DailyTech - Western Digital Settles Drive Capacity Class-Action Lawsuit

Thats ridiculous

By unfalliblekrutch on 6/28/2006 3:50:16 PM , Rating: 2

Crazy lawyers... It's not like WD did anything wrong

RE: Thats ridiculous

By TomZ on 6/28/2006 3:58:25 PM , Rating: 1

I don't know... I always thought their definition of 1MB = 1,000,000 bytes, 1GB = 1,000,000,000 byte, etc. was somewhat deceptive.

Parent

RE: Thats ridiculous

By Tupolev22m on 6/28/2006 4:00:25 PM , Rating: 2

It's deceptive but its technically correct

Parent

RE: Thats ridiculous

By TomZ on 6/28/2006 4:08:54 PM , Rating: 2

Some would say that deception is wrong.

Parent

RE: Thats ridiculous

By creathir on 6/28/2006 4:42:47 PM , Rating: 5

Technically, the OS manuf. are the ones misleading...
A mebibyte is actually what they call a megabyte...
I do find that this is some area where people need to come to a level of agreement though.
- Creathir

Parent

RE: Thats ridiculous

By TomZ on 6/28/2006 5:06:43 PM , Rating: 1

quote:
Technically, the OS manuf. are the ones misleading...

How do you figure? The OS defines a HDD GB the same as most people define a GB = 1024^3 bytes. If I buy 1GB of RAM, I get 1024^3 bytes, not 1,000,000,000 bytes.

Parent

RE: Thats ridiculous

By Morpth on 6/29/2006 11:06:31 AM , Rating: 2

I looked this up, this mebibyte. What do you know? It's a real standard and it's been in use since 1998!! You learn something new everyday! (Link: http://physics.nist.gov/cuu/Units/binary.html)

I want to see the next major lawsuit on CRT TV manufactures. We buy a 34" TV but the screen size is always smaller!! Sue! Sue! Sue!...MUAHAHAHA =)

Parent

RE: Thats ridiculous

By TomZ on 6/29/2006 3:54:39 PM , Rating: 1

It is a "standard," but it is not in common use.

Parent

RE: Thats ridiculous

By Tyler 86 on 6/29/2006 6:11:59 PM , Rating: 2

I present to you a new standard.

From now on, we will measure IQ in pence. An intelligent person has 1 kilo-pence. A genius has 1 mega-pence. A mentally challenged individual has 1 pence.

Followers the SI standard for data storage have used up their 2 pence, and are intellectually bankrupt.

Seriously now... just because a 'standard' is a standard, without application, it is only a confusion.

Take NASA for instance. They moved to SI, and it cost them the Mars rover -- and they're frickin' rocket scientists. :D

Parent

RE: Thats ridiculous

By Araxen on 6/28/2006 5:22:18 PM , Rating: 1

Technially you are wrong.
1gb = 1,073,741,824 bytes (http://en.wikipedia.org/wiki/Gigabyte)
1mb = 1,048,576 bytes
(http://en.wikipedia.org/wiki/Megabyte)

Parent

RE: Thats ridiculous

By TomZ on 6/28/2006 5:26:41 PM , Rating: 3

Your links, unfortunately, contract your assertion, since they explain that both powers-of-two and powers-or-ten definitions are commonly used.

Parent

RE: Thats ridiculous

By xdrol on 6/29/2006 5:28:05 AM , Rating: 2

Check your own link, the powers-of-two meanin of mega/kilo/etc. is only commonly used , the only correct, defined-by-standards meaning of them is the powers-of-ten version.

If you want to use powers-of-two prefixes go kibi/mebi/etc.

International standards, people, use them. Suing a HDD company because it uses the SI international standard correctly, is a way of saying "I'm an ignorant idiot".

Parent

RE: Thats ridiculous

By PedroDaGr8 on 6/29/2006 8:30:07 AM , Rating: 2

Actually, in this case, the words were adopted for the power of 2 system, LONG before the power of 10 system came out. I remember when HD manufactures started releasing their drives in the power of 10 system. I seem to remember it was AFTER that happened that we had the new mebi and what now come out to try to ease confusion.

Parent

RE: Thats ridiculous

By Tyler 86 on 6/29/2006 11:13:20 AM , Rating: 2

1024^2 (1,048,576) & 1024^3 (1,073,741,824) came first, and are implicitly the 'true' way to measure 'byte' based storage, due to it's binary representation...
In hexadecimal, a byte is represented by a pair of hexadecimal digits.
Each hexadecimal digit, 8 "1000", 4 "0100", 2 "0010", and 1 "0001", in binary each represent a lone set bit and the rest zeros. They are sort of the 'building blocks' for the 'powers of' binary.
The hexadecimal representation of a kilobyte is "0x00000400", a megabyte is "0x00100000", a gigabyte "0x40000000". They each are 1 single set bit.
The difference between a kilobyte and a megabyte is the binary function left- or right- shift 10 bits - the same for the difference between a megabyte and a gigabyte .. a gigabyte and a terrabyte, etc...
Decimal '1,000,000' in hexadecimal is '0x000F4240'. Clearly not a single bit. '1,000,000,000' in hexadecimal is '0x3B9ACA00'. The difference between them is multiply or divide by '0x03E8', or 1,000.
A bit is 1 eighth of a byte, so when you're dealing with 'bits', it's only 1, so powers of 1,000 are vaguely more acceptable for measurement for human use - although it's still not entirely appropriate when considering their binary representation on computers.

Parent

RE: Thats ridiculous

By RMTimeKill on 6/30/2006 2:24:31 PM , Rating: 2

Hello,

I have been a reader for a long time and this post just made me have to sign up…

It is funny to see two sides, whom are both correct in their math (but not always their logic) arguing that the other is wrong…

So now I am going to condense all the decimal vs binary arguements of the whole thread into one post, in simpler terms.

FOR THE RECORD, neither of you are wrong.

Also

Mega, Giga, Tera, etc do mean in powers of 10 when referencing decimal, but it is IMPOSSIBLE for them to reference 10 in binary, thus, the conversion factor!!

(I see the power of 10 people going, WTH? Whats this idiot mean impossible?!) Well…its pretty basic actually… a “bit” only can count “2” numbers, 0 and 1, so how could it ever be counted to 10 in order to have a power of 10? Same with a byte, which is merely a string of 8 bits, there is still only “8” bits per byte and so on and so forth continuing on up the binary counting chain 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, which 1024 bytes equals 1 Kilobyte, 1024 Kilobytes = Megabyte, 1024 Megabytes = 1 Gigabyte

Now if we did it by powers of 10 as the HDD companies do then

1000 bytes = 1 Kilobyte, 1000Kilobytes = 1 Megabyte, 1000Megabytes = 1 Gigabyte.

So, with that being said, using conversion from decimal to binary

100GB Decimal = 97.6GB Binary

Just as

97.6GB Binary = 100GB Decimal.

After all that, what’s my point? My point is neither one are wrong, and no one was being deceitful, it simply points out that we have IT degree’s for a reason… because the common non-IT consumer is too stupid to understand what they are reading. It can all be explained in the math, its just simple conversion. Just like Fahrenheit to Celcius.

While Decimal may have been around since the begining of math, binary has been here since the begining of modern computers. They are two completely different maths that happen to use the same prefixs, and you can convert back and forth between them, just like everything else in math.

Also for the record, this Mebibyte crap has been around how long? I have never seen it on a single spec sheet or piece of IT equipment, hell, MS word 2k3 doesn’t even recognize it as a real word!! Standard my arse…

JB

Parent

correct case

By BikeDude on 6/29/2006 2:42:12 PM , Rating: 2

k = metric kilo (1000)
K = more modern kilo :) (2^10)
m = milli (.001)
M = Mega (in our world, that's 1024K)
b = bit
B = Byte

Hence, 1mb is really, really small... I dunno what "gb" is, but it looks messy. Better stick with GB.

The nonsense started in 1998 is just that; nonsense. "mebibyte"? Yes, when pigs fly.

--
Rune

Parent

RE: correct case

By Tyler 86 on 6/29/2006 3:04:58 PM , Rating: 2

It's too bad the average acronym or abbreviation suffers from 'case insensitivity', the computers have had to really endure some changes what with localization and all...

Don't take casing for it's supposed representation on eye-contact from someone you don't know, eg; an ISP or HDD manufacturer - if it's to be factored in, inquire about it's actual long-hand value.

