forgotten the last time I did it”).

Maybe something snuck into the repository between beta9 and rc0?

Justin

--
Justin C. Walker, Curmudgeon at Large
Director
Institute for the Enhancement of the Director's income
-----------
Question 43:
What if the hokey pokey
really *is* what it’s all about?
--

[Matthias Köppe]

There are a number of recent tickets, not yet merged, that fix /usr/local leakage.

First ticket to try: https://trac.sagemath.org/ticket/31552 - singular

[Samuel Lelievre]

Thanks for this new release. Would a ticket upgrading

to the latest OpenSSL still be considered for Sage 9.3?

Or should we just strongly encourage users to install

OpenSSL via some package manager and build Sage

from source using that?

[Kenji Iohara]

On my mac OS 11.2.3, I could compile it from the first built (even if I had to compile it twice).

But for ptestlong, it didn't work; the last word I saw on the screen was <<too many open files. >>

Here is its log-file. 

2021年4月1日(木) 17:24 Samuel Lelievre <samuel....@gmail.com>:

--
You received this message because you are subscribed to the Google Groups "sage-release" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/4eb78673-ec4e-4cb9-b0fc-a4107373ab28n%40googlegroups.com.

[Justin C. Walker]

> On Apr 1, 2021, at 08:04 , Matthias Köppe <matthia...@gmail.com> wrote:
>
> There are a number of recent tickets, not yet merged, that fix /usr/local leakage.
> First ticket to try: https://trac.sagemath.org/ticket/31552 - singular

Thanks! That looks like it should work.

It having been a while since I’ve done this, got a pointer to instructions for incorporating the results of this trac into my local system?

Justin

--
Justin C. Walker, Curmudgeon-At-Large
Director
Institute for the Enhancement of the Director's Income
--------
"Weaseling out of things is what separates us from the animals.
Well, except the weasel."
- Homer J Simpson
--------

[Dima Pasechnik]

On Fri, Apr 2, 2021 at 1:09 AM 'Justin C. Walker' via sage-release
<sage-r...@googlegroups.com> wrote:
>
>
>
> > On Apr 1, 2021, at 08:04 , Matthias Köppe <matthia...@gmail.com> wrote:
> >
> > There are a number of recent tickets, not yet merged, that fix /usr/local leakage.
> > First ticket to try: https://trac.sagemath.org/ticket/31552 - singular
>
> Thanks! That looks like it should work.
>
> It having been a while since I’ve done this, got a pointer to instructions for incorporating the results of this trac into my local system?

git pull trac u/mkoeppe/_usr_local_leaks_into_singular_build
(assuming trac is the name of the git remote)

>
> Justin
>
> --
> Justin C. Walker, Curmudgeon-At-Large
> Director
> Institute for the Enhancement of the Director's Income
> --------
> "Weaseling out of things is what separates us from the animals.
> Well, except the weasel."
> - Homer J Simpson
> --------
>
>

> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/EF0CE8C4-06F3-4014-B5B2-408614C2AAA4%40mac.com.