Re: modified residues() for matlab compatibility

[John W. Eaton]

On  5-Oct-2007, Ben Abbott wrote:

| On Oct 5, 2007, at 5:42 PM, John W. Eaton wrote:
| 
| > On  5-Oct-2007, Ben Abbott wrote:
| >
| > | For the ChangeLog are these recorded in the script, or in a separate
| > | file to permit you to record the change in you version control  
| > system?
| >
| > Logs of changes for Octave sources belong in the ChangeLog files kept
| > along with the sources.  We don't use CVS logs and I prefer to avoid
| > cluttering the sources with this kind of info.
| >
| > jwe
| 
| ok ... I'm <blushing/embarrassed> after looking into through the  
| source distribution ;-)
| 
| I've attached ChangeLog.txt respecting the changes in residue.m and  
| the addition of mpoles.m. As I am new to this, don't hesitate to  
| offer corrections.
| 
| I've made an attempt to properly construct the copyright and texinfo  
| portions to both residue.m and mpoles.m ... they are included as well.

OK, I checked in these functions.

ChangeLog entries should just say what changed.  There is no need to
go into detail about the reason for changes there.  If extra
explanation is needed, it belongs in a comment in the code.

The initial part of the doc string for residue could use some
improvement.  The initial sentence of a doc string should always be
written as a command, like this:

  Compute the singular value decomposition of @var{a}.

not

  This function returns @var{s}, the SVD of the argument @var{a}.

or

  Given a matrix @var{a}, this function returns its singular value
  decomposition in @var{s}.

etc.

The @seealso lines should just be a comma-separated list (no "and"
separating the last two elements).

There is no need for @code{} in an @example environment.

I don't think any other docstrings have index entries in them.  Should
they?  I'm not sure.

We could also use some tests.  Maybe at least a few simple sanity
checks plus the example (bug?) that prompted these changes?

Thanks,

jwe