Parent

RE: correct case

By TomZ on 6/29/2006 4:14:15 PM , Rating: 2

quote:
k = metric kilo (1000)
K = more modern kilo

I don't think this is common usage. I think that most folks recognize 1kB = 1KB; however, as the article points out, 1kb typically means 1 kilobit, as you stated.

Parent

RE: correct case

By Tyler 86 on 6/29/2006 4:43:54 PM , Rating: 2

Yeah, K or k for kilo... the case insensitivity issue is more along the lines of b(it) versus B(yte)... it'd be 'better' if we used the C++ term of char(acter) to represent 8 bits, then with 1024 MCs the GC would be off the hook! but yeah... so that wouldn't work...

Parent

RE: Thats ridiculous

By Dubb on 6/28/2006 4:00:02 PM , Rating: 2

soo...the people "burned" by this get some crappy software, and the lawyers get a ton of cash?

Parent

RE: Thats ridiculous

By marvdmartian on 6/28/2006 4:08:59 PM , Rating: 2

That's the nature of the class action "beast". Unfortunately. Typical.....lawyers laughing all the way to the bank, those who (barely, in this case) got screwed get screwed by the lawyers too!

Parent

RE: Thats ridiculous

By hwhacker on 6/28/06, Rating: 0

RE: Thats ridiculous

By jon1003 on 6/28/2006 4:12:05 PM , Rating: 2

Yep, we get crappy software and lawyers get cash. How about we get cash and lawyers get 1,000,000 software licenses? They can ebay them or something.

I have a few WD drives, but I'm not wasting time on DLing this software, and I bet few will or ever use it. There are free or opensource apps you can use if you needed backup software.

Parent

RE: Thats ridiculous

By brystmar on 6/28/06, Rating: 0

RE: Thats ridiculous

By TomZ on 6/28/2006 4:44:18 PM , Rating: 2

Doesn't anybody recognize sarcasm any more?

Parent

RE: Thats ridiculous

By Randalllind on 6/29/2006 4:22:48 PM , Rating: 2

It has been like this for years and I always thought it was wrong myself. Everyone does it not just WD. The awnser I always got was it was because of how Windows detects it. Gib instead of GB.

Parent

RE: Thats ridiculous

By Feckless Plaintiff on 7/1/2006 11:02:34 AM , Rating: 2

For readers of a supposedly technical forum, there's an immense amount of willful ignorance displayed here. Welcome to the 21st century, and persist in using ambiguous overloading of acronyms in measurements to your own confusion.

Anyone with 0.5 of a brain can look up GiB whenever they encounter it for the first time and perhaps think "nice, now I don't have to deal with the confusion and ambiguity", and so avoid technical errors, not just lawsuits that take money from technologists by exploiting that confusion.

Those of you who persist in using GB to mean 2^30 and expect your readers to get it due to context, how do you do calculations for gigabit network rates transferring gigabytes of data? It's clear if you stick to base 10. If you arbitrarily introduce base 2 at some points and not others, and don't clearly distinguish the two, you create confusion and error.

BTW, Linux documentation has been using GiB for some time now.

Parent

indeed ridiculous

By Wamingo on 6/29/2006 10:16:10 AM , Rating: 2

oh noes! My harddrive is defunct! it's missing 24 bytes per kilobyte! LAWSUIT!

Ignorant consumers should not be allowed to buy computers in the first place!
Harddrive manufacturer's has done nothing wrong.
If anyone, it would be MICROSOFT's and other OS's job to inform the user how the harddrive and definitions are being used.

If WD writes 80GB on the drive it means:
80GigaBytes = 80 billion = 80,000,000,000 bytes
80 billion / 1024 = 78,125,000 KIBI-bytes
80 bn / 1024^2 = ~76,294 MIBI-bytes
80 bn / 1024^3 = ~74.5 GIBI-bytes
__NOT__ 75.5 "GIGA" BYTES!!!

Kibi is a contraction of kilo-binary-byte. Go figure.
http://en.wikipedia.org/wiki/Kibibyte

When you read KB in windows and most places it means Kibi, not Kilo like most people believe, therefor the blame is on the OS, not the harddrive.

Memory modules also use mega and giga, so expect them to be the next in this lawyer's scam, as 512Mega = 500Mibi bytes...

RE: indeed ridiculous

By TomZ on 6/29/2006 2:34:11 PM , Rating: 2

Sorry, wrong. You can read about SI, and kibibyte all you want in wikipedia, but it doesn't change the industry-accepted usage that 1kB = 1024 bytes.

I've worked in this industry for 15+ years, and never once, until today, heard of this term. It is just not accepted by the industry, period.

I can't believe you accept the marketing B.S. from the HDD manufacturers of choosing to define their own units of measure in order to inflate their apparent HDD capacities. As has been noted by other posters, it wasn't always like that, and along the way, HDD manufacturers decided that this would provide a marketing benefit.

Parent

RE: indeed ridiculous

By agnot on 6/30/2006 1:38:22 AM , Rating: 2

quote:
Sorry, wrong. You can read about SI, and kibibyte all you want in wikipedia, but it doesn't change the industry-accepted usage that 1kB = 1024 bytes.

Huh? Every hard drive manufacturer uses 1KB = 1000 bytes, so I fail to see how you can possibly conclude that the industry standard is 1024 bytes.

Memory uses 1kB = 1024 bytes because DRAM modules are inherently binary. The number of cells in DRAM is always a power of 2. On the other hand, there's nothing inherently binary about the capacity of a hard drive. There's no universal law that says the number of platters, tracks, sectors, etc. in a hard drive has to be a power of 2, so I think it's perfectly acceptable, and indeed preferable, to use the traditional meaning of kilo = 1000. The whole 1024 thing was just an approximation anyway.

PC3200 memory is exactly 3,200MB/s = 3,200,000,000 bytes/s. 3GHz is exactly 3,000,000,000 Hz. A 100GB hard drive is exactly 100,000,000,000 bytes. About the only thing that uses the binary definition are memory modules. It's more of an exception rather than the rule.

Parent

RE: indeed ridiculous

By TomZ on 6/30/2006 9:15:13 AM , Rating: 2

quote:
On the other hand, there's nothing inherently binary about the capacity of a hard drive.

Sector size on all HDDs is a power of two. The capacity is the number of sectors times the sector size, which is a power of two. Only up until HDD marketing made the brochures and realized they could boost HDD apparent capacity by redefining the units. Note that hard drives were not marketed this way until just a few years back. Prior to that, capacities were in powers-of-two MB and GB. This is the reason for the lawsuits against the HDD manufacturers.

quote:
PC3200 memory is exactly 3,200MB/s = 3,200,000,000 bytes/s. 3GHz is exactly 3,000,000,000 Hz

That's true, and is because they are based on MHz crystals. But memory storage on computers does not use SI prefixes . Never has been.

Parent

RE: indeed ridiculous

By Tyler 86 on 7/1/2006 11:36:23 AM , Rating: 2

Yeah, Hz means hertz; 'cycles per second'.

Clocks are for all intents and purposes considered 'analog' components. Sometimes they're wrong. Even without EMI reducing spectrum-spread clock technology, a typical Pentium 4 CPU clock can vary by ~10Mhz under normal operation due to the core clock, sometimes even more.
Now, honestly, that's not a big deal for a CPU rated in Gigahertz - but it is Ten Million Hertz.

'Digital' is synonymous with 'binary', and the common native storage unit for it is the byte. Digital storage is measured in powers of 2 because of lack of more than 2 'digit' symbols to represent numbers in.

Parent

RE: indeed ridiculous

By Tyler 86 on 7/1/2006 11:40:42 AM , Rating: 2

er... damn, went off on a tangent there and left off the summary for analog clocks...

The point was analog clocks are in no way related to 'digital', although you can have a digital clock, and true digital crystal-less clocks do exist...

They don't work with integers - they have no definite value.
They operate on fractions, on *real* values, with infinite precision, and measurable inprecision.

Parent

RE: indeed ridiculous

By Tyler 86 on 7/1/2006 11:47:12 AM , Rating: 2

... still didn't finish that off, did I?

Basicly, anything analog can be measured by any standard, because it has no minimum increment.
You don't increase it 'one' hert, or 'one' bit, you increase it by an indefinite fraction of it's current value...

For this reason, the infinite value of 'pi' or 'e' cannot be instantaneously derived in any digital form - although if you wrap measuring tape around a ball, you just beat the hell out of a CPU - you're analogously representing pi in all it's majesty.

With digital data, you have to have definite values in the form of binary. With analog data, you can do whatever the hell you want with it - but you also run into inaccuracy, and alternate dimensions... :P

Parent

RE: indeed ridiculous

By agnot on 7/1/2006 12:17:34 PM , Rating: 2

quote:
Clocks are for all intents and purposes considered 'analog' components.

