I  have now  converted  three  docs to  SCRIPT  format:  LISTFILE &  LISTAFD
(grouped together  in a 'File  Server Functions' manual), and  LISTLPUN (which
had no reason to be grouped with anything else). I can send you LIST3820 files
if you'd like to try them out, but  that's not the main point. I have also had
to develop a set  of SCRIPT macros to allow me to reformat  the files in 'raw'
format and make them look as they did before... which was not easy at all!
 
  I have  replaced the  old version  of LISTFILE  MEMO on  my server  with the
output of my  SCRIPT macros for the corresponding doc.  Note that LISTFILE and
LISTAFD have  been grouped  in a  single manual but  they are  still formatted
separately when in  'raw' mode so as to avoid  having overtly large helpfiles.
You can send an INFO  FILE command to my server if you want  to have a look at
it. I  get the impression  that the  output is too  large and the  preface and
appendixes could be zapped, at least from  some of the docs, while the general
introduction docs might need  to keep the preface. On the  other hand it might
be helpful  to keep the list  of publications at the  end of the doc  since it
would, in the long run, reduce the  number of INFO ? commands sent to LISTSERV
so it might end up actually reducing network load. I just don't know and would
like your opinion. :-)
 
  I have  not had  time to  investigate the  SCRIPT/WSCRIPT problem  yet. Most
important was for me to write the set of macros to regenerate 'raw' files, and
whether they can run  under WSCRIPT or not is irrelevant  -- what is important
is that ONE node  in the network (preferrably mine ;-) ) is  able to run them.
But from what  I've seen it would  seem VERY difficult to write  a SCRIPT file
that would  run under  both processors  with changes only  in the  profile (eg
definition of  a .AA macro  if under WSCRIPT).  Example: IBM's title  page tag
does not have a logo. The logo is supposed to be on the hardcover which is not
SCRIPTed. I don't have  an offset press and I wanted a logo,  so I rewrote the
title page tag. That's no problem,  *EXCEPT* that I must still re-invoke IBM's
title page tag to have it issue  the appropriate .SE commands for its internal
variables (otherwise  the next GML tag  will complaint about not  being in the
title  page). I  do  that by  setting the  appropriate  variables and  calling
DSMETTL. I  bet $10 that  Waterloo's is not  called DSMETTL... and  there will
probably be a lot of problems like that :-( Well, we'll see that later.
 
  Eric