Searching \ for '[OT]Is Kibibyte really an ISO reality?' in subject line. ()
Help us get a faster server
FAQ page: massmind.org/techref/index.htm?key=kibibyte+really
Search entire site for: 'Is Kibibyte really an ISO reality?'.

Exact match. Not showing close matches.
'[OT]Is Kibibyte really an ISO reality?'
2009\04\14@145239 by

I'm not sure if it is a German April Fools - arriving a bit late.

I cut and paste:

' Units are regulated by the International Bureau of Weights and
Measures. It determines the absolute value of a kilo as the
thousandfold value of its original value. In IT a kilo always means
1,024, therefore, in 2000 the International Bureau of Weights and
Measures determined that 1,000 Bytes are a Kilobyte (KB). As a result,
whenever 1,024 Bytes are meant, then the unit Kibibyte (KiB) should be
used. '

They have provided a table explaining,  Kibi, Mebi, Gibi, Tebi and
Pebibyte.

If this is fact, then that fudge is almost as bad as the ridiculous
IEEE logic symbols.

Colin
--
cdb,  on 15/04/2009

On Wed, 2009-04-15 at 04:52 +1000, cdb wrote:
> I've just received a newsletter from a German software company and
> I'm not sure if it is a German April Fools - arriving a bit late.
>
> I cut and paste:
>
> ' Units are regulated by the International Bureau of Weights and
> Measures. It determines the absolute value of a kilo as the
> thousandfold value of its original value. In IT a kilo always means
> 1,024, therefore, in 2000 the International Bureau of Weights and
> Measures determined that 1,000 Bytes are a Kilobyte (KB). As a result,
> whenever 1,024 Bytes are meant, then the unit Kibibyte (KiB) should be
> used. '
>
> They have provided a table explaining,  Kibi, Mebi, Gibi, Tebi and
> Pebibyte.
>
> If this is fact, then that fudge is almost as bad as the ridiculous
> IEEE logic symbols.

It's real, it's sad, and most people think of it as a joke.

While the "ibi" units gained some popularity a few years ago, they are
almost completely gone these days. Nobody I know SAYS kibi, or mebi, a
few publications publish it, but on the whole it's dead as in doodoo.
Some open source computer software uses it, that's probably where I most
often see it.

The horses are out of the barn, there's no putting them back in, we just
have to live with the fact that kilo/mega/etc means something different
depending on context. It's a shame, but that's how it is.

TTYL

cdb wrote:
> If this is fact, then that fudge is almost as bad as the ridiculous
> IEEE logic symbols.

Which ones are they -- the square boxes with "&", "*" and so on inside them,
or the D- and triangular-shaped logic gate symbols that seem to have been in
use since the year dot?

--
Phil.
piclistphilpem.me.uk
http://www.philpem.me.uk/

:: Which ones are they -- the square boxes with "&", "*" and so on
:: inside them

Yup they's (sic) the ones, Elektor have always loved using them, and I
have to look them up almost everytime.

Colin
--
cdb, colinbtech-online.co.uk on 15/04/2009

Web presence: http://www.btech-online.co.uk

Hosted by:  http://www.1and1.co.uk/?k_id=7988359

I did recently come across a list of numeric prefixes which had all
the usual ones, but then also added "guaca" as 10^-23, so that a
that a guacamole=1). Most of the table was serious but that one was a
joke.

On Tue, Apr 14, 2009 at 2:52 PM, cdb <colinbtech-online.co.uk> wrote:
{Quote hidden}

>
cdb wrote:
> :: Which ones are they -- the square boxes with "&", "*" and so on
> :: inside them
>
> Yup they's (sic) the ones, Elektor have always loved using them,

So do Texas Instruments. Though the "IC" ones are slightly less obtuse, and at
least the TI "Digital Logic Pocket Databook" and most of the TI datasheets
have both symbols in them.

> and I
> have to look them up almost everytime.

Welcome to the club...