When it comes down to it, everything is "analog". Something is "digital" only when we choose to interpret the analog values in a discrete fashion. A transistor is an analog device, but it "becomes" digital when we associate it with a clock signal and sample the voltage/current at appropriate times. Transistors can and are used in analog circuits as pure analog devices.

In any case, I don't see what this has to do with the discussion at hand, which is whether kilobyte should mean 1024 bytes or not. The point of bringing up bandwidth and clock speed numbers is that other areas in the computer industry use the traditional meaning of kilo = 1000.

quote:
'Digital' is synonymous with 'binary'

I would say "digital" is synonymous with "discrete". In computers, that usually means binary, but "digital" does not have to be "binary", e.g. digital clocks vs analog clocks.

Parent

RE: indeed ridiculous

By Tyler 86 on 7/3/2006 4:18:53 PM , Rating: 2

Everything is indeed analog at a low-level, but at a particularly high level of definite (1) versus not-definite (0).

In comparison, two fine grains of sand in a low-accuracy digital tray measuring grams goes absolutely unnoticed, effectively, to the digital meter, those two grains of sand do not exist, while ten thousand definantly registers -- on an analog scale, no matter how minutely two fine grains of sand weighs, it affects the meter on the scale by a definite (albeit definitely minute) amount.

Compare an analog speed-dial versus a digital speedometer; the digital meter will never be able to represent speed as accurately as the analog dial.

Although, with analog measurement also comes inaccuracy.
Most of the time the digital reading of scales is just an interpretation of an analog meter with limited precision. Some round up, some round down, some with a value somewhere inbetween.

That is the basis of the binary digital system; what registers, and what doesn't.

Digital clocks are fundamentally different in this way; they absolutely tick, without rise or fall delay, instead of pulse as today's clocks - although many components today rely on the rise and fall, which can be digitally emulated to high precision, as well...

Although you can rate them in Hz, cycles per second, they have absolute ratings measured in distance... from which Hz can be alteratively derived from environmental factors without ever actually ticking...

Digital is a high-level concept of maximum and minimum absolute discrete analogous values, although you can have multiple discrete analogous values that are not simply an absolute - you can have multiple dimensions, or more than 2 absolutes per dimension.

Bits tend to be the ones 'measuring themselves' in binary, and thus powers of two are the only true effective digital measurement scale.

Analog values can be measured in infinite precision, so you can measure them from any power - infinite divisibility.

You can not have 'half of', or 'one third of' a binary digit.

You can have half of a clock cycle.

Parent

RE: indeed ridiculous

By agnot on 7/4/2006 10:51:19 AM , Rating: 2

There is not a one-to-one relationship between "digital" and "binary". Binary is convenient given our current technology, but that's not to say that it will always be the case.

Take a look at multi-level cells for flash memory: http://www.memorynet.co.uk/info/faq-single-layer-c.... Each cell can store four states: off, 1/3 charge, 2/3 charge, and full charge.

Parent

RE: indeed ridiculous

By Tyler 86 on 7/6/2006 1:41:12 PM , Rating: 2

Analogously, those states would be 0.000, ~0.0101, ~0.1010, and 1.000.

However, it's true digital values for each state are absolutely 00, 01, 10, and 11...

It's just a digital binary binary principle of 1/11, or 1/4th, but no actual fraction of state exists. When you digitally quantify something, you assign each state a value, that value is expressed as a binary integer due to it's definity - it is not a 'real' number.

Parent

RE: indeed ridiculous

By Tyler 86 on 7/6/2006 1:41:55 PM , Rating: 2

-- as opposed to true analogous 'real' numbers expressing the curve of pi.

Parent

RE: indeed ridiculous

By Obadiah on 7/6/2006 4:16:20 PM , Rating: 2

Your hands aren't waving fast enough to keep all that BS in the air.

"true digital value," "definity" - You are just making up terms.

The etymology of the term digital is clear - it comes from the word "digit" for fingers and toes. Since humans naturally have 10 of each kind of digit, any sort of "true digital value" would be base-10 not base-2.

Parent

RE: indeed ridiculous

By agnot on 7/1/2006 1:02:59 PM , Rating: 2

quote:
Sector size on all HDDs is a power of two. The capacity is the number of sectors times the sector size, which is a power of two.

Yes, sector sizes are, but the number of sectors per track, number of tracks per platter, and number of platters per disk are not powers of two, and hence the total capacity of the hard drive is not a power of two. Consumers care about the total capacity, not how big each sector is (if they even know what a sector is). If hard drive capacities were powers of 2, hard drives would only come in sizes of 64 GiB, 128 GiB, 256 GiB, etc.

quote:
That's true, and is because they are based on MHz crystals. But memory storage on computers does not use SI prefixes . Never has been.

The correct solution to all of this is for everyone to follow the standards, not for the sake of following a standard, but because it makes things easier/clearer . What's easier, to see "Free space: 12,219,582,624 bytes" and say "I have 12.2 GB of free space", or "Ummm, I have 10 GB of free space, or thereabouts...I guess...hold on, let me get my calculator out."?

I mean, why go against the standard and force hard drive manufacturers to use kilo = 1024, when the standards-compliant solution of having the OS report kilo = 1000 would be simpler to implement (do an OS update vs. having to reprint all retail boxes). It's simpler and technically correct.

Or, like someone already mentioned, have hard drive manufacturers report both values, e.g. 120GB/112GiB and have the OS change "GB" to "GiB" so you see 112GiB in the OS. Or have the OS report both values.

In any case, the most incorrect solution would be to force hard drive manufacturers to change 120GB to 112GB. It's wrong, and it just makes things inconsistent. Take for example bandwidth values. Can you transfer 100MB of data in one second with a bandwidth of 100MB/s? If you use the binary definition, the answer is "no" since 100MB/s is and always will be 100,000,000 bytes/s while 100MB with the binary definition is 104,857,600 bytes.

If you follow the standards, then you can say that you can transfer 100MB of data in one second, but you cannot transfer 100MiB of data. Following the standards makes things consistent and it makes things less confusing (assuming everyone is following the standard). It's just plain better. Use it .

Parent

RE: indeed ridiculous

By Tyler 86 on 7/3/2006 4:22:46 PM , Rating: 2

quote:
hold on, let me get my calculator out

The computer is the calculator. Let it tell you how many gigabytes you have - in it's native format, converted to your decimal format.

If you want it to read in billions of bits, don't redefine the byte - come up with your own measurement system.

Parent

RE: indeed ridiculous

By Tyler 86 on 7/3/2006 4:53:27 PM , Rating: 2

As for 100MB/s at 100MB/s, yes you can transfer 100*1024^3 bytes/sec at 100MB/s...

The problem is, 100Mb ethernet is not 100 Mega Byte ethernet, it is 100 Mega Bit ethernet.

Bits are a true base unit, and can be measured in thousands, powers of 2, powers of 1024, or whatever.

One bit is currently one eighth of a byte.

The HDD manufacturers aren't refering to how many bits are on their drives when they state GB, they're referring to bytes, and they're not using an apropriately measuring it.

Same with some ISPs, although modern ISPs are getting much better about it, and actually sell 2MB/s lines that are capable of transferring 2 raw megabytes per second - excluding packet overhead.

Parent

RE: indeed ridiculous

By Tyler 86 on 7/3/2006 4:54:22 PM , Rating: 2

Also, I'm not 100% certain of this, but I believe 100 megabit ethernet actually measures megabits as 1024^3...

Parent

RE: indeed ridiculous

By agnot on 7/3/2006 6:57:33 PM , Rating: 2

No, 100 megabit/s is 100,000,000 bits/s.

Parent

RE: indeed ridiculous

By agnot on 7/3/2006 6:56:31 PM , Rating: 2

quote:
As for 100MB/s at 100MB/s, yes you can transfer 100*1024^3 bytes/sec at 100MB/s...

The problem is, 100Mb ethernet is not 100 Mega Byte ethernet, it is 100 Mega Bit ethernet.

The 100MB/s (100 megabytes/second) was just a figure I used as an example, it had nothing to do with ethernet. And you are wrong, 100MB/s = 100,000,000 bytes/second, not 100*1024^2 bytes/second.

