21 Feb
2009
21 Feb
'09
5:09 p.m.
The problem with the V20's 8080 emulation was that
it was buggy (I
have a few MicroNotes on the subject and can even claim to have
discovered one of the bugs).
I wonder if it was only cetain chips. I ran cp/m 80 on a v20 for years.
22 Feb
22 Feb
12:42 p.m.
On 21 Feb 2009 at 15:09, Les Hildenbrandt wrote:
Try running a program in CP/M with JRT Pascal 1.0 on a V20. CPU bugs
can be very subtle.
Cheers,
Chuck
The problem
with the V20's 8080 emulation was that it was buggy (I
have a few MicroNotes on the subject and can even claim to have
discovered one of the bugs).
I wonder if it was only cetain chips. I ran cp/m 80 on a v20 for years. 
3:35 p.m.
On Sunday 22 February 2009 12:42:21 pm Chuck Guzis wrote:
On 21 Feb 2009 at 15:09, Les Hildenbrandt wrote:
Try running a program in CP/M with JRT Pascal 1.0 on a V20. CPU bugs
can be very subtle.
Isn't JRT Pascal the one that had really horrible reviews in the magazines
back when? Pournelle, maybe, though I couldn't say for sure as it's been a
while.
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
The
problem with the V20's 8080 emulation was that it was buggy (I
have a few MicroNotes on the subject and can even claim to have
discovered one of the bugs).
I wonder if it was only cetain chips. I ran cp/m 80 on a v20 for years.
5:57 p.m.
On 22 Feb 2009 at 15:35, Roy J. Tellason wrote:
Isn't JRT Pascal the one that had really horrible
reviews in the magazines
back when? Pournelle, maybe, though I couldn't say for sure as it's been a
while.
It *was* pretty buggy and very S-L-O-W, particularly when run from 8"
SSSD floppy on a 2MHz 8080. But it allowed for very large programs
by using some sort of program segmentation (swapped to/from disk) and
had a curious calling sequence, which turned up the V20 snake.
But it *was* cheap--something like $24.95. I still have my original
floppy from them.
I have no idea what became of JRT--ISTR, the company name was the
initials of the author's name.
It's a good illustration of how sneaky some CPU bugs can be. I
suspect that CP/M will run just fine on an original NEC 8080, for
example.
Cheers,
Chuck
6:06 p.m.
On Sunday 22 February 2009 05:57:46 pm Chuck Guzis wrote:
On 22 Feb 2009 at 15:35, Roy J. Tellason wrote:
I do recall seeing it in an ad not too long ago and thinking so too. Cheap is
good. Mostly. :-)
Isn't JRT Pascal the one that had really
horrible reviews in the
magazines back when? Pournelle, maybe, though I couldn't say for sure
as it's been a while.
It *was* pretty buggy and very S-L-O-W, particularly when run from 8"
SSSD floppy on a 2MHz 8080. But it allowed for very large programs
by using some sort of program segmentation (swapped to/from disk) and
had a curious calling sequence, which turned up the V20 snake.
But it *was* cheap--something like $24.95. I still have my original
floppy from them. I have no idea what became of JRT--ISTR, the company
name was the
initials of the author's name.
Borland came along? Not that I was ever a fan of Pascal, though. I remember
going to some meeting of a bunch of computer nuts way back in 1978 sometime,
in Philly, and a guy there was all fired up about it, I guess because there
wasn't much in the way of higher-level languages available around that time.
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
6:19 p.m.
Roy J. Tellason wrote:
On Sunday 22 February 2009 12:42:21 pm Chuck Guzis
wrote:
JRT Pascal went through several versions, and was last
seen being sold as Nevada Pascal. There is a Wikipedia
article on it, http://en.wikipedia.org/wiki/JRT, and several
places you can download it, just google for sources.
It wasn't very good, and if you ordered it you had to
wait a very long time before you received it. Borland's
Pascal (even the early versions) were much better.
Try running a program in CP/M with JRT Pascal 1.0
on a V20. CPU bugs
can be very subtle.
Isn't JRT Pascal the one that had really horrible reviews in the magazines
back when? Pournelle, maybe, though I couldn't say for sure as it's been a
while.
8:54 p.m.
Chuck Guzis wrote:
On 21 Feb 2009 at 15:09, Les Hildenbrandt wrote:
Try running a program in CP/M with JRT Pascal 1.0 on a V20. CPU bugs
can be very subtle.
Well what are the bugs?
The
problem with the V20's 8080 emulation was that it was buggy (I
have a few MicroNotes on the subject and can even claim to have
discovered one of the bugs).
I wonder if it was only cetain chips. I ran cp/m 80
on a v20 for years. Cheers,
Chuck
9:42 p.m.
On 22 Feb 2009 at 18:54, bfranchuk at jetnet.ab.ca wrote:
Well what are the bugs?
Here's the one with JRT Pascal (IIRC):
LXI SP,LABELX
CALL LABELX
LABELX:
...
RET
crashes and burns on a V20. Runs fine on an 8080/Z80. I may not have
it exactly--this was over 20 years ago and it's going to take some
digging to find my paperwork on the V20. Until then, you'll have to
rely on my ever-more-tenuous memory it or try it for yourself.
V20 emulation mode was enough of a problem that we eventually
disabled it in 22NICE. x86 machines were getting powerful enough
that software emulation was faster anyway. We could also do things
in software emulation that we couldn't do in V20 mode, such as trap
I/O accesses and redirect them to user-written servicers.
My contact at NEC Natick was Rich Naro and he indicated that 8080
emulation mode was very low on the "things to be fixed" list in the V-
series chips.
There were also some very strange behavior when an 8087 was included
into the mix.
Cheers,
Chuck
9:52 p.m.
On Sun, Feb 22, 2009 at 8:42 PM, Chuck Guzis <cclist at sydex.com> wrote:
V20 emulation mode was enough of a problem that we
eventually
disabled it in 22NICE. x86 machines were getting powerful enough
that software emulation was faster anyway. We could also do things
in software emulation that we couldn't do in V20 mode, such as trap
I/O accesses and redirect them to user-written servicers.
I understand wanting to minimize user frustration, but back in the day
(1991?), I did stick a V20 in my XT clone specifically to run 22NICE
to run the CP/M versions of the Scott Adams games (since those were
the easiest versions to find at the time). For what I was doing, it
worked great.
-ethan
10:46 p.m.
On 22 Feb 2009 at 20:52, Ethan Dicks wrote:
I understand wanting to minimize user frustration, but
back in the day
(1991?), I did stick a V20 in my XT clone specifically to run 22NICE
to run the CP/M versions of the Scott Adams games (since those were
the easiest versions to find at the time). For what I was doing, it
worked great.
The "disabling" was only in removing the option specification from
the GENCOM utility--the code was still in 22NICE itself. If you used
an older version of GENCOM, or just updated the 22NICE.COM resident,
V20 emulation would work. The change was made cica 1993.
The I/O remapping worked for some of our customers wanting to
substitute a PC in their existing equipment, particularly those
involving CNC. I recall a Strippit sheet metal punch adapted in this
manner.
That was back when the US still manufactured things...sigh.
Cheers,
Chuck
5605
days inactive
5607
days old
9 comments
6 participants
participants (6)
-
bfranchuk＠jetnet.ab.ca -
cclist＠sydex.com -
ethan.dicks＠gmail.com -
kth＠srv.net -
les＠hildenbrandt.com -
rtellason＠verizon.net