[OS X TeX] ps2eps as of 2009 was out of date

[Michael Sharpe]

On Feb 11, 2011, at 6:08 AM, Arthur Snoke wrote:

> I describe below a problem I had and its solution.  It had to do with program ps2eps not producing an optimal BoundingBox.  This may not be an issue for its use within LaTeX-land, or the current package may have a newer version, but this report might be of interest to some on this list.
> 
> This is the version of ps2eps that came with my 2009 distribution:
> 
> Jul 31  2009 /usr/texbin/ps2eps@ -> ../../texmf/scripts/ps2eps/ps2eps.pl
> 
> I used it with a .ps file produced by another program and found the BoundingBox too tight (clipped on the bottom).  I could not find a MAN page, so looked on the Web.  I found this version is version 1.61, and the most recent version on the Web (August 2010) is 1.68.
> 
> To solve my problem, I could use either the -g switch (which forces a Ghostscript BoundingBox) or a -l (loose) option -- which does not seem to exist in 1.61.  These could be automated with an alias such as
> alias ps2eps 'ps2eps -l -q' or, (for -g) an environmental variable.
> 
> Maybe there are MAN pages but in the TeX package and I just don't have the path, or one can use the --help switch (which I had forgotten).

Version 1.64 of ps2eps that comes with MacTeX 2010 is still seriously flawed. Version 1.68 from http://www.tm.uka.de/~bless/ps2eps repairs all the problems, as far as I can see. In my experience, the optimal way to use it is with $hiresprecision=0.1 and the command line

ps2eps -l -f -g -r=3600 filename.ps

making use of the output from the HiResBoundingBox. The -a option is sometimes a little more accurate, but is slower. The older version had problems chiefly with bbox, which had numerous bugs. In fact, I can see no real reason to use bbox, as the gs computation of the bounding box seems very reliable now. 

Michael