[OS X TeX] fragmaster, dvips and pdflatex problem

[Ross Moore]

Hi Bruno,

On 01/07/2006, at 10:41 PM, Bruno Voisin wrote:

> Le 26 juin 06 à 12:56, Michael Williams a écrit :
>
>> On Mon, Jun 26, 2006 at 07:33:02AM +1000, Ross Moore wrote:
>>
>>>  c.  psfrag  is essentially a dead-end, so far as
>>>     future development is concerned, since it requires
>>>     programming features of the PostScript language.
>>>     As such, there cannot be an easy PDF equivalent.

>> Right now this is not a problem, but it may become a showstopper  
>> for me
>> in the future, so it would be good to migrate to something other than
>> psfrag (to which I have no particular attachment).
>
> Publishers generally require figure files to work as stand-alone  
> files,

That depends on the kind of publication that you are providing
material for ...

> not requiring interaction with any external software (here TeX and  
> psfrag) or file (here the .tex input file)

  ... and the kind of material being provided, which then determines
what kind of editing still needs to be done by the publisher
(via a pre-Press house, say).


In my experience, with several monographs and Proceedings volumes,
the modern equivalent of "camera-ready copy" is a completely
self-contained PDF. You certainly don't want the images separately
in this case.

However there may be issues concerning font-embedding and
the color-space, within embedded graphics. But these are issues
that may bite you anyway, whatever methods you use.


> to be ready for public consumption (read here: contain full  
> annotations and lettering). Which is another showstopper for psfrag.

I don't see that this follows at all.

It's easy to put together a simple LaTeX document that
includes just the psfrag-annotated images, each on a page
by itself, just as you can with WaRMreader-annotated
graphics. Then convert the .ps to .pdf in the usual way.

Indeed, when done this way, you can use  pdfcrop
to crop each page to the ideal size for its contents.

Finally, use  pdfpages to insert these annotated graphics
at the end of a new PDF containing the full paper + graphics.


Maybe I just don't understand what is your point here.

>
> Plus: in my experience, tags used in psfrag should generally be one- 
> letter long, not more. Depending on the exact DVI processor that is  
> used (for example dvips or Textures), the letters composing the tag  
> may be separated in the PS output (by spaces or PostScript  
> instructions). In that case, psfrag does no longer see the tag and  
> fails.

Do you have examples of this effect ?
Are you saying that the TeX processing can break up
longer words that are intended as tags? With which
version of psfrag macros can this happen?

The last time I examined a .ps file that used  psfrag ,
the labels were all inside a PostScript array of strings.
There's no possibility of such an effect there.
Perhaps what you say was possible with earlier versions.

>
> In short: I recommend against psfrag.

I prefer other techniques too, but not for the above reasons.

To my way of thinking,  psfrag  doesn't give you the proper
control over the size and placement of labels, while retaining
full flexibility over scaling of the graphic itself, to suit
the context in which it will appear.

With psfrag  if you need to scale the graphic, then you will
scale the labels as well, so that the fonts are no longer
at the same size(s) as what is used in the surrounding material.
To me, that's poor aesthetics.

The only way to fix it is to rescale the picture itself,
before saving as .eps .  Then the positioning of the label
may not look so gaod anymore, so it's back to the image for
more editing --- this time the position of the label(s).
To me, this is a poor model for the work-flow.


Note that pre-typesetting the labels then inserting them
within the graphic itself, is also prone to these issues.
(Think LaTeXiT and the old Textures+Illustrator way.)
If you are very careful about the exact sizing of your
graphics, upon creation, then this may not be important.
Frequently you may not have such control.



With WaRMreader, you construct the (unlabelled) graphics once,
then add the labels later, using LaTeX.  This way you can make
all final aesthetic adjustments from within the LaTeX source,
employing the Preview to judge how well it all looks. (Best
to work with 1 image at a time, until you are fully satisfied;
then put it into the full document.)

This way you don't have to revisit the graphic-producing
software, unless you are unhappy with some of the non-label
elements, of course. (Think about the cases where you have
paid a Graphic Artist for their work, or a Matlab/Mathematica
graphic where you no longer have the source to reproduce it.)

Furthermore, the same graphic can be re-used in different
ways, with and w/o labels, without need for extra editing.
Also, the graphics are not constrained to being in .eps format.


>
> Bruno Voisin


Sorry for the length of this posting.
Hope it helps,

	Ross

------------------------------------------------------------------------
Ross Moore                                         ross at maths.mq.edu.au
Mathematics Department                             office: E7A-419
Macquarie University                               tel: +61 +2 9850 8955
Sydney, Australia  2109                            fax: +61 +2 9850 8114
------------------------------------------------------------------------


------------------------- Info --------------------------
Mac-TeX Website: http://www.esm.psu.edu/mac-tex/
          & FAQ: http://latex.yauh.de/faq/
TeX FAQ: http://www.tex.ac.uk/faq
List Archive: http://tug.org/pipermail/macostex-archives/