Opened 13 years ago

Closed 13 years ago

# Update Maxima to 5.20.1

Reported by: Owned by: Karl-Dieter Crisman tbd major sage-4.3.1 packages: standard maxima Robert Marik, Burcin Erocal sage-4.3.1.alpha0 Karl-Dieter Crisman Robert Marik N/A

### Description

Maxima is now updated, and that will incorporate a number of our bugfixes as well as things we have reported, not to mention other improvements. See http://groups.google.com/group/sage-devel/browse_thread/thread/bbd2c801032392f7 for the announcement.

### comment:2 follow-up:  4 Changed 13 years ago by Karl-Dieter Crisman

Status: new → needs_work

Okay, I have a successful spkg. BUT there are a number of things that will have to be fixed. I have most of them (just doctests) but there are a couple bigger ones. So this needs work (and a patch), but hopefully will be ready for 4.3.1, when that comes.

1. For some reason, a certain integration with infinity is not working properly. Maybe Maxima regressed on it? Or maybe we aren't parsing it properly?
```sage: integrate(t*cos(-theta*t),(t,-oo,oo))
```
1. Because of overall improvements to to_poly_solve/%solve, there are some annoying things we will have to fix. Some are just in parsing the new %solve and some other things from its new capabilities, like
```TypeError: unable to make sense of Maxima expression '[If(cos(pi*...!=0,[x=-...],union())]' in Sage
```

though this used to be nicely behaved from

```sage: solve(cos(x) * sin(x) == 1/2, x, to_poly_solve=True)
```

but unfortunately one of them is yet another hang in the algsys which doesn't automatically resolve (this is line 5948 in symbolic/expression.pyx):

```sage: a = .004*(8*e^(-(300*t)) - 8*e^(-(1200*t)))*(720000*e^(-(300*t)) - 11520000*e^(-(1200*t))) +.004*(9600*e^(-(1200*t)) - 2400*e^(-(300*t)))^2
sage: a.solve(t, to_poly_solve=True)
<hang>
```

I'm contacting the author of that for info.

### comment:4 in reply to:  2 Changed 13 years ago by Karl-Dieter Crisman

1. For some reason, a certain integration with infinity is not working properly. Maybe Maxima regressed on it? Or maybe we aren't parsing it properly?
```sage: integrate(t*cos(-theta*t),(t,-oo,oo))
```

Update: this integral doesn't converge! It was reported in #6816 but we never checked that it made sense, since Maxima did give an answer - zero, because the limit of the indefinite integral from -N to N is zero. Maxima now (sensibly) doesn't give that any more, though it would be even better if it returned divergent; however, that would be a different ticket.

Still working on fixing the remaining doctest issues.

### comment:5 Changed 13 years ago by Karl-Dieter Crisman

Status: needs_work → needs_review

Came up with fixes of one kind or another for all these things. The real remaining issue is that there isn't a super-simple way to parse %if when it comes from to_poly_solve, so for now I am just leaving that alone. As soon as anyone can think of a way to do this, please open a ticket and post a patch. Otherwise this patch should make the new spkg work.

### Changed 13 years ago by Karl-Dieter Crisman

Based on 4.3.alpha1

### comment:6 Changed 13 years ago by Karl-Dieter Crisman

Authors: → Karl-Dieter Crisman

See #6423 and #4142 for other bugs fixed by this spkg.

### comment:7 Changed 13 years ago by Robert Marik

Installs fine. Still running tests, but have the following (trivial) errors.

```sage -t  "devel/sage/sage/interfaces/maxima.py"
**********************************************************************
File "/opt/sage-4.3.rc0/devel/sage/sage/interfaces/maxima.py", line 1204:
sage: maxima.version()
Expected:
'5.19.1'
Got:
'5.20.1'
**********************************************************************
File "/opt/sage-4.3.rc0/devel/sage/sage/interfaces/maxima.py", line 2723:
sage: maxima_version()
Expected:
'5.19.1'
Got:
'5.20.1'
**********************************************************************
```

### comment:8 Changed 13 years ago by Karl-Dieter Crisman

If those are the only ones you get, can you put a reviewer patch of those? I always forget little things like that, and unfortunately the horsepower I have precludes running full tests if I ever want to get anything done. I did get the one in rings/number_field, though.

### Changed 13 years ago by Robert Marik

apply on the top of the previous patch and #6423 and #4142

### comment:9 Changed 13 years ago by Robert Marik

Status: needs_review → positive_review

Positive review! Tested on 4.3.rc0 and got only errors which have been reported on sage-devel and are not relevent to this ticket. Tested together with #6423 and #4142.