Releasing Sage-6.0.rc0

[Volker Braun]

Sage 6.0 Release Candidate 0

We have the first release candidate for Sage 6. As you probably have heard by now, this marks the transition to git and a unified repository. As far as mathematical functionality goes, this should be identical to Sage-5.13.

Source tarball: http://boxen.math.washington.edu/home/vbraun/src/sage-6.0.rc0.tar.gz

If you use git, you can do

    git clone git://github.com/sagemath/sage.git

    cd sage

    make

And finally, as a note to the experts: There is a new branch "develop" on our git server, this will always point to the most recent beta/rc version. The master branch will always point to the most recent "release", except right now it will be identical to "develop" since there is no sage-on-git release yet.

Remaining issues:

http://trac.sagemath.org/15524 sage-fix-pkg-checksums does not preserve tarball name

http://trac.sagemath.org/15526 banner is not updated, will erroneously say Sage-5.13

http://trac.sagemath.org/15517 sage -update might not work yet

[Harald Schilly]

Cool!

Just a short question, would it be ok to use github's feature of
rolling up a .zip archive for us?
e.g. https://github.com/sagemath/sage/archive/6.0.rc0.zip
points to the tag 6.0.rc0.

Harald

> --
> 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 post to this group, send email to sage-r...@googlegroups.com.
> Visit this group at http://groups.google.com/group/sage-release.
> For more options, visit https://groups.google.com/groups/opt_out.

[Volker Braun]

The source tarball is the git repo + tarballs for standard components so you can build it without internet connection (after the initial source tarball download). The github zip would just be the repo. If we place a link somewhere on the sage web page then we need to make this distinction very clear.

[R. Andrew Ohana]

On Mon, Dec 16, 2013 at 9:27 AM, Volker Braun <vbrau...@gmail.com> wrote:

As far as mathematical functionality goes, this should be identical to Sage-5.13.

This might not be quite true -- I haven't finished exporting sage-5.13 quite yet -- so right now it should be identical to 5.13.rc0.

--

Andrew

[R. Andrew Ohana]

Never-mind, seems like nothing was added in 5.13 (except for a version change, which you already handled). (I figured something would have been done about the many tmp files that were being left around before stable status.)

--

Andrew

[Volker Braun]

There was no change between 5.13.rc0 and 5.13. So I just took the 5.13.rc0 and commited the version change on top...

[Volker Braun]

On Monday, December 16, 2013 10:17:03 PM UTC, R. Andrew Ohana wrote:

Never-mind, seems like nothing was added in 5.13 (except for a version change, which you already handled). (I figured something would have been done about the many tmp files that were being left around before stable status.)

At the end of the day, thats just a minor annoyance. Thats what tmpwatch is for.... 

[Jeroen Demeyer]

On 2013-12-16 23:17, R. Andrew Ohana wrote:
> I figured something would have been
> done about the many tmp files that were being left around before stable
> status.

Well, it's really just an annoyance. It's also not a regression (it must
have existed in Sage 5.12 also), so I found it hard to justify as
blocker issue.

Jeroen.

[Nathann Cohen]

Hellooooo !

I also got the 3+2 errors on cmdline+maxima, but also 5 in sagedev.py and 6 in patch.py. The errors are there :

http://www.steinertriples.fr/a.log

I also build it from github, not from the tarball O_o

Nathann

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

[Volker Braun]

Your git is configured to spit out color escape sequenctes in scripts, that sounds generally like a bad idea. Whats the output of 

$ git config color.ui

You should set it to "auto" to show colors interactively but not in scripts. 

Having said that, we should probably work around borked git configurations by explicitly overriding it "git diff --color=never". Volunteers?

On Tuesday, December 17, 2013 12:40:15 PM UTC, Nathann Cohen wrote:

Hellooooo !

I also got the 3+2 errors on cmdline+maxima, but also 5 in sagedev.py and 6 in patch.py. The errors are there :

http://www.steinertriples.fr/a.log

I also build it from github, not from the tarball O_o

Nathann

On 17 December 2013 10:59, Jeroen Demeyer <jdem...@cage.ugent.be> wrote:

On 2013-12-16 23:17, R. Andrew Ohana wrote:

I figured something would have been
done about the many tmp files that were being left around before stable
status.

Well, it's really just an annoyance. It's also not a regression (it must have existed in Sage 5.12 also), so I found it hard to justify as blocker issue.

Jeroen.

--
You received this message because you are subscribed to the Google Groups "sage-git" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sage-git+u...@googlegroups.com.

[Nathann Cohen]

Yoooooooooooooooo !!

> Your git is configured to spit out color escape sequenctes in scripts, that
> sounds generally like a bad idea.

Oh. So that's forbidden ? :-P

Whats the output of
>
> $ git config color.ui

Absolutely nothing. But I guess that's where the problem comes from (it appears in my .gitconfig file)

[color]
        diff = always

> You should set it to "auto" to show colors interactively but not in scripts.

Done.

Aaaaaaaaaaand all tests pass as a result !

> Having said that, we should probably work around borked git configurations
> by explicitly overriding it "git diff --color=never". Volunteers?

I can review it I guess.. But I really don't get what's going on in the dev script, so I can't do much more :-P

Nathann