As a more concrete example, take DDR400 memory. It is 64 bits wide, i.e. 8 bytes wide, clocked at 200Mhz double data rate, or 400 MT/s (400 megatransfers/s). That's 400,000,000 transfers/second, since 200Mhz is 200,000,000 cycles/second. Each transfer is 8 bytes, so you come up with 3,200,000,000 bytes/second, or more commonly 3,200MB/s or 3.2GB/s.

If you use 1GB = 1024^3 bytes, you cannot transfer 3.2 GB of data in one second with DDR400 (3.2GB/s bandwidth), which is a counterintuitive result of having different definitions for the prefix giga.

To resolve the inconsistency, you either have to define 1GB = 1,000,000,000 bytes, or you define 1GB/s = 1024^3 bytes/second. If you think about it, the latter approach will essentially mean that 1KHz = 1024Hz (e.g. DDR400 would have 3,052 MB/s bandwidth under the new definition, and with 8 bytes/transfer, that means 381.5 megatransfers/second, which accounting for double data rate, leads to the ludicrous conclusion of 190.7 MHz = 200,000,000 Hz, i.e. essentially 1 MHz = 1,048,576 Hz = 1024*1024 Hz).

The only logical way to resolve the inconsistency is to define 1GB = 1,000,000,000 bytes. Believe it or not, there actually are good reasons for coming up with the standard. It's not like somebody decided one day to arbitrarily change things.

Parent

RE: indeed ridiculous

By Tyler 86 on 7/6/2006 3:18:51 PM , Rating: 2

Bits are in this gray area - they're a true base unit when taken out of digital context.

However, you cannot take bytes out of context.

Using powers of 2 does not have a damn thing to do with hard drives 'only' coming in sizes of 64 GB, 128GB, etc...
It would if they only increased or changed by powers of two, but we're only talking about measurement.

They can be 11GB to 33GB to 69GB to 711GB to whatever...

Using powers of 2 IS about being clearer and easier.

Filesystems from the begining to have a minimum storage unit for uncompressed data - typicly ranging from 512 (bitmap of 4 bits) bytes to 4096 bytes, or 4KB (12 bits).
RAID stripes are typicly between 16384, or 16 kilobytes (14 bits), to 65536, or 64 kilobytes (16 bits).

These are for a number of reasons; security, mapping, and performance - all done in binary.

Your usable quantities are all powers of 2 - using a power of 2 to group them only makes sense.

This is related to the argument Microsoft makes with .NET, and Sun makes with Java.

At first, it was not at all about efficiency, it was about security, management, and simplicity.
Then they realised it's ridiculous to ignore low-level efficiency, and performance.
Now, Java and .NET even have their own 3D graphics interfaces, and have subsets of heavily optimized native machine assembly routines.

If you're going to hide them from the user, don't redefine the true industry's standard, make up your own.

That's why they use 'Transfers' or just Bits.

Say 64 billion bytes, not 64 GB*, G means Billion!

Everything that operates on these storage devices doesn't give a damn about fractions inbetween these set power of 2 quantifiers -- with the exception perhaps of optical drives such as CDs & DVDs, due to their analog relationship with audio and finely detailed media characteristics.

Let GiB and MiB stand for multiples of 1,000 - leave the damn GB and MB alone.

Parent

RE: indeed ridiculous

By Tyler 86 on 7/6/2006 3:22:48 PM , Rating: 2

Also, Megahertz and Gigahertz are not storage, and are not in question.

They are analog values, and can be represented any which way but wrong, and be right.

Parent

RE: indeed ridiculous

By Obadiah on 7/6/2006 4:38:08 PM , Rating: 2

They are analog values, and can be represented any which way but wrong, and be right.

Every which way but loose, man.

You are a real hoot! What are you a first year CS student?

Parent

THIS IS NOT A FORMATTING ISSUE.

By namechamps on 6/28/2006 5:52:07 PM , Rating: 1

THIS IS NOT A FORMATTING ISSUE.
THIS IS NOT A FORMATTING ISSUE.
THIS IS NOT A FORMATTING ISSUE.

(Maybe if I write it 3 times people will read it, maybe not).

Everyone claiming that HD companies are doing nothing wrong and this is a windows issue are COMPLETELY wrong.

The issue is that computers measure a gigabyte as
1024 Megabytes
or
1024 * 1024 Kilobytes
or
1024 * 1024 * 1024 Bytes.

To make their hard drives look bigger EVERY HD company "measures" their drives as
1 GB = 1000 MB = 1000 * 1000 KB = 1000 * 1000 * 1000 bytes.

This is simply wrong.

If you buy a "80 GB drive" it really is 75 Gigabytes BEFORE FORMATTING.
If you buy a "500 GB drive" it is 465 GB BEFORE FORMATTING.

Now after formatting the OS will use up some space but not much so your 75GB drive may have 72GB space and your 465GB drive has 459GB of space. That is not the issue.

The issue is simple they are selling a 75GB drive as 80. They are selling a 500GB drive as 465GB. Basicly because of the faulty math they oversize their drive by 7%.

Imagine how pissed you would be if I started a new memory company and sold 1GB dimms. They are a great deal and u buy a pair only to find out ther really are 953MB each. I say sorry I count a 1GB as 1000 * 1000 * 1000 not 1024 * 1024 * 1024.

All HD companies do this and it would be nice if they would simply stop. When drives were small the difference was minor however on say a 1TB drive you lose nearly 80GB due to this "rounding".

It is simply stupid. All HD companies should simply sell drives based on the real capacity (in 1024^3). I am sad the lawsuit is so small. Hopefully the next one will be larger. The truth is that companies rarely stop doing the wrong thing until it cost them real $$$. If Seagate gets sued for 22 million and it affects their stock price you can bet all the companies will simply start measuring correctly. $500K plus some free software will not convince them to change.

If everyone measures correctly then nobody gains an advantage, and the consumer is being told the truth.

THIS IS NOT A FORMATTING ISSUE.
THIS IS NOT A FORMATTING ISSUE.
THIS IS NOT A FORMATTING ISSUE.
(3 more times just in case. $1 says someone will post that this is a formatting issue anyways).

RE: THIS IS NOT A FORMATTING ISSUE.

By 13Gigatons on 6/28/2006 6:28:57 PM , Rating: 3

It's just a formating issue :) :) :) :)

Parent

Système International d'Unités

By Tyler 86 on 6/29/2006 4:15:14 PM , Rating: 2

The French didn't realise at the time that the Holy Decimal System is not the be-all end-all of standardization when they came up with their standards, they pretty much just said 'We are ze french, and we are changing ze definition. Ten is ze number of fingers on deu hans so, and now everything is ze ten... We do not care, Hau hau. Now we take ze 4 hour lunch break.'

The '1024' in decimal is 10 physical places in binary.
That means '1024' is '10000000000' in binary. It's not 9 places like a million, it's 10 places, and for fairly good reasons... for one you don't have 2s in binary, you just have to trust me, it's an intelligent representation.

People more intelligent than the those at SI developed these methods of storage with these names for a friggin' reason.

The SI is overstepping it's "authority", which is in itself an oxymoron, when it re-defines units of data storage.
They say whatever they want, and expect people to adhere to it.

We of the world, meaning everywhere except france in this case, don't work that way.

Hell, even in france, french programmers don't run around saying, "pas mebibit! pas mebibit!"... They know what the hell a megabyte is supposed to be.

Parent

RE: Système International d'Unités

By Le Québécois on 6/30/2006 3:19:18 AM , Rating: 2

I know i'll probably get a bad rating for saying this but on this one you're right and wrong at the same time.

First of the French came whit this system while they were the biggest Empire (hence around Napoleon frame of time). Like many empire tends to do when they are an empire is to force their rules on other peoples ( a little bit like the US is trying to do now and then ). If you thing about it the SI is almost perfect in every way. Take the freezing temp of the water, in SI "celcius" it's 0, why because water is one of the most basic element surrounding us, in the "english" system I don't even know at what F the water freeze. Another exemple, density, the density of water is 1 g per cm^3. To my knowledge only a couple of nation don't use it complitly and those happen to be english speaking nation.

Now to the you're right point. 1024 does in fact come from the BIT aspect of computing where 10000000000 is in fact 1024. Quite confusing to someone who don't know jack about computer or binary.
That's why in french we( here I say we because I'm french Canadian, please not to be confuse with French, it's almost an insult sometime ) use two very different words to talk about bit and byte to not confuse them. Basicly for byte in french we use "octet" so we never confuse Kb and Kb...

