Re: DOCBOOK-APPS: More on PDF

[Lane Stevens <lane at cycletime dot com>]

See inline...
On Thu, Nov 30, 2000 at 01:35:57PM -0700, Lane Stevens wrote:
> I am creating some graphics with StarOffice to include in a document.  I am
> able to export the graphics into eps format, and then convert to (e?)pdf
> format with the epstopdf tool.
> 
> My questions:
> 1)  If I use a statement like this:
>     <figure float="1" id="DFD0">
>       <title>Context-level dataflow diagram</title>
>       <mediaobject>
> 	<imageobject>
> 	  <imagedata format="pdf" align="Center" fileref="./images/SDGDFD0.pdf">
> 	</imageobject>
>       </mediaobject>
>     </figure>
> 
> in my sgml document, then I get the following complaint from openjade:
> 
> openjade:sdd.sgml:237:25:E: value of attribute "FORMAT" cannot be "PDF";
> must be one of "BMP", "CGM-CHAR", "CGM-BINARY", "CGM-CLEAR", "DITROFF",
> "DVI", "EPS", "EQN", "FAX", "GIF", "GIF87A", "GIF89A", "JPG", "JPEG",
> "IGES", "PCX", "PIC", "PNG", "PS", "SGML", "TBL", "TEX", "TIFF", "WMF",
> "WPG", "LINESPECIFIC"
> 
> If I leave
> the format attribute out entirely, no graphic appears, but the validation
> step goes much better.  How should my graphic be declared in the file in
> order for me to successfully use the PDF graphic?
> 
After some research, I found that if I edit the dbnotn.mod file I could
eliminate the validation error.  I added PDF to notation.class, and then
added a line among the other notation declarations:
<!NOTATION PDF SYSTEM "PDF">

My DTD is docbook 4.1, and for some reason it doesn't feel right having to
make a change to the DTD to make this work.  Is there a more appropriate way
to have solved this problem? or am I completely missing the boat somewhere?

> 2) In my source (e.g., the StarOffice file), I have sized the graphic to be
> the size that I need for my target document.  I suspect that there are
> better ways to scale and specify the target size.  Are there?
> 
> 3) It is my intention to export the graphic in GIF format to use in the
> production of HTML documents, and then use multiple imageobject statements
> to select based on the type of target document that is to be produced.  Is
> this an acceptable approach or is there a better way?  By way of background,
> I have set the preferred-media-object-notations in the respective driver
> files to cause the selection to take place, and I use a separate driver for
> HTML and print (pdf) output and am willing to stick with this approach
> for now.
> 
> I would appreciate advice and pointers to information from anyone that can
> offer some help.
> 
> Thanks,
> Lane.
> -- 
> Lane Stevens
> Terrapin Technologies, Inc.
> http://www.cycletime.com

-- 
Lane Stevens
Terrapin Technologies, Inc.
http://www.cycletime.com