Update to Frobby 0.9.0

[mwhansen]

The spkg is at

http://sage.math.washington.edu/home/mhansen/frobby-0.9.0.spkg

This should fix #7436 and #8783.

CC: @sagetrac-drkirkby

Component: packages: experimental

Keywords: sd40.5

Author: Mike Hansen

Reviewer: Benjamin Jones

Issue created by migration from https://trac.sagemath.org/ticket/13007

[benjaminfjones]

Reviewer: Benjamin Jones

[benjaminfjones]

Changed keywords from none to sd40.5

[benjaminfjones]

comment:2

The new spkg builds successfully using gcc 4.6.3 and SAGE_CHECK=yes on x86_64 debian squeeze. I'll test some other archs / platforms tomorrow, but looks good so far!

[benjaminfjones]

comment:3

The new spkg builds successfully using gcc 4.7.0 and SAGE_CHECK=yes on x86_64 Fedora 16 (skynet/eno). I'll test ubuntu and call it a wrap.

[benjaminfjones]

comment:4

Building with SAGE_CHECK=yes on Ubunutu with gcc 4.4.3 (redhawk) works as well. I'm giving this a positive review.

[benjaminfjones]

comment:5

I changed the milestones for #7436 and #8783 to duplicate since they are fixed by this ticket.

[haraldschilly]

comment:8

i just added it to the experimental spkges + updated master mirror

[kcrisman]

comment:9

See also #8515 - Ben, do you have access to one of those Solaris machines?

I just had it pass with SAGE_CHECK on Mac OS. However, the tests in sage/interfaces/frobby.py don't pass.

sage -t --optional "devel/sage/sage/interfaces/frobby.py"   
**********************************************************************
File "/Users/.../sage-5.1.beta1-flask/devel/sage/sage/interfaces/frobby.py", line 42:
    sage: frobby("analyze", input="vars x,y,z;[x^2,x*y];", #optional
          options=["lcm", "iformat monos", "oformat 4ti2"]) #optional
Expected:
    'x^2*y\n'
Got:
    ' 2 1 0\n\n2 generators\n3 variables\n'
**********************************************************************
File "/Users/.../sage-5.1.beta1-flask/devel/sage/sage/interfaces/frobby.py", line 47:
    sage: frobby("do_dishes") #optional
Expected:
    Traceback (most recent call last):
    ...
    RuntimeError: Frobby reported an error:
    ERROR: Unknown action "do_dishes".
Got:
    Traceback (most recent call last):

    RuntimeError: Frobby reported an error:
    ERROR: No action has the prefix "do_dishes".
    <BLANKLINE>
**********************************************************************
File "/Users/.../sage-5.1.beta1-flask/devel/sage/sage/interfaces/frobby.py", line 96:
    sage: decom = frobby.irreducible_decomposition(id) #optional
Exception raised:
 
        raise RuntimeError("Format error: incorrect matrix dimensions.")
    RuntimeError: Format error: incorrect matrix dimensions.
**********************************************************************
File "/Users/.../sage-5.1.beta1-flask/devel/sage/sage/interfaces/frobby.py", line 98:
    sage: set(decom) == set(true_decom) # use sets to ignore order #optional
Exception raised:

        set(decom) == set(true_decom) # use sets to ignore order #optional###line 98:
    sage: set(decom) == set(true_decom) # use sets to ignore order #optional
    NameError: name 'decom' is not defined
**********************************************************************
File "/Users/.../sage-5.1.beta1-flask/devel/sage/sage/interfaces/frobby.py", line 117:
    sage: for ring in rings: #optional
          id1 = ring.ideal(Integer(1)) #optional
          decom1 = frobby.irreducible_decomposition(id1) #optional
          allOK = allOK and decom1 == [id1] #optional
Exception raised:

    RuntimeError: Format error: encountered non-number.
*******************************************************************

Otherwise we could really make it an optional package again after all this testing if it worked on Solaris.

[benjaminfjones]

comment:10

I don't have access to drkirkby's machines. Maybe we can get him to test the package out. In any case I don't know anything about Solaris or SPARC platforms.