Here's a quote from wikipedia:
On most computers the smallest unit of memory addressing—or byte—is 8 bits, so the terms "byte" and "octet" are often used interchangeably. However, the size of a byte is determined by the architecture of a particular computer system: some old computers had 9, 10, or 12-bit bytes, while others had bytes as small as 5 or 6 bits. An octet is always exactly 8 bits. As a result, computer networking standards almost exclusively use "octet" to refer to the 8-bit quantity.

So excuse me if I'm rude but let's come back to the part where you are wrong. First in french we use Mo of Go for megabyte so it is not to be confuse with bit. Second nobody is wrong here. In fact in a perfect world where a normal person would not pays 50$ to Bestbuy so they can have their new ram stick install evereybody would know that in computing M or G means x1024. But we don't live in a perfect world and the company take advantage of that fact by missleading consumer.

Now to the part where you say WE OF THE WORLD...I hope by that you mean the US, England and their colonies (or former colonies) because open you eyes for a sec and you'll see that everywhere else they use the SI.

So yes in France you are right, they go around saying pas mebibit (what ever you meant by that because mebibit is not even a word). They say MegaOctet or GigaOctet.

Hope this open your eyes a littles and teach you that there is a world outside the USA.

Probably going to get a lot of bad rating from this but at least I have made my point...and if you don't belive me do a little searching around.

Why did I see all this, probably because I'm Quebecois and I live every day in this raging war between the SI and the "english(USA)"system.

Parent

RE: Système International d'Unités

By TomZ on 6/30/2006 9:18:52 AM , Rating: 2

quote:
pas mebibit (what ever you meant by that because mebibit is not even a word).

http://en.wikipedia.org/wiki/Mebibit

Parent

RE: Système International d'Unités

By Le Québécois on 6/30/2006 1:55:03 PM , Rating: 2

Thx, I'll be less dumb when I go to bed tonight(a Quebecer idiom) but never the less it isn't a French word at all.

I have never used it or heard it in all the time I did study binary in my electronic course( which I did in french ).

Parent

RE: Système International d'Unités

By Tyler 86 on 7/3/2006 3:33:04 PM , Rating: 2

Many measurements in non-english speaking countries exist too.

'Metrication' implements purely new standards of measurement, the meter, the gram, etc...

The issue is that they're redefining bytes, and even octets.

The SI overstepped it's supposed authority on this one, unless you're telling me they came first -- and had a concept of data storage in the Napoleonic era...

Parent

RE: Système International d'Unités

By Tyler 86 on 7/3/2006 3:43:23 PM , Rating: 2

On a note of agreement, the byte is too close to the bit in it's abbreviation; I proposed the C++ term 'character', as this represents a minimum storage unit of binary data, while leaving it open to changing from 8 to 16 to however many bits are needed.

An octet is very definitively 8 of something, commonly bits. I have an octet of non-opposing fingers... Supposedly, a bit could be re-interpreted, from it's origin, b inary dig it , as just a digit. I have 5 digits on my hand...

When nations accepted the SI units as their national standards, although they converted from their previous method of measurement, they didn't overwrite their originals.

They also defined the decimal system to work by representation on paper, and in a human mind - ignoring binary. This 'holy' decimal system was the one system to rule them all.

As you are implicitly aware, there had to be units measurement opposed to the metric system previously in every nation that 'converted', and I'm certain that's a heck of a large number of non-metric measurements.

Parent

RE: Système International d'Unités

By Tyler 86 on 7/3/2006 5:04:55 PM , Rating: 2

Oh, and if you really think about it, like, break out the calculator, the laws of physics, and study existentialism, binary and powers of 2 are even more perfect in every way than decimal and powers of 10. O_o

Parent

RE: Système International d'Unités

By Tyler 86 on 7/3/2006 5:06:16 PM , Rating: 2

^ and tack on programming.

Parent

RE: THIS IS NOT A FORMATTING ISSUE.

By Motley on 6/28/2006 6:46:05 PM , Rating: 2

It would be better if people stopped reporting 1 MB as 1024^2 (Or K as 1024), or came up with their own name rather than using a bastardized metric system (I do believe the correct term is binary deciplex=1024, or binary dodeciplex=1024^2).

Kilo=1,000.
A kilogram is 1000 grams.
A kiloton is 1000 tons.
A kilometer is 1000 meters.
A kiloliter is 1000 liters.
A kilowatt is 1000 watts.
A kilobyte is 1000 bytes.

To say a kilobyte is 1024 is wrong. The metric system was in use LONG before computers. Computers don't "measure" anything. It's easier for computers to deal with numbers that are multiples of 2, but that excuse only shows laziness.

Parent

RE: THIS IS NOT A FORMATTING ISSUE.

By almvtb on 6/28/2006 9:55:32 PM , Rating: 2

There are real standard prefixes for Binary but a gibibyte just sounds stupid to me http://en.wikipedia.org/wiki/Binary_prefix
So I am going to stick with the bastardised SI prefixes

Parent

RE: THIS IS NOT A FORMATTING ISSUE.

By chekk on 6/28/2006 10:12:56 PM , Rating: 1

Thank you for helping to clear away all this nonsense.
To reiterate for everyone for whom this is still unclear:
While the difference between 1000 and 1024 bytes is deceptive, it is NOT the HDD manufacturers fault.
For example:
Mega is an SI prefix meaning 1 x 10^6 not 1024^2.
I agree that the "mebi" prefix sounds dumb, but get over it, people.
For the last time,
kilo = 1000
mega = 1000000
giga = 1000000000

Move along, nothing to see here.

C.

Parent

RE: THIS IS NOT A FORMATTING ISSUE.

By Schrag4 on 6/29/2006 9:18:46 AM , Rating: 2

The only reason people get upset about this is because the HD manufacturers and Microsoft Windows use the same unit of measure (GB) to report your hard drive space, but they claim completely different numbers from each other. So, in my opinion, if GigaByte really means 1,000,000,000 bytes, then your OS is what's telling you the wrong number.

Mind you, the people getting upset are users who don't know much about computers but see 80 GB on the HD box and when they put it in only see 74 GB (or whatever). Seems like they have a right to complain, but perhaps their target is misguided...

Parent

Actually, we should all blame Toshiba

By AkaiRo on 6/28/2006 7:14:08 PM , Rating: 2

It's not really the hard drive manufacturers who started this. Years ago, Toshiba's 4x000 series notebooks were getting trounced in the speeds and feeds war against Compaq and IBM. They started referring to the capacity of the hard drives used in the 4x00 series notebooks using Million Bytes instead of MegaBytes to make it seem like you got more for the money than with the LTE and Thinkpad.

Besides, a 1GB DIMM gives 1 GigaByte of RAM, not a Billion Bytes of RAM.

Parent

RE: THIS IS NOT A FORMATTING ISSUE.

By lemonadesoda on 6/28/2006 9:30:37 PM , Rating: 4

It is very simple. 1024 <> kilo = 1000

The standard scientific index k is 1000. The wiki is helpful http://en.wikipedia.org/wiki/kilo as is http://en.wikipedia.org/wiki/Category:SI_prefixes

Since 2^10 = 1024 this is, technically, not a kilo byte. However, for reasons of popular simplicity in naming, it became a convention to use kilo as representing 1024 for binary data. A convenient shortcut.

Technically, at the scientific level, it is incorrect to have k represent 1024.

So the truth is WD is correct in their scientifically accurate naming. And companies that sell, e.g. RAM modules, are incorrect.

The legal case should have been thrown out.

The lawyers should be suing the memory manufacturers for selling RAM with a free extra%.

Ha. Ha. Doh.

Anybody that continues to argue the 1024 is simply a noodle.

Parent

RE: THIS IS NOT A FORMATTING ISSUE.

By TomZ on 6/29/2006 4:18:58 PM , Rating: 2

COMPUTER MEMORY MEASUREMENTS ARE NOT BASED ON SI
COMPUTER MEMORY MEASUREMENTS ARE NOT BASED ON SI
COMPUTER MEMORY MEASUREMENTS ARE NOT BASED ON SI
COMPUTER MEMORY MEASUREMENTS ARE NOT BASED ON SI
COMPUTER MEMORY MEASUREMENTS ARE NOT BASED ON SI

There, that should do it.

Your argument that 1 kilobyte = 1000 is wrong because you assume that computer memory adheres to SI, which it doesn't, and it never has, until some smart marketing guy/gal realized they could get a big capacity boost by changing the definition.

Parent

RE: THIS IS NOT A FORMATTING ISSUE.

By Tyler 86 on 6/29/2006 4:26:21 PM , Rating: 1

quote:
Anybody that continues to argue the 1024 is simply a noodle.

You are no programmer, so your comments suck, mon ami.

