Kingston "512MB" SD flash cards - only 488MB

While there may a more correct abbreviation, I do believe that in terms of
hard drive and RAM byte capacities, the 1024 = 1K thesis is in fact,
correct. For years computer users have been either slightly and usually
inconsequently off, but if you ever puzzled over the numbers that a drive
manufacturer offers for capacity and then look at the report generated by a
directory listing, there has always been that strange and wonderful
disparity between the "real" K or Mb measurement and the binary K or Mb
measurement. Somehow the disks never quite measure up to their claims.
Transmission bandwidth is a different story. Bandwidth has often been
reported as "bits per second" and then again as "bytes per second" or even
"baud" (which I understand is the number of waveform edges per second).
Bytes of throughput may be deceptive since a transmitted byte can be various
number of bits depending on ASCII convention followed and checkbits
transmitted, etc. So in computerese it seems there is no *absolute* 1K or
1Mb. Goodness knows, there ain't no "correct", just conventions.

"Malcolm Weir" wrote in message
...
On Sat, 05 Feb 2005 19:45:37 +0000, Alun Saunders
wrote:

Alan wrote:
In message , Alun Saunders
wrote

most OS's define it (correctly) as 1024*1024=1048576 bytes.

Er, no. The 1000x1000 figure is correct

It depends on your definition of "correct", I suppose.

How about the SI units?

Ever since I've been working in the computer field, which is getting for
27 years now, a kByte has been 1024 bytes and a MByte has been 1048576
bytes, at least for RAM, ROM and suchlike.

True, and they've incorrectly been using SI prefixes all the time!

You do know that a Kg is not 1024g, and a Km is not 1024m, right?

I can't honestly say whether
hard disks way back then (which were in the order of 5 or 10 MBytes)
used that or the 10^6 version, and let's face it with capacities like
that the absolute difference between the two is pretty small, but I
still find it misleading to use one definition for memory and another
for storage purposes, and historically I find the 1 MByte=1048576 bytes
the more "correct" one. You can think what you like, however.

How many bits per second can you send down a 1 megabit/sec line?

It has *never* been anything than 1,000,000. Because frequencies
(1MHz) have always been correct.

Still, in the early days, disk manufacturers sometimes quoted
capacities in units of 2000 sectors, each of 512 bytes, i.e. 1024000
byes (also called "megabytes").

The correct prefix for the binary power is "Mebi", written "Mi", as in
"Mebibytes" (MiB).

Malc.

Recalling some fundamentals ---
Here's why the Kilo-, Mega-, and Giga- terms were appealing.
2^10 ~ 1K
2^20 ~ 1M
2^30 ~ 1G
2^40 ~ 1T
etc.

It's a curious blend of base-2 numbers and base-10 groupings.
Early computer people liked that, and ran with it.
FWIW, that predated the time when a byte was universally 8 bits.

"Pieter Litchfield" wrote in message =
...
While there may a more correct abbreviation, I do believe that in =
terms of=20
hard drive and RAM byte capacities, the 1024 =3D 1K thesis is in fact, =

correct. For years computer users have been either slightly and =
usually=20
inconsequently off, but if you ever puzzled over the numbers that a =
drive=20
manufacturer offers for capacity and then look at the report generated =
by a=20
directory listing, there has always been that strange and wonderful=20
disparity between the "real" K or Mb measurement and the binary K or =
Mb=20
measurement. Somehow the disks never quite measure up to their =
claims.
Transmission bandwidth is a different story. Bandwidth has often been =

reported as "bits per second" and then again as "bytes per second" or =
even=20
"baud" (which I understand is the number of waveform edges per =
second).=20
Bytes of throughput may be deceptive since a transmitted byte can be =
various=20
number of bits depending on ASCII convention followed and checkbits=20
transmitted, etc. So in computerese it seems there is no *absolute* =
1K or=20
1Mb. Goodness knows, there ain't no "correct", just conventions.

On Tue, 08 Feb 2005 02:29:43 GMT, "Pieter Litchfield"
wrote:

While there may a more correct abbreviation, I do believe that in terms of
hard drive and RAM byte capacities, the 1024 = 1K thesis is in fact,
correct.

The SI folks disagree. You lose.

