home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Usenet 1994 October
/
usenetsourcesnewsgroupsinfomagicoctober1994disk2.iso
/
unix
/
volume16
/
pcomm2
/
part02
/
Release.notes
< prev
next >
Wrap
Text File
|
1988-09-14
|
5KB
|
157 lines
Pcomm version 1.1
15 Aug 88
Incompatibilities with v1.0:
I've added a few more fields to the pcomm.param and pcomm.modem
files. This means that those files created under previous
releases will not be compatible with the v1.1 release. It might
be a good idea to abandon the older files and start over with
the sample files in this release.
Things that have changed:
An awful lot of sloppy programming on my part has now been
cleaned up. This means that porting Pcomm to different versions
of Unix will be considerably less painful than before.
All of the tunable parameters are now in "config.h".
Pcomm now uses buffered I/O routines. This yields higher
performance I/O at large baud rates.
File transfers now use high performance CRC calculations.
If a UUCP lockfile does exist, Pcomm now attempts to determine
if the process is still active.
An external file transfer program can be invoked from the file
transfer menu.
The keyboard macro feature is now supported (although it is
rather crude).
The virtual screen routine now understands a limited subset of
escape sequences.
The virtual screen can now be held in shared memory or in a file.
The input routine is now contained in a separate program called
pcomm_input.
You can now include a "modem break" in a dial or initialization
string.
A separate "initialization speed" feature was added to support
9600 baud modems such as the Telebit Trailblazer. See Appendix
C of the Doc file for more info on the use of this feature.
I've added a transparent pass through mode to be used in a
communications daisy chain.
Known limitations:
Whenever Pcomm makes the transition from the terminal mode to
the command mode, the true screen contents are lost. Pcomm
tries to keep track of what the true screen looks like in an
internal "virtual screen". This means that the escape sequences
known to the hardware have to be emulated in the virtual screen.
Unfortunately, the number of escape sequences that are emulated
is quite small, so the representation of the true screen is
sometimes wrong.
For the VCS (video command sequence) emulation to work, the
terminfo entries on the host machine must match the entries on
the remote.
The disp_tty() routine in s_tty.c currently does not support a
NUM_TTY definition of greater than 10.
Future directions:
Continue work on the virtual screen routines.
Add dial back support.
Use shared memory to overcome the current limitation of outgoing
characters in the half duplex mode not appearing in the virtual
screen, log files, and print logs.
Many thanks to the following people for bug reports and ideas for
improving the code:
Andy Pitts ...{mtune,pacbell,kd4nc}!gladys!rbdc!andy
Chris Wiener ...ihnp4!{killer,attnssc}!crlabs!cwiener
David Brierley ...{sun,decuac,cbosgd,gatech,necntc,ukma}!rayssd!dhb
Karl Fox ...cbosgd!mstar!karl
Michael Young ...panda!genrad!mrst!sdti!mjy
Mark Mendel ...ihnp4!umn-cs!hyper!mark
Viet Hoang ...ihnp4!drutx!vgh
--------------------------------------------------------------------------------
Pcomm version 1.0
12 Mar 88
Incompatibilities with the beta release:
I've added a bunch of new parameters to the pcomm.param and
pcomm.modem files, so the files used with the beta release won't
work with the v1.0 release.
Things that have changed:
Pcomm will attempt to determine if the modem has synchronized at
a baud rate different than what is expected, and make changes to
the line settings as appropriate.
The quit and interrupt signals are now ignored.
All of the file transfer protocols are now functional.
The directory search order used to find the support files has
been changed slightly.
Known limitations:
The keyboard macro feature is not implemented at this time.
The true screen contents are lost when the "hot key" is pressed.
Pcomm attempts to compensate (rather poorly) by repainting a
virtual screen of what it thinks the true screen should look
like. Escape sequences in the virtual screen image will be
ignored when the background is repainted. For example, if
you're on a vt100 and you recieve a ^[[2J to clear the screen,
the screen *will* be cleared... but when the screen is
repainted, it will contain the characters ^[, [, 2, J (instead
of performing the function).
The disp_tty() routine in s_tty.c currently does not support a
NUM_TTY definition of greater than 10.
For some strange reason, the first keystroke is "lost" after a
file transfer is complete or after starting data logging.
Future directions:
The virtual screen routines need a lot of work. The most
commonly used escape sequences (known to terminfo) will be
processed.
I plan to have an option at compile time to have the virtual
screen buffer held on disk [] or in shared memory [].
The input routine is designed so it could be a standalone
program that gets called from Pcomm. []
-------------------------------------------------------------------------------
Have fun...
Emmet P. Gray US Army, HQ III Corps & Fort Hood
...!uunet!uiucuxc!fthood!egray Attn: AFZF-DE-ENV
Directorate of Engineering & Housing
Environmental Management Office
Fort Hood, TX 76544-5057