Cygwin(64) port status

[Jean-Pierre Flori]

Hi all,

With http://trac.sagemath.org/ticket/15649 and http://trac.sagemath.org/ticket/17365 which need reviews and touch Cygwin-only code we'll (almost) have support to build Sage trivially (type "make") on Cygwin (again) and Cygwin64 (for the first time).
Only the MPIR update at http://trac.sagemath.org/ticket/15015 will be missing for Cygwin64. It is ready for merging, but it may need some more discussion before being merged.

Best,
JP

[David Roe]

Nice! Well done to those of you who have been working on the Cygwin port.
David
>
> Best,
> JP
>
> --
> 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+...@googlegroups.com.
> To post to this group, send email to sage-...@googlegroups.com.
> Visit this group at http://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.

[kcrisman]

>
> With http://trac.sagemath.org/ticket/15649 and
> http://trac.sagemath.org/ticket/17365 which need reviews and touch
> Cygwin-only code we'll (almost) have support to build Sage trivially (type
> "make") on Cygwin (again) and Cygwin64 (for the first time).
> Only the MPIR update at http://trac.sagemath.org/ticket/15015 will be
> missing for Cygwin64. It is ready for merging, but it may need some more
> discussion before being merged.

Nice!  Well done to those of you who have been working on the Cygwin port.
David
>

Yes, congrats!  Sorry I haven't been able to help much the past year or more. 

[Jean-Pierre Flori]

Note that everything gets built, but at this point I don't promise it is really functional.

Cygwin32 should be almost fully functional.
Cygwin64 is very far from it (e.g. PARI builds but seems completely broken: it says 131 is not prime).

Anyway getting back to a point where just typing "make" would finish without errors and build everything without requiring human interaction was a first step before debugging libraries behaving badly..

[mmarco]

The result of the compilation is relocatable? I mean, would it be eventually possible to have something that windows users just unzip and runs? Or would they always need to compile it?

[Jean-Pierre Flori]

On Tuesday, November 25, 2014 10:46:56 PM UTC+1, mmarco wrote:

The result of the compilation is relocatable? I mean, would it be eventually possible to have something that windows users just unzip and runs? Or would they always need to compile it?

I would say so, i dont really know if any paths get hardcoded but would be inclined to say it can be fixed just as on Linux.
For sure some decision need to be made (should we provide a full cygwin install or just the sage folder?) and it will need more work to ensure we ship a one click install zipfile (what to chose as a base address for dlls if we only ship Sage or do we run a rebase script once when we install?).

[Jean-Pierre Flori]

And I forgot http://trac.sagemath.org/ticket/17365 and http://trac.sagemath.org/ticket/15649 which need a little love (and as it only affects Cygwin should be easy to review, at least if you trust me)!!!

[Jean-Pierre Flori]

On Thursday, November 27, 2014 6:39:13 PM UTC+1, Jean-Pierre Flori wrote:

And I forgot http://trac.sagemath.org/ticket/17365 and http://trac.sagemath.org/ticket/15649 which need a little love (and as it only affects Cygwin should be easy to review, at least if you trust me)!!!

In addition of the two above still needing review, http://trac.sagemath.org/ticket/17358 is also needed!

And if anyone has anything to say about updating to an MPIR alpha as there is a strong probability the final version will be released in a long time, please have a look at http://trac.sagemath.org/ticket/15015.

[Travis Scrimshaw]

We've included dev releases before into Sage, and having closer to out-of-the-box support for Cygwin64 gets a +1 from me.

Best,
Travis

PS - I should have some time to look again at Cygwin stuff soon.

[Jean-Pierre Flori]

Bumpy bump.
http://trac.sagemath.org/ticket/17365
http://trac.sagemath.org/ticket/15649

These two are among the ones Karl-Dieter mentioned the other day and are really easy, you don't even need to know how Cygwin works, just check they make sense!
Being able to work on Cygwin(64) from Sage 6.5 would be nice!

[Jean-Pierre Flori]

Thanks to Travis, only the ATLAS patch needs review!
http://trac.sagemath.org/ticket/17365

[John Cremona]

By the way, there is discussion of building Pari under Cygwin 64 on
the pari-dev list; bugs are being discovered and fixed as I type.

John

[Jean-Pierre Flori]

On Tuesday, January 6, 2015 3:51:29 PM UTC+1, John Cremona wrote:

By the way, there is discussion of building Pari under Cygwin 64 on
the pari-dev list;  bugs are being discovered and fixed as I type.

Great news!
My last findings were that PARI behaved very badly on Cygwin64.
For example 131 was not prime.
(And I'm not receiving anything from pari-dev yet though I am subscribed to it, strange.)

[Jean-Pierre Flori]

Oh, it was pari-users.