Coherent note:

As you will see from below, I didn't write this. As a matter of fact, I
appear to be at least the 3rd Person involved. Below you will find comments
by the original author and also by another person who set the current
package up.

I was able to get it running with a minimum of trouble. I used the C source,
which worked with minor changes. I've included everything, including the
original pascal source.

If you have any problems, you can email me at:  kranitz@sundown.mil.wi.us

Good Luck.....

Wes Kranitz

------------------------- Start of original Read Me --------------------



The translator is a single huge Pascal program. If you are working on a UNIX
system that has a Pascal compiler what you need to do is to unpack and
concatenate the four parts forming the translator sourcecode, "ptc.p",
unpack the "makefile", say "make ptc" and take a break.

If you are on a non-UNIX system proceed with discretion.

If compilation fails due to the program size you may consider attacking the C
version, "ptc.c", if this too is too large to compile you may as well give up.
If it fails to compile or run for other reasons you may try to contact me.

When/if the ptc program is created you may try saying "make tests", this should
not produce any output at all. If it does it means that ptc obtained from ptc.p
differed in behaviour from cptc obtained from ptc.c obtained from ptc.p. In that
case your Pascal and/or C compilers depart seriously from the ones availible to
the author, this may imply a bug in ptc.

Read the section "Tuning" in the implementation notes and do what you believe
is appropriate for your installation; recompile and install binary and man-page
as necessary.

Send vitriolic comments to /dev/null;
checks, constructive critisism and bug-reports to the address below.


Sincerely,

	Per Bergsten

Email:	perb%holtec.uucp@chalmers.csnet
	... mcvax!enea!chalmers!holtec!perb
aka:	bergsten@chalmers

Smail:	Holistic Technology AB
	Grona Gatan 59
	414 54 Gothenburg
	Sweden

Fax:	+46 31 247456



[This is a BIG program, I spend a day trying to compile it for MS-DOS using
 the Microsoft compiler but with no luck.  I think that it makes implicit
 assumptions that call for a 32-bit machine but I'm not sure yet.

 I decided to upload the source and let someone else have a crack at it.
 It looks like it would be very valuable to someone who works in both
 PASCAL and C.  If you can compile the PASCAL source that would probably
 be the best place to start.  The C source is a translation, by the
 program, of the PASCAL source.  It also serves as a test case for
 checking whether or not you successfully compiled the program.  See
 the MAKEFILE for details.

 Couple of notes on the C source.  The source file as distributed (250K)
 will choke some compilers.  The Microsoft compiler couldn't handle it
 as one file, and neither could the C compiler on a brain damaged UNIX
 machine I tried.  I was able to compile it with the Microsoft compiler 
 by breaking the source into 12 files of 800 lines or so.  If you take this
 approach be aware of a couple of things.  The code which proceeds the first
 subroutine is a mixture of typedefs, global declarations, and defines.
 The typedefs and defines need to be moved to an include file so that they
 are available in every source file.  The global declarations need
 to be included in one source file with extern references in the others.
 Also, there are routines and variable declarations near the end of
 the source which are declared 'static'.  If you split the source you
 will have to remove the static declarations or else you will get
 unresolved names at link time.  If your compiler complains when a subroutine
 is used and then later declared as something other that 'int' (Microsoft
 does) you will have to create an include file with declarations for
 the return types of the subroutines.

 In the comments at the head of the source is a note which says that
 code which is dependent on the CPU is marked as such.  The comments
 about CPU dependencies appear in the PASCAL source, but are not in
 the C source.

 The file PTOC.TXT is the nroff output of the file PTOC.DOC.  There
 are places where it looks strange, apparently my nroff version didn't
 like some of the macros used, but it is readable.



 ken brown

 in the MS-DOS world i hang out on the
 Computer Connection BBS @ (202) 547-2008

 in the UNIX world
 ...seismo!irs3!irs1!kbrown

 ]
 
