10 Jun
2009
10 Jun
'09
5:14 a.m.
I have in my grubby little hands an Elektronika MK-85 calculator, which
externally is a spot-on clone of the Casio FX-700P BASIC programmable
calculator. Functionally, it's identical as well, with a few
enhancements (pixel-addressable graphics, a Cyrillic character-set, etc)
Internally though, they're 100% different -- the processor in the MK-85
is a Russian PDP-11 knockoff. It's not a power-efficient CPU by any
means, so to ensure decent battery life the CPU speed is severely
limited. Not sure exactly what speed it runs at (anyone out there
know?) but the result is by far the slowest calculator I've ever used.
Computing the sine of an angle in degrees using the built-in "SIN"
function takes anywhere between 2 and 7 seconds by my stopwatch.
So the MK-85 takes an otherwise-elegant and useful 16-bit CPU, attaches
it to a mere 2K of RAM and 8K of ROM and runs at an insanely low clock
speed in order to allow the battery life to be more than a few
minutes... I just have to wonder why they decided to go with a PDP-11
clone over, say, a Z80 clone... :).
(The BASIC implementation is also incredibly buggy, mostly due to poor
argument checking... see http://www.pisi.com.pl/piotr433/mk85mc1e.htm
for a cool example of exploiting a bug in INPUT to do machine-language
coding, in a way only a contortionist could love...)
I just find this machine fascinating, for a number of reasons. A
handheld PDP-11! (Sort of.) Anyone else know of examples of odd-duck
machines like this, where the hardware is probably not the best choice
for the application? (But it's cool anyway?)
- Josh
10 Jun
10 Jun
5:30 a.m.
On Tue, 9 Jun 2009, Josh Dersch wrote:
I have in my grubby little hands an Elektronika MK-85
calculator, which
externally is a spot-on clone of the Casio FX-700P BASIC programmable
calculator. Functionally, it's identical as well, with a few enhancements
(pixel-addressable graphics, a Cyrillic character-set, etc)
With all the stuff that the Russians did with PDP-11 knockoffs, I wonder
what would have happened if DEC took the market of home computers
seriously. The UKNC (which to me resembles an MSX), is particularly
interesting.
--
David Griffith
dgriffi at cs.csubak.edu
A: Because it fouls the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?
6:49 a.m.
On 9 Jun 2009 at 22:14, Josh Dersch wrote:
I have in my grubby little hands an Elektronika MK-85
calculator,
which externally is a spot-on clone of the Casio FX-700P BASIC
programmable calculator. Functionally, it's identical as well, with a
few enhancements (pixel-addressable graphics, a Cyrillic
character-set, etc)
Very cool, Josh! I've admired the universal application of the
Soviet LSI-11 clone chips. Probalbly the one item that represents
the height of it all is the MK-87--a pocket-sized calculator:
http://rk86.com/frolov/mk-87.htm
I have no idea how long the batteries lasted on that one...
--Chuck
12:39 p.m.
(The BASIC implementation is also incredibly buggy,
mostly due to poor
argument checking... see http://www.pisi.com.pl/piotr433/mk85mc1e.htm
for a cool example of exploiting a bug in INPUT to do machine-language
coding, in a way only a contortionist could love...)
That is outstanding. This reminds me of the stack-pave trick for writing to
Tomy Tutor VDP RAM by exploiting bugs in SCELL and GCELL (although no
explicit execution of 9900 assembly ... yet).
--
------------------------------------ personal: http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com
-- This manual has been carefully for errors to make sure correct. -- classiccmp
12:47 p.m.
Josh Dersch wrote:
Internally though, they're 100% different -- the
processor in the MK-85
is a Russian PDP-11 knockoff. It's not a power-efficient CPU by any
means, so to ensure decent battery life the CPU speed is severely
limited. Not sure exactly what speed it runs at (anyone out there
know?) but the result is by far the slowest calculator I've ever used.
That's interesting. Does it have provision for AC input? Just curious if it
auto-magically ups the clock speed when not running from the battery, as that
would be kinda cool (and an early example of a power-saving mode :-)
(The BASIC implementation is also incredibly buggy,
mostly due to poor
argument checking... see http://www.pisi.com.pl/piotr433/mk85mc1e.htm
for a cool example of exploiting a bug in INPUT to do machine-language
coding, in a way only a contortionist could love...)
Gah, one of the home micro BASICs did something similar, so you could throw MC
in there as a character string and 'trick' the BASIC into executing it by
tripping the parser up - but my brain's refusing to tell me which one it was now.