--
Phil.
piclistphilpem.me.uk
http://www.philpem.me.uk/
In message <1239735685.16325.13.camel@not-the-one>, Herbert Graf
<hkgrafgmail.com> writes
>While the "ibi" units gained some popularity a few years ago, they are
>almost completely gone these days. Nobody I know SAYS kibi, or mebi, a
>few publications publish it, but on the whole it's dead as in doodoo.
Apart from with storage manufacturers, it's the weasel words which allow
them to advertise disks as xGB when in fact you're 24MB short on every
GB.

--
Clint Sharp
On Wed, Apr 15, 2009 at 2:52 AM, cdb <colinbtech-online.co.uk> wrote:
>
> They have provided a table explaining,  Kibi, Mebi, Gibi, Tebi and
> Pebibyte.
>
> If this is fact, then that fudge is almost as bad as the ridiculous
> IEEE logic symbols.
>

The following links may shed some lights.
http://physics.nist.gov/cuu/Units/binary.html
"Faced with this reality, the IEEE Standards Board decided that IEEE
standards will use the conventional, internationally adopted, definitions
of the SI prefixes. Mega will mean 1 000 000, except that the base-two
definition may be used (if such usage is explicitly pointed out on a
case-by-case basis) until such time that prefixes for binary multiples
are adopted by an appropriate standards body."

http://en.wikipedia.org/wiki/Kibibit
http://en.wikipedia.org/wiki/Kilobyte

--
Xiaofan http://mcuee.blogspot.com
On Tue, Apr 14, 2009 at 22:03, cdb <colinbtech-online.co.uk> wrote:

> Yup they's (sic) the ones, Elektor have always loved using them, and I
> have to look them up almost everytime.

You're saying that a square box with half of a circle to the end is
easier to remember it's an AND-gate, than a square with an "&" inside?

--
- Rikard - http://bos.hack.org/cv/
>You're saying that a square box with half of a circle to the
>end is easier to remember it's an AND-gate, than a square with
>an "&" inside?

It is easier to compare to OR and XOR gate symbols, rather than have to look
inside square boxes to see what the logic symbol is. When a schematic is
drawn large, and then reduced to small paper sizes for printing, the logic
symbol inside the square box can become unreadable. At least the 'American'
symbols are distinguishable from each other when reduced in a similar
manner.

There are things about some ISO symbols that I like, e.g. counters and shift
registers, where the control lines get separated into their own area. At
least some thought went into how these should appear, but the reasoning
behind having these low level symbols as identical boxes without
distinguishing shape escapes me.

>> You're saying that a square box with half of a circle to the
>> end is easier to remember it's an AND-gate, than a square with
>> an "&" inside?

OK.  No.  But it's easier to recognize the old-style gates "at a
glance", and some (many?) of the DIN (in Eagle it's the DIN logic
library that has the "offending" symbools?) symbols for the more
complex gates are really incomprehensible (though in retrospect it's
not clear that the older symbols there are any better.)
The whole "everything is a squarish box" philosophy that seems to
pervade is a giant step backward in schematic clarity, IMO.  It's like
like the thought process was:

1) Americans and Europeans can't agree on symbols for gates, so we'll
use something completely different from either one.

2) high-res graphics terminals/cards are just too expensive, so
everything has to be drawable using the linemode character-based
graphics in the IBM PC font...

BillW

On Wed, Apr 15, 2009 at 3:56 PM, William "Chops" Westfield
<westfwmac.com>wrote:

> 1) Americans and Europeans can't agree on symbols for gates, so we'll
> use something completely different from either one.
>

Same with the resistor. I used to use the European one, but now I have the
feeling that it is easier to recognise the American one with that zig-zag,
and is definitely easier to draw with a pen.

With the logic gates I am not comfortable looking at the little circles on
the joints or the double curves but in that manner those rectangular ones
are not even better.

Look at the difference between the AND and NAND
http://en.wikipedia.org/wiki/Logic_gate

Tamas
--
http://www.mcuhobby.com
On Wed, 2009-04-15 at 09:04 +0100, Clint Sharp wrote:
> In message <1239735685.16325.13.camel@not-the-one>, Herbert Graf
> <hkgrafgmail.com> writes
> >While the "ibi" units gained some popularity a few years ago, they are
> >almost completely gone these days. Nobody I know SAYS kibi, or mebi, a
> >few publications publish it, but on the whole it's dead as in doodoo.
> Apart from with storage manufacturers, it's the weasel words which allow
> them to advertise disks as xGB when in fact you're 24MB short on every
> GB.

