graphic device MetaPost

Sun, Mar 7, 2004 8:41 AM #

Users may have insights that the developers may not have and 
those ideas may usefully find their way into R, even if the 
promoter of the idea does not have the capabilities or 
resources to implement it.

Surely, good ideas such as this one should be mentioned.  
Maybe that user will implement it or maybe someone from 
the R core will or maybe some other user will.  Who knows?  
No one will if no one knows about it.

Date:   Sun, 7 Mar 2004 14:42:17 +0000 (GMT) 
From:   Prof Brian Ripley <ripley at stats.ox.ac.uk>
To:   =?gb2312?q?Jinsong=20Zhao?= <zh_jinsong at yahoo.com.cn> 
Cc:   rhelp <r-help at stat.math.ethz.ch> 
Subject:   Re: [R] graphic device MetaPost

On Sun, 7 Mar 2004, [gb2312] Jinsong Zhao wrote:

I don't think so: you would still need character metrics for the fonts you 
use to be able to centre them, for example.

We would welcome your contributing such a device. Note though that 
there is a public API for graphics devices, and so you could just 
contribute the device to CRAN.

If perchance you meant `I want the R core team to write a metapost device
for me', then you have not grasped how Open Source projects work.

There are some plans for internationalization of R via UTF-8, but this
will be a considerable amount (man months?) of work, and of very little
benefit to any of the core developers. Volunteers would be welcome.

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

Itay Furman

Sun, Mar 7, 2004 11:12 AM #

The TeX world offers several comapnion systems to typeset 
graphics; metapost, xy-pic, and PSTricks are the most powerful, 
and have facilities for typesetting data plots. Recently, the pgf 
LaTeX package, that can coop with PDF and PS from the same 
source, was submitted to CTAN; but it is more restricted than the 
others.

mentioned device. The trouble is, that I feel it is beyond my R 
and TeX skills (for the moment, at least).
If no one picks up this glove -- then I will attempt this 
venture.
If you would like to see such a device(s), and would like to 
contribute, maybe we can team up.

To the original poster: could you use the pictex() device
for your needs?

	Thanks,
	Itay Furman

On Sun, 7 Mar 2004, Gabor Grothendieck wrote:

Users may have insights that the developers may not have and 
those ideas may usefully find their way into R, even if the 
promoter of the idea does not have the capabilities or 
resources to implement it.

Surely, good ideas such as this one should be mentioned.  
Maybe that user will implement it or maybe someone from 
the R core will or maybe some other user will.  Who knows?  
No one will if no one knows about it.

Date:   Sun, 7 Mar 2004 14:42:17 +0000 (GMT) 
From:   Prof Brian Ripley <ripley at stats.ox.ac.uk>
To:   =?gb2312?q?Jinsong=20Zhao?= <zh_jinsong at yahoo.com.cn> 
Cc:   rhelp <r-help at stat.math.ethz.ch> 
Subject:   Re: [R] graphic device MetaPost 

 
On Sun, 7 Mar 2004, [gb2312] Jinsong Zhao wrote:

By default, MetaPost passes all text through TeX. This has the
advantage of allowing essentially any TeX symbols in titles and labels.
It give us, who use the multibyte character in ordinary communication,
much convenience. Gnuplot has fulfilled this function, and it give me a
deep impression for I could use Chinese character in plots with a minor
modification to the MetaPost file.

I don't think so: you would still need character metrics for the fonts you 
use to be able to centre them, for example.

I hope the R Development Core Team could consider MetaPost as a graphic
device in future R version.

We would welcome your contributing such a device. Note though that 
there is a public API for graphics devices, and so you could just 
contribute the device to CRAN.

If perchance you meant `I want the R core team to write a metapost device
for me', then you have not grasped how Open Source projects work.

There are some plans for internationalization of R via UTF-8, but this
will be a considerable amount (man months?) of work, and of very little
benefit to any of the core developers. Volunteers would be welcome.