Anyone else know of examples of odd-duck machines like
this, where the
hardware is probably not the best choice for the application?
Anything ever done using an IBM-compatible PC?
(But it's cool anyway?)
Oh. Scratch that, then.
;-)
1:07 p.m.
On Jun 10, 2009, at 8:47 AM, Jules Richardson wrote:
This is a pretty common trick on the ZX81/TS-1000. You'd embed
machine code into a REM statement at the top of a program, as a
string of characters. It was a real bitch to enter that stuff from
magazines. :-)
-Dave
--
Dave McGuire
Port Charlotte, FL
(The BASIC
implementation is also incredibly buggy, mostly due to
poor argument checking... see http://www.pisi.com.pl/piotr433/
mk85mc1e.htm for a cool example of exploiting a bug in INPUT to do
machine-language coding, in a way only a contortionist could love...)
Gah, one of the home micro BASICs did something similar, so you
could throw MC in there as a character string and 'trick' the BASIC
into executing it by tripping the parser up - but my brain's
refusing to tell me which one it was now.
1:26 p.m.
(The BASIC implementation is also incredibly buggy,
mostly due to
poor argument checking... see http://www.pisi.com.pl/piotr433/
mk85mc1e.htm for a cool example of exploiting a bug in INPUT to do
machine-language coding, in a way only a contortionist could love...)
Gah, one of the home micro BASICs did something similar, so you
could throw MC in there as a character string and 'trick' the BASIC
into executing it by tripping the parser up - but my brain's
refusing to tell me which one it was now.
1:33 p.m.
On Jun 10, 2009, at 9:26 AM, Cameron Kaiser wrote:
This is a pretty common trick on the ZX81/TS-1000. You'd embed
machine code into a REM statement at the top of a program, as a
string of characters. It was a real bitch to enter that stuff from
magazines. :-)
Right, you could do this on a C64 also and probably many machines,
but this
was just a convenient means of storage -- execution of arbitrary ML
was
still permitted by the interpreter. The MK-85 hack is noteworthy
because it
allows you to run stuff on the CPU even though the BASIC supposedly
doesn't
let you (no CALL, SYS, USR(), etc).
OH! Oh yes, I see what you mean. Different trick entirely. Very
cool.
-Dave
--
Dave McGuire
Port Charlotte, FL
(The BASIC
implementation is also incredibly buggy, mostly due to
poor argument checking... see http://www.pisi.com.pl/piotr433/
mk85mc1e.htm for a cool example of exploiting a bug in INPUT to do
machine-language coding, in a way only a contortionist could
love...)
Gah, one of the home micro BASICs did something similar, so you
could throw MC in there as a character string and 'trick' the BASIC
into executing it by tripping the parser up - but my brain's
refusing to tell me which one it was now.
9:30 p.m.
Cameron Kaiser wrote:
Right, you could do this on a C64 also and probably
many machines, but this
was just a convenient means of storage -- execution of arbitrary ML was
still permitted by the interpreter. The MK-85 hack is noteworthy because it
allows you to run stuff on the CPU even though the BASIC supposedly doesn't
let you (no CALL, SYS, USR(), etc).
Yes, the one I'm remembering was similar to that - I wish I could remember
more, but it was entered via some function which had a utterly different
purpose during normal use, but invoked in a certain way it could be used to
run sequences of MC.
Dave's message rings a bell though, in that I think the actual MC data was
encoded into a character string (possibly as a REM statement, but I can't be
sure).
If my memory coughs up the name of the system I'll shout :-)
11 Jun
11 Jun
4:34 a.m.
Jules Richardson wrote:
Cameron Kaiser wrote:
Then there's the fun undocumented 'CALL "EXEC"' command for the
Tektronix 4051, which executes 6800 assembly instructions in a
specially-encoded 7-bit ASCII string, two characters per byte...
Josh
Right, you could do this on a C64 also and
probably many machines,
but this
was just a convenient means of storage -- execution of arbitrary ML was
still permitted by the interpreter. The MK-85 hack is noteworthy
because it
allows you to run stuff on the CPU even though the BASIC supposedly
doesn't
let you (no CALL, SYS, USR(), etc).
Yes, the one I'm remembering was similar to that - I wish I could
remember more, but it was entered via some function which had a
utterly different purpose during normal use, but invoked in a certain
way it could be used to run sequences of MC.
