Final Word: PT3 / CTS bug fix

[badbunny@tfsp.saic.com (Brendan Hoar)]

Now that I've downed a veggie burger, some cajun beans & rice (vegan) and a
wine cooler, I feel better about posting this.  :)

Morgan (and Dave/Chris), you were absolutely right in insisting that it was a
software problem.  Greg Schaefer was intializing the SCC with $56 (from the
example in the Zilog docs - Zilog assumes that you aren't going to do something
silly like tie TRxCB and CTSB together...) instead of $50 (from the GS
docs...or technotes?).  So Greg was not only enabling everything he needed, but
also enabling the transmit clock as well - and as we all know, having an input
and an output enabled on the same line is bad.  Because I was unable to get a
fix/waveform on the so-called interference/noise I detected (the reason is now
obvious - my ineptitude with an oscilloscope) Greg assumed that it wasn't the
transmit clock that was interfering and that it was some other problem.  Oops.
:(  I screwed that one up - definitely.

One sort of good thing that came out of all of this is that while Greg was
working on ways to get around the bad reads of the CTS signal, he ran out of
room in the code space he had allotted for the driver.  So he had to do a
little optimization to make room.  So there are a coupla cycles removed here
and there in some loops and a few bytes saved.  Even though he has now taken
out all of the work around code, he has kept the optimizations in.  Cool.

[Note: Due to the different tolerances of the Motorola 26LS32s and other
26LS32s, the transmit clock only caused problems with GS's that had Motorola
buffer chips (some TIs).  So there is a hardware DIFFERENCE, but not a hardware
PROBLEM.  I've got it straight now.  It just happened by chance that of the GS
[I] have available to me, at least half of them had Motorola buffer chips.  My
guess is that this is not representative of the 'general GS population',
otherwise this would have come up long ago.]

Turning off the transmit clock, however, deals with it a lot better than
soldering on an extra pull up resistor or two.  And speaking of that...

Apologies to David Huang who soldered on a resistor to get around the problem.
However, he says that the 'fix' works great for him.  And while I'm at it,
apologies to my auxiliary GS which is also modified.  Poor thing.  :(

*sigh* And to think I almost had a scoop!  :)

To everybody:

Greg has posted (yet again) a new PT3.CODE0 on the InSync BBS.  Grab it!  As
usual, it has some more even cool improvements.  There is yet again, a better
VT-100 plug in emulation too (PTM.V1).  Greg has been working hard at both
writing the Macintosh version of ProTERM *AND* supporting the Apple II version.
Do him and InSync a favor and if you use the program, please buy it.  Thanks.

Oh yeah, sorry - no cute ASCII art this time.  :)  Just smileys.

____________                                        | Arbitrary in our pity.
Loc.|Address|We_are_in_our_minds, 5 billion PIECES  | Selective in our shame.
Work|Brendan_Hoar@notes.pw.com|_______so_defined__  | Recognizing our lives as
Play|badbunny@tfsp.saic.com        (703) 998-5687 | | pinpricks, we conclude
IPay|BrendanHr@aol.com_(ex-GS_Hardware_Terrorist)_| | exemption from blame.