Parent

Haven't noticed anything with my Segates...

By Trisped on 6/28/2006 4:14:43 PM , Rating: 2

I haven't noticed any problems with my Segates not actually having 2^10 sizes, but I do lose space to the file systems. Does anyone count those, or are manufactures suppose to add extra space for those too?

RE: Haven't noticed anything with my Segates...

By TomZ on 6/28/2006 4:37:12 PM , Rating: 2

I think Seagate has the same issue, at least for some of their drives. I have a "74GB" Raptor that has a real capacity of only 69GB with the conventional GB unit of measure. So Seagate "shorted" me 5GB.

Parent

RE: Haven't noticed anything with my Segates...

By ghuoof on 6/28/2006 4:48:55 PM , Rating: 2

a seagate raptor?

Parent

RE: Haven't noticed anything with my Segates...

By TomZ on 6/28/2006 4:57:59 PM , Rating: 2

Sorry, duh, my brain's not working today. :o(

Parent

RE: Haven't noticed anything with my Segates...

By Tyler 86 on 6/29/2006 3:36:35 PM , Rating: 2

Yeah, the WD740GD Raptors @ 74 GB are actually 69 GB... and the older revision of the WD740GD Raptors had even ~200MB less.

But yeah, it's thousand versus thousand-twenty-four, same issue, with Seagate drives .. eg, '100GB' is actually less than 100GB, to the tune of only ~%95.37.

Parent

RE: Haven't noticed anything with my Segates...

By Tyler 86 on 6/29/2006 3:37:29 PM , Rating: 2

Woops, transposition + bad math.. ~%93

Parent

RE: Haven't noticed anything with my Segates...

By Master Kenobi

(blog) on 6/28/2006 4:45:56 PM , Rating: 2

It's not a sizing issue, its formatting. Anyone with a clue when it comes to hard drives knows that from an 80GB drive your likely to get 74ish GB of usable space, thats just how its always worked out. Drive manufacturers since the beginning of time have done this, theres always been a bit shaved off for various purposes. The real question here is how the heck this lawsuit actually made it into court, any judge worth his salt should have smacked them down when this got filed.

Side note, the grounds for this lawsuit, or lack there of, reeks of the raw deal McDonalds got a few years back because some dumb woman burned herself by spilling her coffee on herself, and then sued McDonalds because the coffee was "hot".

For this style of lawsuit I might as well sue McDonalds everytime I buy a large fry because if i dont get the exact same amount of fries quantity/weight, then I didn't get what I paid for each and every time. Oh yea, if that actually flies, I lost the rest of the minute bit of faith in the American legal system.

What a friggin JOKE 'nuff said.
/rantoff

Parent

RE: Haven't noticed anything with my Segates...

By TomZ on 6/28/2006 5:02:49 PM , Rating: 2

No, I don't think you're understanding the real issue.

My 74GB WDC (not Seagate!) Raptor has a capacity of 74,340,044,800 bytes. Their marketing calls that "74GB" because they define 1GB = 1,000,000,000 bytes.

But to most people, 1GB = 1024^3 = 1,073,741,824 bytes. Therefore, the "74GB" Raptor is actually only 69.2GB.

Therefore, I buy a 74GB drive and only get 69GB.

It's really not any different than buying a gallon of milk at the store, and then getting it home and finding small print saying that they define 1 gallon = 120 oz. instead of the normal 128 oz.

Parent

RE: Haven't noticed anything with my Segates...

By xdrol on 6/29/2006 5:30:48 AM , Rating: 2

Never anyone defined kilo/mega/giga as 2^x0. It is and was always 10^x. Whoever started to use them for powers of two should be blamed.

Parent

RE: Haven't noticed anything with my Segates...

By TomZ on 6/29/2006 9:38:37 AM , Rating: 2

1 kilobyte has always been defined as 1024 bytes since the early days of computers. Same for megabyte and gigabyte.

I'll admit it is not the best system, borrowing SI prefixes and re-defining them as powers-of-two, but it is such common usage that it won't change any time soon.

Parent

RE: Haven't noticed anything with my Segates...

By Tyler 86 on 6/29/2006 3:45:39 PM , Rating: 2

No one ever defined kilo/mega/giga to be 2^x0, in fact it is 10^x when dealing with meters or whatever...

Kilo- means 'thousand', from Greek 'Chilia-'.
However, 'Mega' just means 'Very Large'.

Welcome to English. My name is Smith, pronounced "Janofski", and the "Système International d'Unités" can go shove it up their french fannies.

See also; http://dictionary.reference.com/search?q=fannies
Note sub; [Perhaps from Fanny, a nickname for Frances.]

Parent

RE: Haven't noticed anything with my Segates...

By Tyler 86 on 6/29/2006 3:52:08 PM , Rating: 2

Oh, guess where Giga- came from? Greek as well, the first four letters of the word for giant, '???a?' (gamma iota gamma alpha), by it's sound directly to english as 'giga'.

The french made it 1,000,000 long after it's conception for gigawatts and gigabytes.

Parent

RE: Haven't noticed anything with my Segates...

By Tyler 86 on 6/29/2006 3:54:03 PM , Rating: 2

Argh, why'd I submit gigawatts... that's just the Back to the Future reference in Wikipedia... He wasn't working with gigawatts even, he was just 'emphasizing magnitude'...

Parent

RE: Haven't noticed anything with my Segates...

By TomZ on 6/29/2006 4:11:25 PM , Rating: 2

quote:
No one ever defined kilo/mega/giga to be 2^x0, in fact it is 10^x when dealing with meters or whatever...

I'm not talking about SI prefixes kilo, etc. I realize that k=kilo=1000 in that context. But in the context of bytes, 1 kilobyte = 1024 bytes.

Computer industry units of measures for memory are not based on SI .

Parent

RE: Haven't noticed anything with my Segates...

By Tyler 86 on 6/29/2006 4:49:13 PM , Rating: 2

The very prefix itself, kilo-, represents "one thousand", not "one thousand twenty four". Kilo is in the gray area.

Kilo came before SI, and strictly defined as one thousand.

Mega, Giga, etc.. are the ones bastardized by SI.

Giga- being the one in question here, representing billion instead of the industry's accepted 'giant' 1024^3 representation.

Parent

RE: Haven't noticed anything with my Segates...

By 05SilverGT on 6/28/2006 5:03:21 PM , Rating: 2

That's why McDonalds puts this warning on the Quarter Pounter.

+ Based on the weight before cooking 4 oz. (113.4g)

Someone I'm sure went after them saying that it actually didn't way a quarter pound cooked.

Parent

RE: Haven't noticed anything with my Segates...

By Lakku on 6/29/2006 4:43:14 AM , Rating: 1

In the case of McDonald's, I think people fail to realize just how hot the coffee was. I dislike idiotic lawsuits, however, the coffee was extremely hot, around 200 degrees, when coffee is usually around 125 to 140F. It caused third degree burns and she required skin grafting. She sought to settle for something like 20k. McD's refused so she took em to court. At court, they produced documentation showing McDonalds had had many complaints of bad burns for almost a decade, but did nothing about it. This is why she won and why it's important to know that in this case, McD's deserved what it got.

Parent

RE: Haven't noticed anything with my Segates...

By TomZ on 6/29/2006 4:22:13 PM , Rating: 2

Please don't confuse us with the facts. We just like to complain about lawyers and assume all lawsuits are for purposes of extortion. :o)

Parent

Why not PC manufacturers??

By othercents on 6/29/2006 12:22:54 PM , Rating: 2

So wouldn't PC manufacturers know that the drives where the wrong size? Since they knew and still advertised their computers as 80gb when it was actually 75gb shouldn't they be liable too? Maybe we will see more class action suits against anyone that builds computers.

Other

RE: Why not PC manufacturers??

By TomZ on 6/29/2006 2:29:14 PM , Rating: 2

I think the same liability might apply there, but it is may be a harder case to make, since they can just "pass through" and say that they simply advertised the manufacturer ratings.

Parent

RE: Why not PC manufacturers??

By Tyler 86 on 6/29/2006 6:05:24 PM , Rating: 2

Yeah.

Though I bet this can apply to ISPs... They aren't "passing through" the advertising of the internet from anyone.

I'd get a real kick out of it if they were "passing through" 384KB/s as 384,000 baud from the backbone at a university or something... That little bandwidth would make 5.25" floppy disk access speed seem like a real luxury. :D

Parent

RE: Why not PC manufacturers??