Dave's message rings a bell though, in that I think the actual MC data
was encoded into a character string (possibly as a REM statement, but
I can't be sure).
If my memory coughs up the name of the system I'll shout :-)
8:53 a.m.
Jules Richardson wrote:
:)
Josh
Josh Dersch wrote:
It has an external power input, via some weird connector on the side.
It _does_ have a "high speed" mode -- if you turn it on while holding
"+" it runs 6 or 7 times faster (tested by running a program that just
runs a FOR loop from 0 to 1000 -- ~28 seconds in "normal" mode, ~4
seconds in "high speed" mode). No idea how it affects battery life, but
I can't imagine it's good :).
Internally though, they're 100% different --
the processor in the
MK-85 is a Russian PDP-11 knockoff. It's not a power-efficient CPU
by any means, so to ensure decent battery life the CPU speed is
severely limited. Not sure exactly what speed it runs at (anyone out
there know?) but the result is by far the slowest calculator I've
ever used.
That's interesting. Does it have provision for AC input? Just curious
if it auto-magically ups the clock speed when not running from the
battery, as that would be kinda cool (and an early example of a
power-saving mode :-) (The BASIC implementation is also incredibly
buggy, mostly due to
poor argument checking... see
http://www.pisi.com.pl/piotr433/mk85mc1e.htm for a cool example of
exploiting a bug in INPUT to do machine-language coding, in a way
only a contortionist could love...)
Gah, one of the home micro BASICs did something similar, so you could
throw MC in there as a character string and 'trick' the BASIC into
executing it by tripping the parser up - but my brain's refusing to
tell me which one it was now.
Anyone else know of examples of odd-duck machines
like this, where the
hardware is probably not the best choice for the application?
Anything ever done using an IBM-compatible PC?
(But it's cool anyway?)
Oh. Scratch that, then.
;-)
10 Jun
10 Jun
8:35 p.m.
Josh Dersch wrote:
I have in my grubby little hands an Elektronika MK-85
calculator,
which externally is a spot-on clone of the Casio FX-700P BASIC
programmable calculator. Functionally, it's identical as well, with a
few enhancements (pixel-addressable graphics, a Cyrillic
character-set, etc)
Pixel-addressable characters. I wouldn't really call it
"graphics"
since there are gaps between the character cells.
Internally though, they're 100% different -- the
processor in the
MK-85 is a Russian PDP-11 knockoff.
There is also the MK-90, with a larger
graphics display and removable
RAM cartridges. Purportedly came with a RAM cartridge containing a
Tetris game, but there are no known surviving copies of the game.
Eric
11 Jun
11 Jun
4:39 a.m.
Eric Smith wrote:
Yeah, the MK-90 looks really neat. love to find one one of those someday...
As far as Tetris (and other games)... looks like there's at least one
known copy out there: http://www.youtube.com/watch?v=48pmWw4TG1o (See at
about 2:13).
Josh
Josh Dersch wrote:
Well, that's just being
pedantic :). I can plot a point at any
arbitrary x,y coordinate. It's just that there are gaps here and there
along the X axis :).
I have in my grubby little hands an Elektronika
MK-85 calculator,
which externally is a spot-on clone of the Casio FX-700P BASIC
programmable calculator. Functionally, it's identical as well, with
a few enhancements (pixel-addressable graphics, a Cyrillic
character-set, etc)
Pixel-addressable characters. I wouldn't really call it
"graphics"
since there are gaps between the character cells. Internally
though, they're 100% different -- the processor in the
MK-85 is a Russian PDP-11 knockoff.
There is also the MK-90, with a larger
graphics display and removable
RAM cartridges. Purportedly came with a RAM cartridge containing a
Tetris game, but there are no known surviving copies of the game.
Eric
12:47 p.m.
I have in my grubby little hands an Elektronika MK-85
calculator,
which externally is a spot-on clone of the Casio FX-700P BASIC
programmable calculator. Functionally, it's identical as well, with
a few enhancements (pixel-addressable graphics, a Cyrillic
character-set, etc)
Pixel-addressable characters. I wouldn't really call it
"graphics"
since there are gaps between the character cells.
5690
days inactive
5691
days old
13 comments
7 participants
participants (7)
-
cclist＠sydex.com -
derschjo＠mail.msu.edu -
dgriffi＠cs.csubak.edu -
eric＠brouhaha.com -
jules.richardson99＠gmail.com -
mcguire＠neurotica.com -
spectre＠floodgap.com