R help - Aug 2003 - Encapsulated postscript and the family argument

> version
         _                
platform i686-pc-linux-gnu
arch     i686             
os       linux-gnu        
system   i686, linux-gnu  
status                    
major    1                
minor    7.1              
year     2003             
month    06               
day      16               
language R                

>From my investigations, it would appear that if I wish to make an eps
file instead of a 'normal' postscript file, I can specify paper as
'special' and onefile to FALSE which will then mean the bounding box
is used as the size.

However, what wasn't obvious to me was that it is necessary to specify
what family to use.  If no family is specified, the default family
does appear to be used, BUT, the resulting file is no different from a
'regular postscript' file.  The value in ps.options does not seem to
be used in the same way.

Is this intentional behaviour?

The help file refers to the existence of two forms of the argument
family, but it doesn't seem to me to be what is at stake here.  There
is a reference earlier in the file to the need to be a postscript
expert if tinkering with .ps.prologue is contemplated.  Maybe
something similar is required to understand what's happening here.

best

-- 
Patrick Connolly
HortResearch
Mt Albert
Auckland
New Zealand 
Ph: +64-9 815 4200 x 7188
~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~
I have the world`s largest collection of seashells. I keep it on all
the beaches of the world ... Perhaps you`ve seen it.  ---Steven Wright 
~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~.~

On Mon, 25 Aug 2003, Patrick Connolly wrote:
> > version
>          _                
> platform i686-pc-linux-gnu
> arch     i686             
> os       linux-gnu        
> system   i686, linux-gnu  
> status                    
> major    1                
> minor    7.1              
> year     2003             
> month    06               
> day      16               
> language R                
> 
> 
> >From my investigations, it would appear that if I wish to make an eps
> file instead of a 'normal' postscript file, I can specify paper as
> 'special' and onefile to FALSE which will then mean the bounding
box
> is used as the size.
> 
> However, what wasn't obvious to me was that it is necessary to specify
> what family to use.  If no family is specified, the default family
> does appear to be used, BUT, the resulting file is no different from a
> 'regular postscript' file.  The value in ps.options does not seem
to
> be used in the same way.
The family used is nothing to do with EPS.  The code is always
EPS-conformant (but may not be a single page), but the *header* is only
sometimes, the times being documented.
> Is this intentional behaviour?
Is what, exactly?  Consider

postscript(width=8, height=6, horizontal=FALSE)

postscript(file="Rplots.eps", width=8, height=6, horizontal = FALSE, 
onefile = FALSE, paper = "special")

The first does not give an EPS header: the second does.  Try it and see:  
`family' has nothing to do with it.  But for the record, if family is set 
in ps.options and not in the postscript call it is used.  Try
ps.options(family="Times") and repeat those calls.

A long-timer such as yourself really, really should know not to send in
vague statements not backed up by the code used to leap to these 
conclusions!

BDR

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595