By Tyler 86 on 7/3/2006 4:56:44 PM , Rating: 2

Well, 5.25" floppy speed is actually quite high with buffering now, so nevermind... heh...

Parent

RE: Why not PC manufacturers??

By Tyler 86 on 7/3/2006 4:59:43 PM , Rating: 2

Since a 56k modem is 33.6K baud... DSL @ 384K baud is faster... but still, 'yesterday' they say bytes, and give bits... recently, looking around, they now accurately measure bytes, using powers of two as well.

Parent

RE: Why not PC manufacturers??

By Obadiah on 7/6/2006 5:24:56 PM , Rating: 2

Since a 56k modem is 33.6K baud.

No, a 56k modem (v.92) is at most 8K baud.

Parent

RE: Why not PC manufacturers??

By neothe0ne on 7/2/2006 2:06:35 PM , Rating: 2

Because PC manufacturers constantly write on spec sheets and on boxes that they cannot be held liable for "mistakes" in the specs. Such is written on HP/Compaq boxes, as well as (I'm assuming) eMachine boxes. This is also written online at Dell and other manufacturers, I'm sure.

And even if they were sued for these hard drive space concerns... lawyers win, hard drive manufacturer could care less, (ignorant) consumer is still screwed.

Parent

RE: Why not PC manufacturers??

By rushfan2006 on 7/5/2006 9:33:03 AM , Rating: 2

quote:
And even if they were sued for these hard drive space concerns... lawyers win, hard drive manufacturer could care less, (ignorant) consumer is still screwed.

That's just what you guys don't understand though. Through this entire 115+ comment thread back and forth on this issue. The "ignorant" consumer DOES NOT CARE. Its the old story, what you don't know can't hurt you (well in medical situations that's false of course..lol). The average consumer buying a PC for email, web surfing and maybe some digital picture work -- isn't gonna stress over the fact that her "80 gig" drive only really has "74 gig"...or whatever.

Its only us our crowd that cares...the system builders, the IT geeks, et al.

And believe it or not the average joe/joan consumer out numbers us vastly..thus companies like WD are slow to address the issue (if at all).

Parent

A worthless settlement for a worthless case

By Aesir on 6/29/2006 5:19:02 AM , Rating: 2

Whether you like it or not, and as stupid as the word gibibyte is, the OSes are the ones lying to you, as what they report as a Gigabyte is actually a Gibibyte. There are rules to how these (stupid) prefixes come about, believe it or not there ARE organizations in charge of this (IEEE and IEC come to mind. If you've not heard of them, leave before you embarass yourself). Just because general usage has become something other than the correct usage, does not mean a correct usage does not exist.

If you sued Microsoft for under reporting harddrive capacity you would have just as strong, and probably stronger, of a case, as technically the harddrive manufacturers are in the right (remember those committees that make the rules?) and Microsoft, Linux, Apple, etc. are all wrong.

RE: A worthless settlement for a worthless case

By TomZ on 6/29/2006 9:35:14 AM , Rating: 1

quote:
Whether you like it or not, and as stupid as the word gibibyte is, the OSes are the ones lying to you, as what they report as a Gigabyte is actually a Gibibyte. There are rules to how these (stupid) prefixes come about, believe it or not there ARE organizations in charge of this (IEEE and IEC come to mind. If you've not heard of them, leave before you embarass yourself). Just because general usage has become something other than the correct usage, does not mean a correct usage does not exist.

Wrong. This case is based on consumer expectations that a MB and GB are defined the way they expect - powers of two. HDD manufacturers choose powers-of-ten definitions in order to inflate their apparent drive capacities. HDDs internally have powers-of-two sector sizes, so it is clear their internal technology is also powers-of-two, so it would be more natural for them to also express capacity in powers-of-two.

Consumers, and the courts, don't give a crap what IEC/IEEE defines. They are standards bodies, and in no position to create "rules" as you put it. They are not in any position to define "correct" usage for commerce.

Operating systems are coded correctly because in computer science, powers-of-two have always been used to measure memory capacity. They have no need to otherwise support HDD manufacturers' inflated "marketing" capacities.

Parent

RE: A worthless settlement for a worthless case

By Graviton on 6/29/2006 11:41:19 AM , Rating: 2

quote:
This case is based on consumer expectations that a MB and GB are defined the way they expect - powers of two.

Wrong. The SI prefixes have specific power of 10 meanings (sometimes the greek and latin words are used elsewhere but the computer industry uses the whole SI system and abbreviations). Anyone who expected powers of two would have to be familiar with the way the industry sometimes incorrectly uses those prefixes as a form of shorthand when talking about large numbers of bytes (since binary prefixes have only recently been defined). Of course, if they already knew about the dual meanings of the prefixes, then they'd have no complaints and there goes their case against WD. As it stands, people wind up getting more space than the rounded-down advertised capacity of the drive would suggest - certainly nothing to complain about. But if WD labeled their drive capacities in the binary shorthand, then consumers would wind up with less space than advertised and would have a genuine case against them.

quote:
HDD manufacturers choose powers-of-ten definitions in order to inflate their apparent drive capacities.

While I'm sure the bean-counter types wouldn't mind people thinking their HDD's were larger, they are absolutely defining the size correctly, so kudos to them.

quote:
HDDs internally have powers-of-two sector sizes, so it is clear their internal technology is also powers-of-two, so it would be more natural for them to also express capacity in powers-of-two.

It takes less space to write out a large number that way if it's a power of two. But it's shorthand and that's all. Neither the real capacities nor the definition of the SI prefixes have changed. And how is it "natural" to express sizes (anywhere other than internally) in a way that can't be added, subtracted, multiplied, and divided, without resort to a calculator? The binary numbers are useless and frustrating to someone trying to just work with files and folders on their computer.

quote:
Consumers, and the courts, don't give a crap what IEC/IEEE defines. They are standards bodies, and in no position to create "rules" as you put it. They are not in any position to define "correct" usage for commerce.

True to a point. Still, most places where an O.S. would use the binary shorthand, they tend to also display the actual number of bytes. The usage is obviously for shorthand and not a redefinition of the term. Perhaps that's a subtle point, but just think of the example of a 2x4 inch piece of lumber. No, it's not actually 2 inches by 4 inches (by the time it's done being processed anyways), but neither have we redefined what an inch is. It's just rounding to make it easier to refer to that piece of lumber by a short name. At least simple rounding is less confusing than using a binary shorthand, but it's the same idea.

quote:
Operating systems are coded correctly because in computer science, powers-of-two have always been used to measure memory capacity.

How computers address memory is totally irrelevant to how sizes are displayed to the user. Sizes are what they are. At the end of the day what we are talking about is how sizes are displayed to the end-user. Using binary of any form is completely inappropriate. Yes, it has been used within the computer industry for a great many years. But, for example, it's also traditional to label electricity flowing from + to - in the physics world and some related industries. It works the way that it's used, but if someone ever says "electricity flows from + to -" they are dead wrong. Likewise, hardware designers and a tiny number of programmers can use their version of the SI system all they want if it's easier for them, but that doesn't change what those words mean to the rest of the world.

So, does anybody else have the sudden desire to be a lawyer?

Parent

RE: A worthless settlement for a worthless case

By TomZ on 6/29/2006 2:28:14 PM , Rating: 2

First, I can't believe you are arguing this point. 1 kilobyte has been 1024 bytes since computers first had enough memory to count that high. 1 kilobyte has never meant 1000 bytes ever . To argue against this is just foolish, because it is a historic fact and common usage.

quote:
The SI prefixes have specific power of 10 meanings (sometimes the greek and latin words are used elsewhere but the computer industry uses the whole SI system and abbreviations)

As "the goat" points out below, the computer industry took a different definition of these prefixes. It is not bound by SI - SI is not any natural or man-made law. Again, you may not agree with the decision here, but you cannot deny the fact that it is common usage and has been for many years.

quote:
As it stands, people wind up getting more space than the rounded-down advertised capacity of the drive would suggest - certainly nothing to complain about. But if WD labeled their drive capacities in the binary shorthand, then consumers would wind up with less space than advertised and would have a genuine case against them.

Wrong, the stated capacity is less than people would expect. For example, my "74GB" Raptor only has 69GB of capacity. This is the basis for the lawsuit in the first place, or did you not understand this?

quote:
While I'm sure the bean-counter types wouldn't mind people thinking their HDD's were larger, they are absolutely defining the size correctly, so kudos to them.

