On Thu, Oct 28, 2010 at 9:04 PM, Dave Close <[email protected]> wrote:
> Brandon S Allbery KF8NH wrote:
>
>>I'll add that "octet" is itself something of a leftover from when the 36-bit
>>dinosaurs walked the earth.

Hi!  My name is Tom, and I'll be your dinosaur for the evening.

> That doesn't seem right to me. Certainly 36-bit machines (and 12-bit and
> 18-bit ones) frequently divided instruction words into 3-bit units and
> used octal notation to represent them. But 36%8 != 0.

36-bit machines used several character sets.  Most 36-bit machines
(IBM, Honeywell, DEC) started with 6-bit character sets.  BCD, SIXBIT,
etc.

Note that EBCDIC is the *Extended*  Binary Coded Decimal Interchange
Code, eg it was an 8-bit extension of the earlier 6-bit character set.
 When EBCDIC and ASCII came along, some hardware offered multiple ways
to pack 8-bit characters into 36 bits.

Honeywell/GE Multics and GCOS (36-bit) hardware offered either 4x8-bit
characters all packed into the lowest bits plus padding bits at the
top of the 36-bit word -or- 4x 9-bit bytes with the most significant
bit of each byte as padding.  And there were *hardware* instructions
to process all these formats, and to convert to/from each format and
between character sets, even.

--tep, the dinosaur
_______________________________________________
Tech mailing list
[email protected]
http://lists.lopsa.org/cgi-bin/mailman/listinfo/tech
This list provided by the League of Professional System Administrators
 http://lopsa.org/

Reply via email to