Re: Need C++ tutor for 10th grade student

["Nicholas Bodley" <user=1375@nabble.blu.org>]

 On Wed, 13 Feb 2008 11:28:35 -0500, Nathan Meyers <[hidden email]>   
wrote: 

> But did you have a Digi-Comp I 3-bit plastic computer   
> (http://www.oldcomputermuseum.com/digicomp_1.html)? 

That was a real gem. I should have kept mine. It was three J-K mechanical   
flip-flops, with programmable feedback inputs. I made a pseudo-random   
counter, I think, among others. I also designed and built myself some   
small perfboard discrete-component logic; had fun with that. If you like   
that, Konrad Zuse's first machine (iirc) had floating point,   
decimal->binary input, binary->decimal output, and was totally mechanical.   
The technology was a stretch for floating point. (A Selectric's keyboard   
creates binary codes that go to two mechanical D to A converters to   
position the "golf ball".) 

Digi-Comp II was a marble machine (I didn't own one); iirc, it was   
essentially a four-arithmetic-function logic device. I read that some   
low-cost manufacturing method (thermoformed plastic sheet?) didn't work   
out well. 

My first language was assembler on the BMEWS DIP computer at the NORAD COC   
in Colorado Springs, before the went under the mountain. No CRT, and the   
console typewriter keyboard (custom Flexowriter) did not feed the machine;   
you had to punch a tape, then read it in, but keep parity correct; machine   
had a 19-bit word length*, so four consecutive tape characters were   
spliced to make a word. I used the maintenance console (lighted pushbutton   
for every bit in every register; arranged in three rows so you could read   
quickly in octal). Machine was ready to roll several weeks (iirc) ahead of   
schedule, so they OK'd midnight hacking; only proviso was that I be   
bright-eyed and keen-witted next morning. This was 1960. *Probably the   
only computer with a 19-bit word. I fairly recently saw a photo of a   
digital device from that era, maybe a navigational system, with 19-bit   
registers. Machine was fixed point, and 19 bits gave enough precision for   
the job. 

Also had borrowed time on the Philco 2000 orbiting-object   
cataloging/monitoring machine. (That machine had no clock; all logic was   
completion-recognition variety, asynch.) 
I learned its assembler, too.  Both machines had rather small instruction   
repertoires; the DIP had a five-bit op code, and no options for any   
instruction. 

When I left Colo. Springs in 1962, I was away from computers totally until   
1981, when I encountered CP/M, iirc on a Vector Graphic (it was neither,   
but a nice-sounding name) machine that used hard-sectored 5¼-in floppies,   
about 550 kB each, not too shabby. Been a computer type to some degree   
since. 

Still have the HP Vectra 386/16N, DOS 6.22 (fairly sure); also had Debian   
(2.1?) on it. It was/is one of very few machines with software to switch   
among ISO 8859 codepages. It could run Latin-1 (Codepage 819) as well as   
any up through ISO 8859-10. Of course, Codepage 437 box graphics got   
totally munged. It's now a monitor riser, with a housing quite strong   
enough to take the ~85-lb weight of my lovely curb find, a DEC VRC-21WA in   
very good condition, VGA-to-five-BNC cable included. 

(Until I learn how to fix it, new incoming e-mail (Opera browser) is   
unlisted. Re-indexing should do it.) 

Best regards, 

-- 
Nicholas Bodley 
Waltham, Mass. 
On the exaltation of ignorance: 
<http://preview.tinyurl.com/2cdwr5> 
soon to have six dozen 

-- 
This message has been scanned for viruses and 
dangerous content by MailScanner, and is 
believed to be clean. 

_______________________________________________ 
Discuss mailing list 
[hidden email] 
http://lists.blu.org/mailman/listinfo/discuss