🚀 go-pugleaf

I received responses from Mark B. Tooley, Matt O'Kelley, Doug Gibbons,
Mathew Atkins and one that got lost.  Most were enthusiastic about the
Wyse-60, and M. Akins sent an entry for the 30.

We decided to go with the 60 emulating a vt100.  Why?

1) both termcap entries for the 30 gave me problems with the (Gosling)
   emacs mode line update.  It was just plain wrong.  Also, the 30 has only
   24 lines.

2) The 60 emulates a vt100 just fine, and can in fact be persuaded to
   emulate a 43 * 80 vt100 just fine.  Also, when emulating a vt100 it
   works more sensibly with the ANSI keyboard.
   I realize that it can be configured to do almost anything, but it is so
   much easier if you can just pop it out of the box and use default
   settings.

3) When running in Wyse-60 mode the terminal for some reason decided that
   the data coming in was 7 data + 1 stop bit, when in fact I think that it
   is 8 + 1.  It processes 8+1 correctly in vt100 mode.

Notes on vt100 emulation.

When running in 42/43 line mode, it is important to *turn off* character
attributes instead of *turn to normal* character attributes.  "Return to
normal" also returns the character size to normal, though the line spacing
stays the same (yuck).  Here is the modified termcap entry:

d0l:\
	:is=200\E[41h\E[1;42r\E[42;1H:\
	:se=2\E[27m:\
	:ue=2\E[24m:\
	:me=2\E[22;25;27m\
	:li#42:tc=vt100:

The "is" attribute places the terminal into 42 line mode, defines a 42
line scrolling region and places the cursor at 1,42.
The new strings for "se", "ue", and "me" all specifically *turn off* their
respective attributes, leaving the characters at the same size.  Using the
sequence "\E[m" to revert to normal loses pretty badly.

I may be using an old termcap file, so this may be old news to other people
out there.  Do vanilla vt100s recognize these sequences, and if so, should
termcap files be updated to reflect this?  This is probably just a quirk of
the Wyse-60.

David Chase, Olivetti Research Center, Palo Alto