For years computer users have been either slightly and usually
inconsequently off, but if you ever puzzled over the numbers that a drive
manufacturer offers for capacity and then look at the report generated by a
directory listing, there has always been that strange and wonderful
disparity between the "real" K or Mb measurement and the binary K or Mb
measurement. Somehow the disks never quite measure up to their claims.

Nope. They live up to their claims, you just didn't read what the
claim really WAS. You just presumed that they were making the same
mistake as you...

Transmission bandwidth is a different story.

No, it isn't. Same deal. Same prefix.

Bandwidth has often been
reported as "bits per second" and then again as "bytes per second" or even
"baud" (which I understand is the number of waveform edges per second).

Not really. Baud is the number of information units per second. An
information unit may be one or more than one bit.

Bytes of throughput may be deceptive since a transmitted byte can be various
number of bits depending on ASCII convention followed and checkbits
transmitted, etc.

Sure. But a 1Mbit/sec line is not deceptive: it transmits 1,000,000
bits per second.

By the way, you know your PCI bus at 32 bits wide and 33MHz transmits
132,000,000 bits per second, right? As in 132MB/sec.

So in computerese it seems there is no *absolute* 1K or
1Mb. Goodness knows, there ain't no "correct", just conventions.

False. There *is* a correct usage, and common (incorrect usage).

I mean, where did we *get* the prefixes K, M, G, and T, etc...?

Look, we've been misusing those SI prefixes for years. We know what
we mean. But let's not pretend that the misuse is "correct".

Malc.

Malcolm Weir writes:
By the way, you know your PCI bus at 32 bits wide and 33MHz transmits
132,000,000 bits per second, right? As in 132MB/sec.

(you meant bytes per second, but the number's right.)

Look, we've been misusing those SI prefixes for years. We know what
we mean. But let's not pretend that the misuse is "correct".

Yup.

Anton

In article ,
Pieter Litchfield wrote:
While there may a more correct abbreviation,

They are defined. It is the computer people that are
misusing the metric prefixes and positive powers fof 1024
prefixes have been defined but as of yet are not commonly used.

http://physics.nist.gov/cgi-bin/cuu/...ts/binary.html

http://physics.nist.gov/cgi-bin/cuu/.../prefixes.html

(( For the EU residents, this is an international standard
which is applicable to the EU as well as the rest of
the world.
))

The ``marketeers'' prefer to use the SI/metric prefixes
rather than the binary prefixes as they give bigger numbes.

The computer ``nerds'' prefer the binary sized meanings
as they more correctly their pattern of usage.

Now just get the computer ``nerds'' to switch from
1.073741824 to 1GiB (512MB = 488.28125MiB).

1Mb. Goodness knows, there ain't no "correct", just conventions.

There is a ``correct.'' But, generally, people are
not using it as they are continuing their old practices.
The correct prefix for the binary power is "Mebi", written "Mi", as in
"Mebibytes" (MiB).

Malc.

Ipse dixit.

(( In the original meaning of the Latin expression:
he/she/it said it (him/her/it)self.
))

Randolph J. Herber, , +1 630 840 2966, CD/CDFTF PK-149F,
Mail Stop 318, Fermilab, Kirk & Pine Rds., PO Box 500, Batavia, IL 60510-0500,
USA. (Speaking for myself and not for US, US DOE, FNAL nor URA.) (Product,
trade, or service marks herein belong to their respective owners.)

In article ,
Malcolm Weir wrote:
On Sat, 05 Feb 2005 19:45:37 +0000, Alun Saunders
wrote:

Alan wrote:
In message , Alun Saunders
wrote

most OS's define it (correctly) as 1024*1024=1048576 bytes.

Er, no. The 1000x1000 figure is correct

It depends on your definition of "correct", I suppose.

How about the SI units?

Ever since I've been working in the computer field, which is getting for
27 years now, a kByte has been 1024 bytes and a MByte has been 1048576
bytes, at least for RAM, ROM and suchlike.

True, and they've incorrectly been using SI prefixes all the time!

You do know that a Kg is not 1024g, and a Km is not 1024m, right?

I can't honestly say whether
hard disks way back then (which were in the order of 5 or 10 MBytes)
used that or the 10^6 version, and let's face it with capacities like
that the absolute difference between the two is pretty small, but I
still find it misleading to use one definition for memory and another
for storage purposes, and historically I find the 1 MByte=1048576 bytes
the more "correct" one. You can think what you like, however.

