🚀 go-pugleaf

You brought up a good point -- there is practically no material on
how to program for an X.25 link. There is also no standard programming
inteface. 
One reason is that most applications
interface at a level higher than X.25. (Consider why noone writes applications
for an RS-232 interface.)

The X.25 which you get with SVr2 UNIX is almost useless
for general X.25 networking. 

X.25 is typically used with higher, PAD (packet assembler/disassembler) layers 
such as:
X.3, X.28, or X.29 which map TTY data streams to X.25
QLLC (qualified link level control) for mapping SNA to X.25
OSI layer 3A, which maps OSI network layer to X.25
DSP which maps 3270 BSC data to X.25, or
BPAD which maps 2780/3780 BSC to X.25

Though one could write an application to interface directly to X.25,
its interconnectability would be limited.

For a description of how we created a useful interface for the X.25
on the NCR TOWER, see my article in the Feb. '87 issue of Data Communications
magazine on UNIX and X.25.

	Robert Heath
	heath@Columbia.NCR.COM
	ncrcae!heath

In article <24642@sun.uucp> guy%gorodish@Sun.COM (Guy Harris) writes:
>	-rw-rw-r--  1 sys             0 Dec 31  1969 bx25b.c

Those aren't just 0-length files; from the date one can guess that
the inodes have been zapped.  I vaguely recall that I once had a
batch of non-null bis-X25 files in one of our AT&T distributions
(I don't maintain them on-line since we don't use them here).  I
don't know WHAT happened to the ones Guy has.  "Very interesting..."

> Those aren't just 0-length files; from the date one can guess that
> the inodes have been zapped.

Most likely, they were zapped by AT&T.  When I read in the X.25 tape
at CCI, all the X.25 source files had zero length; it seems extremely
unlikely that the exact same zapping occurred on two different
machines.

Somebody else reported to me that their S5R2 and S5R2V2 tapes both
had zero-length files for the X.25 source - but with different
modification times!
	Guy Harris
	{ihnp4, decvax, seismo, decwrl, ...}!sun!guy
	guy@sun.com

> Those aren't just 0-length files; from the date one can guess that
> the inodes have been zapped.

Not necessarily.  A more likely possibility is that the files were
truncated to zero length, and the modification times cranked back to
the epoch with something like the "touch" command, before they cut
the S5R2 tape at AT&T.

An additional interesting data point, from the S5 "Release 1" tape:

	gorodish$ cd /archbeluga/s5/usr/src/uts/vax/io
	gorodish$ ls -l *x25*
	-rw-rw-r--  1 sys             0 Nov  3  1982 x25b.c
	-rw-rw-r--  1 sys             0 Nov  3  1982 x25m.c
	-rw-rw-r--  1 sys             0 Nov  3  1982 x25r.c
	-rw-rw-r--  1 sys             0 Nov  3  1982 x25s.c
	-rw-rw-r--  1 sys             0 Nov  3  1982 x25u.c

Still zero-length, but with a non-zero mod time.
	Guy Harris
	{ihnp4, decvax, seismo, decwrl, ...}!sun!guy
	guy@sun.com