BBC News Blunder

On Wed, 2 Aug 2017 16:43:37 +0100, Max Demian
wrote:

If my memory serves me correctly, apart from the ICL 1900 series with
6-bit bytes (and escape sequences for the less likely to be used range),
the DEC10 had 7-bit bytes, five to a 36-bit word (the other one was
either a sign bit or parity, I can't remember which).

As I remember, the DEC-10 represented characters (with capital letters
only) in six bits, six to a word. Hence the packed and unpacked arrays
of char found in Pascal. I don't know whether the term 'byte' was used
in this context.

The DEC-20 certainly used 5 lots of 7 in 36 because we used to get
2560 characters per 512 word page of disk storage.
I can't remember definitively about the DEC-10 as I spent a lot less
time with it (use of it coming after the 20 and thus being somewhat
inferior and annoying).

On Wed, 2 Aug 2017 17:03:16 +0100, Max Demian
wrote:

"Byte" has two big advantages over "octet", as a word:

1. It is a single syllable

I'll second that

2. It starts with a consonant, so you don't get the problem of
having to elide the vowel at the end of SI prefix (kilo, mega,
giga, tera) with the vowel at the beginning of octet - or have the
intrusive R, as in "laura norder" :-)

though kiloctet and megoctet might have become common usage (much as
with ohms)

I don't recognise that usage for Ohms

It's kilohm, not kiloohm.

yeah

but it's mega-ohm

No it's not.

Not officially, but I think most of us would understand what was
meant, and maybe put it down as a personal affectation.

Likewise Big Clive's "millamps", though I don't recall hearing that
one recently, so maybe somebody's told him.

Rod.

On 02/08/2017 16:51, Peter Duncanson wrote:
On Wed, 02 Aug 2017 09:01:54 +0200, Martin wrote:

On Tue, 01 Aug 2017 22:32:04 +0100, Indy Jess John
wrote:

On 01/08/2017 11:17, Robin wrote:

Now if octet[1] had prevailed ...

[1] https://en.wikipedia.org/wiki/Octet_(computing)

It would have caused considerable confusion for users of the ICL 1900
range. It had a 24-bit word which divided into 4x6-bit bytes but would
have looked like 3 octets.

or even three bytes.

No doubt a program could have been written to store 3 8-bit thingies in
a 24-bit word, but that would have been unusual and internal to the
program.

The character input and output on paper tape used 7 bits, a 6-bit
character plus a parity bit.

Well ASCII has 7 bit paper tape origins, hence the NUL code (no holes
for the lead in) and the DEL code (7 holes to backspace and delete).

--
Max Demian

On 02/08/2017 18:53, charles wrote:
In article , R. Mark
Clayton wrote:

By the late 90's I had upgraded to 256MB and 400MHz clock, in 2004 to 1GB
and ~1,300MHz clock and in 2014 to 8GB, ~4GHz clock and four 64bit
processors plus six? GPU's all on the same die, so my current PC is about
100 times faster for each core, with ~3,000 times more memory than what
was the fastest computer in the world when I started. Oh and it cost
about one ten thousandth of the price, not even allowing for inflation...

Talking of which the first transistors I bought in 1967 each cost 6s 8d
(33p) = ~£6 today. You can now buy a trillion for around the same money.

and, on that topic I reemember a collegue buying LEDs at that time afor £1
each and he blew up 3 before he realised they had polarity.

Actually wiring LEDs the wrong way doesn't damage them - they just don't
light (as they are diodes). It's connecting them directly to a supply
without a ballast resistor that blows them.

--
Max Demian

In message , Max
Demian writes
On 02/08/2017 18:53, charles wrote:
In article , R. Mark
Clayton wrote:

By the late 90's I had upgraded to 256MB and 400MHz clock, in 2004 to 1GB
and ~1,300MHz clock and in 2014 to 8GB, ~4GHz clock and four 64bit
processors plus six? GPU's all on the same die, so my current PC is about
100 times faster for each core, with ~3,000 times more memory than what
was the fastest computer in the world when I started. Oh and it cost
about one ten thousandth of the price, not even allowing for inflation...

Talking of which the first transistors I bought in 1967 each cost 6s

(33p) = ~£6 today. You can now buy a trillion for around the same money.
and, on that topic I reemember a collegue buying LEDs at that time
afor £1
each and he blew up 3 before he realised they had polarity.

Actually wiring LEDs the wrong way doesn't damage them - they just
don't light (as they are diodes). It's connecting them directly to a
supply without a ballast resistor that blows them.

Oh no?
http://bit.ly/2vuqD8C
--
Ian

On 02/08/2017 22:58, Ian Jackson wrote:
In message , Max
Demian writes
On 02/08/2017 18:53, charles wrote:
In article ,
R. Mark
Clayton wrote:

Talking of which the first transistors I bought in 1967 each cost 6s
(33p) = ~£6 today. You can now buy a trillion for around the same
money.
and, on that topic I reemember a collegue buying LEDs at that time
afor £1
each and he blew up 3 before he realised they had polarity.

Actually wiring LEDs the wrong way doesn't damage them - they just
don't light (as they are diodes). It's connecting them directly to a
supply without a ballast resistor that blows them.

Oh no?
http://bit.ly/2vuqD8C

I meant a voltage similar to the forward voltage, though, apparently
there isn't so much difference between the maximum forward and reverse
voltages with blue (and presumably white) LEDs.

--
Max Demian

In article ,
Max Demian wrote:

https://en.wikipedia.org/wiki/Byte#History
"It is a deliberate respelling of bite to avoid accidental mutation to
bit." (FWIW)

Whereas parity is added to detect accidental mutation to bit


--
--------------------------------------+------------------------------------
Mike Brown: mjb[-at-]signal11.org.uk | http://www.signal11.org.uk

In article , Robin
writes
On 31/07/2017 19:59, Ian Jackson wrote:

Nick Ferrari wasn't wrong. He said (rather proudly) that he didn't
really know what 500kb/s or 28Mb/s meant. His tone was somewhat one of
'people like me don't need to bother with such trivia',

Spot on. Sadly the chattering classes don't just lack a sense of shame
that they are bad at maths, they announce the fact as if it wear a
badge of honour. Perhaps an implicit "My jobs far too important for me
to have to do sums" or "I'm far too wealthy to need to worry what 12
and a half per cent of anything is".


Perhaps they have an ambition to be Home Secretary?
--
bert

In article , AnthonyL
writes
On Mon, 31 Jul 2017 17:20:49 +0100, Ian Jackson
wrote:

In message , NY
writes
"Ian Jackson" wrote in message
...
As for always using the same units, it's horses for courses (miles or
furlongs).

All I'm saying is that maybe there is a case for a given application
(eg internet comms rates) always using the same units to make it easier
for people who aren't as conversant as us in SI prefix multipliers.
Actually, it's worse than that because k, M, G and T really imply steps
of 1000x, whereas in computing, it's common to use them to mean steps
of 1024x - except for hard disk manufacturers who use 1000 because it
results in slightly larger, more impressive numbers.

But is this really relevant to the situation where at least five people
didn't have a clue about the ratio between 500kb/s and 28Mb/s? As they
bothered to ring in, I would presume that they thought they knew
something about the subject. The closest wrong answer was 57 - but
heaven knows how the guy arrived at that conclusion.

To paraphrase a well known saying, there are 10 type of people, those
who understand bits and those that don't.

Or What's a parity error?




Pieces of 7, pieces of 7.

The old'uns are the best.
--
bert

bert wrote:

[snip]


Or What's a parity error?




Pieces of 7, pieces of 7.

No, that's a parroty error ...

--
Graham J