How many bits per second can you send down a 1 megabit/sec line?

It has *never* been anything than 1,000,000. Because frequencies
(1MHz) have always been correct.

Bzzt.

If you are thinking about the same 1-megabit lines as the G.703/G.704
lines standardize (and common throughout Europe and with some availability
in the rest of the world) i have to disappoint you.

G.703 specifies the bit rate as 2048000 +/- 2 bits per second, and
G.704 divides this into 32 timeslots of 64000 bits per second. All
designed to transmit 8000 samples of 8 bits per second.

A 1-megabit line has 16 timeslots for 1024000 bits per second.

Still, in the early days, disk manufacturers sometimes quoted
capacities in units of 2000 sectors, each of 512 bytes, i.e. 1024000
byes (also called "megabytes").

The correct prefix for the binary power is "Mebi", written "Mi", as in
"Mebibytes" (MiB).

Yeah, right.

-- mrr

In article ,
Alan wrote:
In message , Peter
wrote

I have three of these, all brand new. All show 488MB as the size when
formatting them and this is matched by the size of the files that can
be transferred to them from a PC.

Even allowing for 1k being 1024 etc this is a very big shortfall.

1k is actually 1000 when referring to hard disks, and presumably SD
cards.

A computer may report 1k byte as 1024 bytes. M$ Windows reports both
figures.

Converting from the true figure to the alternative (approximate) form of
reporting size gives

(512 x 1000 x 1000) / (1024x1024) = 488

So when the computer salesman (it is usually a male) sold you 512 meg
of storage for $0.1k, and you ended up with 488000000 bytes for $102.40.

(sorry about old joke, just appropriate for the setting).

-- mrr

Standards are only useful when adopted.

Although very intelligent people define the standards to which we
should adhere, the standards are ultimately set by the people who would
use them.

I understand the need for differentiating between 2^20 and 10^6, but
you'll be hard pressed to find more than 1 in 10 computer "nerds" who
are willing to refer to "Kilobits" as "Kibibits" or "Kaybee's" as
"Kibee's." *cringe*

Sometimes technically-oriented people overlook the real-world aspect of
what they do.

I think the lack of wide-spread adoption of this standard after seven
years is saying something. I think that until the manufacturers, namely
Memory Manufacturers, adopt this "new" standard, it's going to remain
jargon that only the truly savvy understand. Until then, the rest of us
are going to have to refer to the fine print if we want real numbers.

On 24 Feb 2005 19:24:15 -0800, wrote:

I understand the need for differentiating between 2^20 and 10^6, but
you'll be hard pressed to find more than 1 in 10 computer "nerds" who
are willing to refer to "Kilobits" as "Kibibits" or "Kaybee's" as
"Kibee's." *cringe*

Personally, as a "computer nerd", if I heard someone saying "Kaybee" I
would assume they meant Kilobytes.
--
Andrew, contact via interpleb.blogspot.com
Help make Usenet a better place: English is read downwards,
please don't top post. Trim replies to quote only relevant text.
Check groups.google.com before asking an obvious question.

If you have a Linux box available, try to login and run "/sbin/ifconfig eth0|grep bytes".
Surprise, surprise !
And Linux is a system that is run (mostly) by nerds...

wrote:
Standards are only useful when adopted.

Although very intelligent people define the standards to which we
should adhere, the standards are ultimately set by the people who would
use them.

I understand the need for differentiating between 2^20 and 10^6, but
you'll be hard pressed to find more than 1 in 10 computer "nerds" who
are willing to refer to "Kilobits" as "Kibibits" or "Kaybee's" as
"Kibee's." *cringe*

Sometimes technically-oriented people overlook the real-world aspect of
what they do.

I think the lack of wide-spread adoption of this standard after seven
years is saying something. I think that until the manufacturers, namely
Memory Manufacturers, adopt this "new" standard, it's going to remain
jargon that only the truly savvy understand. Until then, the rest of us
are going to have to refer to the fine print if we want real numbers.


--
Yuri Pismerov, Sr. System Administrator,
TUCOWS.COM INC. (416) 535-0123 ext. 1352