home *** CD-ROM | disk | FTP | other *** search
- PMMIBYE5.DOC (as of 7/20/80)
-
- The PMMIBYE5.ASM program is out for distribution,
- but it has not been thoroughly tested in all of the
- possible modes of operation.
- (Please note that version 5.4 is substantially longer
- than previous PMMIBYE versions. Because of the great
- number of conditional switches, the program could require
- as many as 1060 bytes of RAM, or it could be smaller than
- the 900 bytes used by the original PMMIBYE. It all
- depends on the way you set it up.)
-
- Basically, several new conditional assembly routines
- have been added to allow user selection at time of
- assembly of the following:
- 1. Password access
- 2. User log (counts number of system callers)
- 3. Memory-mapped console I/O
- Also added to V5.4:
- 1. 710 Baud option for PMMI modem
- 2. Modifications to AUTO-ANSWER routines
- 3. "/R" option to allow reset of user log counters
-
- Password access was made conditional to allow
- use with systems that offer unrestricted access.
- The user log was added to allow the system operators
- to keep track of the number of people who called the
- system. In addition, it can also count the number of
- unsuccessful log-ons (i.e. didn't know password) and
- the number of voice calls, and conditionally
- display both numbers on the system console when the
- program is exited with a control-C from the system
- console. The counters can be initialized with
- the "/R" option to all zeroes, or preset by using
- DDT, or SID. The number of successful log-ons is also
- conditionally output to port 255 (0FFH) for display on the
- computer's front panel.
- The memory-mapped I/O conditionals have been added
- to allow use of the serial I/O port on the Morrow's 2D
- controller board as the system console; however, the
- routines can be used with any memory-mapped system.
- The 710 Baud option was added for "short-haul"
- callers, and because operation at 710 Baud with
- the PMMI modem is "iffy" at best, it will probably
- be of very little use.
- The AUTO-ANSWER routines were modified to allow
- more dependable ringback operation. Unfortuneately,
- the ringing signal at the caller's end of the phone line
- does not always correspond exactly with the ringing
- signal at the system's end. Although a system-end ring
- does occur sometime during the caller's ring cycle (ring-
- no ring), it is possible for the system to receive no
- or 2 rings in the time that the caller thinks just
- one ring has occurred (The DIAL program is notorious for
- this). The solution chosen for this problem was to
- check for one or two rings before call-back. This
- would allow for the "phase" difference of the caller-
- system ring signals. This solution also means that
- the phone must ring at least three times before the
- system will assume that the call is voice, and is not
- to be answered. The advantage of this arrangement is
- that it is virtually impossible for the computer to
- "miss" a call that is intended for it.
-