freebsd-dev/usr.bin/tn3270/distribution
Bruce Evans 89e45c9b7f Don't use gets().
sys_curses/system.c:
Don't use gets() better.  Neither gets() nor fgets() is appropriate for
discarding a line of input.
1995-09-29 18:44:53 +00:00
..
arpa
sys_dos Don't use gets(). 1995-09-29 18:44:53 +00:00
telnet
utilities Remove trailing whitespace. 1995-05-30 06:41:30 +00:00
makefile_4.2 Change all references to LIBTERM and -ltermlib to LIBTERMCAP and -ltermcap 1994-09-11 22:01:45 +00:00
README
ultrix.curses

Welcome to the new tn3270 - version 4.1.1.

This version consists entirely of bug fixes to the August 1987 beta release
of tn3270.  This version will now deal with CICS and VM/XA on the IBM
side, and with SunOS 4.0 on Sun 3's and Sun 4's.

This version has been tested on various versions of BSD Unix, including
4.2 and 4.3 (but there are never vanilla versions) and the post-4.3 systems
running at Berkeley.  It has been tested on CCI's, Vaxen, Sun 3's and Sun 4's.
However, it doesn't necessarily work on all systems (nor has the testing
been as intensive as one might like).

This version should build on any Berkeley-derived system.

There are two alternate make files:  ./makefile_4.2 and telnet/Makefile_ultrix.

****    Try ./makefile_4.2 if you get compile-time errors, or get
	"multiply defined" messages for "_putchar" from the loader.

****    Try telnet/Makefile_ultrix if your make(1) utility doesn't
	support VPATH.  Also try this makefile if your ld(1) doesn't
	support the -r flag correctly.

The bad news is that I've had to drop MS-DOS support.  The good news here is
that there are various versions available for MS-DOS (from FTP Software in
Cambridge, Mass.; from IBM; from Excelan; and probably from others).  The
hooks are still there, as well as some code to update the screen.  However,
I just haven't had the time to produce a fully integrated version that would
"just make".  I suspect that a future release may have MS-DOS support back
in it.

There is no Mac support.  Contact Peter DiCamillo at Brown University if
you need a Mac tn3270.

The main code change in this version is to what used to be called "telnet.c".
This is now replaced with a version of telnet (substantially what appeared
in the "4.3tahoe" release from CSRG) which is broken into separate files.

Here is an overview of the directory structure:

    api/		General library of function needed by API
			(and, to some extent, by the rest of tn3270).

    arpa/		Location of "telnet.h" (for non-4.3 systems).

    ascii/		Routines necessary to handle the case of running
			from an ASCII-oriented system (ie: unix).

    ctlr/		The main part of the emulator.  Handles 3270 scan
			codes, 3270 data stream, 3270 display codes,
			and EBCDIC.  Also, the internal API function
			lives here.

    general/		Some general subroutines and data structures of
			interest to the emulator only.

    sys_curses/		System-dependent code for a curses-based environment.

    sys_dos/		System-dependent code for an MS-DOS-base environment.
			Remember that this is included for your developmental
			needs (ie: it doesn't work).

    telnet/		Where the telnet portion of tn3720 is built.

    tools/		Various tools.  Most of these are used during the
			build process.  One (prt3270) is a debugging tool.
			One (mkmake.y) is quite horrible, and attempts to
			transform Unix makefiles into PC makefiles.

    utilities/		The source for tnrecv, which receives files
			(fairly slowly) from an IBM host.  We don't
			include the IBM side, because we really aren't
			happy with very much of it (except that it does,
			sometimes, work).  Hopefully, when we get past
			the beta stage we will have more robust (and
			complete) code to share.

The fact that system dependancies are isolated should make it easy
to port to other systems.  I would like to hear about problems porting
to new areas.

In the August, 1987 README file, the following appeared:

> WHAT IS NOT IN THIS VERSION (sigh):

> 1)	We don't have a native X version yet.  I am waiting for X version 11
> 	(though this is mostly an excuse; I could have done version 10,
> 	but I haven't had the time).

> 2)	We don't process structured fields.

> 3)	We don't do 3270-style graphics (ala 3193, say).

> The above three items WILL be in the next version, which should come
> along "any day now" (say 6 months) (but, they WON'T be in the production
> release of this version).

The next piece of bad news is that none of the above have happened yet,
and I don't know when they might occur.