RDROYA01@ULKYVX.BITNET.UUCP (07/08/86)
Since I last wrote I've gone a step further with the [Command] buffer
idea that I wrote about. Now you are able to suspend a background
task, like DDT, return to the editor for further editing, and then go
back to the suspended program where you left off. Debugging under
this system is much easier than it was before because not only can I
display, and move around in the source code, but I can also suspend
the background program to inspect that buffer's contents, for
example, look at a trace four screens previous, and then return to
the debugger or whatever is running.
I think the code that handles this switching about could be made a
simple memory resident module that any program could call. It could
be set up like the Plu*Perfect products for CP/M-80 that "watch" for
a key typed at the console, and then do their chores. The code is
about 150 lines of assembler with only one outside reference, the line
allocation for the word processor. For a mem resident program that
reference could be replaced by a simple line output routine. The
bios conin could check each typed char for a "switch" key of some
sort, and when found could save the current environment and call the
ccp. When the switch is typed again, the background task is
suspended, and you return to the original program. A CTRL-C instead
of the switch would exit the background task and return you to
background ccp. The resident program as a separate module is about
12K. With the additional conin and conout routine, it would likely
be only around 20K, really not large at all.
Perhaps this kind of function is already common and available on most
CP/M-68K systems that allow multiple users, but on my single user
system it is quite a boon. Anyone interested in code?
rdroya01@ulkyvx.bitnet