To be fair the storage manufacturers technically are the only ones that
have it RIGHT. 1MB is SUPPOSED to be 1,000,000 bytes, that's what "mega"
means. It's everyone else that has redefined "mega" to mean 2^20.

If any blame is to be laid it should really go to the rest of the
industry for redefining KB/MB/GB/TB. That said, the storage
manufacturers could just "go with the flow" and end the discrepancy, but
as you mention that move would NOT be in their best interests.

TTYL

Clint Sharp wrote:

>> While the "ibi" units gained some popularity a few years ago, they
>> are almost completely gone these days. Nobody I know SAYS kibi, or
>> mebi, a few publications publish it, but on the whole it's dead as
>> in doodoo.

FWIW, I use them in documentation I write when I mean the power of two.
However, there are not many cases anymore where powers of two are
important, so it usually doesn't matter whether I write 2.5 G-whatever
or 2.3 Gi-whatever, and I use the former.

> Apart from with storage manufacturers, it's the weasel words which
> allow them to advertise disks as xGB when in fact you're 24MB short
> on every GB.

Try to figure out how many bits can be transferred in a second over a
pipe with 8 Mb/s. Then tell me how many MB/s this are (given that there
is no compression or transfer/encoding overhead, and that we're using
the standard 8 bits per byte). Then tell me how many GB of available
memory and disk space I need to store 1000 seconds of this transfer.
Should be simple, shouldn't it? :)

It's like with all the confusing domain-specific units: they seem to
work "well" as long as you stick your head between two big panels and
don't look left or right. The moment you leave that little "world",
things don't work that well anymore. There are many contexts where the
traditional memory-related re-definition of kilo, mega etc. seem to be
obvious. But there are many contexts where it isn't quite that clear
which one is used.

Besides -- ever wondered how many bytes a 1.44"MB" floppy has?

Gerhard

> Besides -- ever wondered how many bytes a 1.44"MB" floppy has?
>
1.44*1000*1024 IIRC

On Fri, Apr 17, 2009 at 7:41 AM, peter green <plugwashp10link.net> wrote:
>
>> Besides -- ever wondered how many bytes a 1.44"MB" floppy has?
>>
> 1.44*1000*1024 IIRC
>
You are right.
http://en.wikipedia.org/wiki/Floppy_disk

--
Xiaofan http://mcuee.blogspot.com
Gerhard Fiedler wrote:

> Clint Sharp wrote:
>
>>> While the "ibi" units gained some popularity a few years ago, they
>>> are almost completely gone these days. Nobody I know SAYS kibi, or
>>> mebi, a few publications publish it, but on the whole it's dead as
>>> in doodoo.
>
> FWIW, I use them in documentation I write when I mean the power of two.

Also FWIW, the authors of Filezilla (an open source ftp client) seem to
use them, too.

Gerhard
Tamas Rudnai wrote:
>> 1) Americans and Europeans can't agree on symbols for gates, so we'll
>> use something completely different from either one.
>>
>
> Same with the resistor. I used to use the European one, but now I have the
> feeling that it is easier to recognise the American one with that zig-zag,
> and is definitely easier to draw with a pen.

FWIW, I like the European one better, because for me it's definitely easier
to draw (it's just a rectangle). I have trouble remembering exactly how many
zig-zags I have to draw. I use the American symbol most of the time, of
course...

> With the logic gates I am not comfortable looking at the little circles on
> the joints or the double curves but in that manner those rectangular ones
> are not even better.
>
> Look at the difference between the AND and NAND
> http://en.wikipedia.org/wiki/Logic_gate

For gates, I like the US symbols, hands down. Even the little "inversion"
circles.

Vitaliy

More... (looser matching)
- Last day of these posts
- In 2009 , 2010 only
- Today
- New search...