[Debian-science-sagemath] Sage 7.6

[Tobias Hansen]

Hi,

turned out the floating point exceptions were caused by ECL 16.1.3
(instead of 16.1.2). This is known in the sage trac (#22191).

Now everything seems to be fine, 29 failing doctests. We can now do some
final testing and polishing and then upload the new versions of eclib,
ecl, maxima and sage 7.6 to experimental.

Julien, do we have your ok to upload eclib 20170122 when we are ready?

Regarding the failing doctests, they look mostly harmless. The following
were all already present in sage 7.5.1 (see [1]) and are still there.

- In addition to the two failed tests from singular.version() in
interfaces/singular.py there is now also a "double free or corruption"
in Singular from tests/cmdline.py.

- There's still the cython issue with factory/factory.h not being found.

- There are again two instances of gap related "gzip: stdout: Broken
pipe" although u0-fix-preexec-fn.patch was applied upstream.

Best,
Tobias

[1]
https://buildd.debian.org/status/fetch.php?pkg=sagemath&arch=amd64&ver=7.5.1-1&stamp=1489792615&raw=1

On 03/26/2017 11:54 AM, Tobias Hansen wrote:
> sage 7.6 was released yesterday. It's already packaged and the main
> blockers are that jupyter-core, jupyter-notebook and ipywidgets (and
> maybe ntl) need to be updated and several floating point exceptions
> during doctesting.