If you say so, but I disagree. If you buy a 1GB memory DIMM, guess how many bytes that is - hint: it is not 1,000,000,000. If you have a file that is "1k," that means its size is 1024 bytes. If you have a sector size of "32k" on your HDD, that means 32768 bytes, not 32,000 bytes, and so on.

quote:
It takes less space to write out a large number that way if it's a power of two. But it's shorthand and that's all. Neither the real capacities nor the definition of the SI prefixes have changed. And how is it "natural" to express sizes (anywhere other than internally) in a way that can't be added, subtracted, multiplied, and divided, without resort to a calculator? The binary numbers are useless and frustrating to someone trying to just work with files and folders on their computer.

You missed my point, which is that the implementation of HDDs are all in powers of two, i.e., design/engineering/manufacturing. HDD sector sizes are not 1000, 2000, 4000, etc. bytes - they are 1024, 2048, 4096, etc. bytes. It is only the marketing side that uses powers-of-ten. Do you think marketing does this because they want to be "SI correct"? Of course not. That might be something an engineer would do, but they didn't.

quote:
How computers address memory is totally irrelevant to how sizes are displayed to the user. Sizes are what they are. At the end of the day what we are talking about is how sizes are displayed to the end-user. Using binary of any form is completely inappropriate. Yes, it has been used within the computer industry for a great many years. But, for example, it's also traditional to label electricity flowing from + to - in the physics world and some related industries. It works the way that it's used, but if someone ever says "electricity flows from + to -" they are dead wrong. Likewise, hardware designers and a tiny number of programmers can use their version of the SI system all they want if it's easier for them, but that doesn't change what those words mean to the rest of the world.

You are wrong to try to marginalize powers-of-two units, because they are actually the most common representation. Now the average consumer probably doesn't know what a KB or MB or GB actually equate to, but everyone in industry does. You say, "tiny number of programmers," but that is not right - more like all programmers . That is one of the first and most basic things that you learn, and it has always been that way.

Parent

RE: A worthless settlement for a worthless case

By Tyler 86 on 6/29/2006 4:55:18 PM , Rating: 2

Indeed, all programmers .
All computer developers, hardware and software.
It's due to the binary representation, not the decimal representation.

You can represent 1024 different symbols with 10 fingers, zero to one-thousand-twenty-three.

Parent

RE: A worthless settlement for a worthless case

By Obadiah on 7/5/2006 8:48:47 PM , Rating: 2

quote:
You missed my point, which is that the implementation of HDDs are all in powers of two, i.e., design/engineering/manufacturing. HDD sector sizes are not 1000, 2000, 4000, etc. bytes - they are 1024, 2048, 4096, etc. bytes. It is only the marketing side that uses powers-of-ten.

You keep asserting this, but all it does is indicate that your level of understanding is superficial. Hard disk sector sizes are NOT powers of two.

The data payload sections of the sectors are indeed usually powers of two, as a convenience for the software that must interact with them. Currently 512 bytes per sector, the sizing is arbitrary, c.f. Microsoft's push for disks with data payloads of 4096 bytes per sector. This is an OS manufacturer's decision to chose 4096 byte payloads because it is convienent for them.

The actual sectors contain a lot more than just the payload. They contain start and stop bits, sector id bits, lots of reed-solomon error correction bits and probably some more stuff I can't remember off hand. Rarely, if ever,is the size of the actual sector on disk a power of two.

You can easily see this for yourself - while sector size will differ from model to model of hard disk - standard media like CDs and DVDs all have constant sized sectors. The physical size of a DVD sector is always 2064 bytes and the size of a CD sector is 2352 bytes - neither are powers of two. http://www.opticaldisc-systems.com/2002NovDec/DVD8...

Furthermore, the actual number of sectors on a hard disk is almost never a power of two. Nor are the number of sectors per cylinder - as that varies by the cylinder's distance from the center of the disc.

So, your claim that, "the implementation of HDDs are all in powers of two, i.e., design/engineering/manufacturing." is patently false. The design, engineering and manufacturing of hard disks is always cognizant of being used in systems where it is most convenient for user data to use units that are sized in powers of 2, but thats all there is to it - EVERYTHING else is not a power of 2. Same thing for just about any storage medium - CDs, DVDs, 4mm/8mm/dlt/etc tape, floppy disks, whatever.

Even computer RAM has not always been measured in powers of 2. Many systems from before your 15 years in the industry had 18-bit and 36-bit words, not 16-bit or 32-bit. Some used 7-bit bytes, some 9-bit bytes. Examples of such machines include the IBM 701/704x/709x line, the DEC PDP-1, PDP-6, PDP-9, PDP-10 and PDP-15.

Parent

RE: A worthless settlement for a worthless case

By the goat on 6/29/2006 9:47:24 AM , Rating: 1

Wow all the replies for this article made me register so I could respond even though I've been reading here since day one.

Any body who thinks giga, mega and kilo must mean powers of 10 is wrong. The SI system doesn't have the magical power to control what words mean. Those words/prefixes are taken from Greek and Latin. The computer industry took the same words from Greek and Latin and applied their use to the computer memory unit byte. The computer industry did not take the prefixes from the SI system.

The computer industry gets to define its own prefixes. And since the beginning of computing they have been defined as follows:

kilo: 2^10
mega: 2^20
giga: 2^30
tera: 2^40
and so on. . .

Just because many of you seem to think that the SI system invented these words doesn't mean they can redefine the meanings that have been in use inside the computer industry for over half a century.

All computer companies should use the same standard of measurement when discussing bytes. HDD manufacturers were wrong to switch their measurements. Yes they did use the correct definitions back in the day.

Also this settlement is completely worthless. Free backup software? Does anybody actually "backup" their hard drive anymore? I copy certain important files to a RAID array as a backup. I haven’t done a full partition backup in years.

Parent

RE: A worthless settlement for a worthless case

By Tyler 86 on 6/29/2006 4:35:03 PM , Rating: 2

I think the term 'back-in-the-day' is screwing with people.
True, 'back-in-the-day' this was a non-issue. It was as it was, and was accepted to always be, because there was no alternative, because the only guys that used it actually used it; developers, enthusiasts, and the like.
This isn't an issue of back in the day at all though, the binary basis for these units of measurement is correct, not the decimal system.

The decimal system is flawed. Don't get me wrong, trailing digits is a fantastic idea - but it system of 10 is wrong.
You can represent 1024 variations with 10 digits - that means you can represent 0 to 1 byte under a kilobyte (1023) on both frickin hands.

The decimal system can go to hell. Long live binary!

Parent

Win/Win for WD

By techfuzz on 6/28/2006 4:11:31 PM , Rating: 2

Not only do they settle and avoid a lengthy and costly trial that they may or may not win (plus any appeals), they give away their own branded backup software driving a wedge into the personal backup software market. Even if 1% of the people who end up with the backup software eventually upgrade it to a new version then WD wins.

WD will keep reporting drives sizes higher than the OS does, gains new market share in the personal backup software market, and gets profit from any upgrades to their backup software that they almost certainly would have never gotten without this settlement.

Win/Win situation for WD all around!

RE: Win/Win for WD

By Squidward on 6/28/2006 4:14:53 PM , Rating: 2

eh true, but really a win/win situation for the lawyers. Does anyone think they really filed this lawsuit in the interest of the consumer? It was all about the cash.

Parent

RE: Win/Win for WD

By brystmar on 6/28/2006 4:20:22 PM , Rating: 2

You really think that some law firm just randomly decided to sue WD one day? No, some consumers went to them with their case and they wisely decided to take it. Lawyers always win in almost every lawsuit; that's just the way it works. I'm in the wrong profession...

Parent

RE: Win/Win for WD

By Squidward on 6/28/2006 4:26:10 PM , Rating: 2

Oh I know consumers did, I remember the article when it first came out last year when this was initially being filed. But it isn't a deception by the vendors like WD per se, it's that the consumers don't understand how the file systems work. It was a quick easy lawsuit but the only one who really got anything out of it is the lawyers.Half a mil for this one, another half mil or so against Seagate when that settles. What do the consumers get? Nada. Anyone who knows hard drives knew they weren't getting ripped off.

Parent

RE: Win/Win for WD

By TomZ on 6/28/2006 4:31:43 PM , Rating: 2

It's not a file system issue, it is a HDD manufacturer marketing issue. They decided to redefine a MB and a GB to be something else other than the industry-accepted unit of measure.

This lawsuit, AFAIK, does not relate to any overhead required by the file system.

Parent

The common mistake

By mindless1 on 6/28/2006 10:56:35 PM , Rating: 2

Everyone seems to be getting it backwards as to why the capacity was misleading. It is not how many bytes were in a "Giga", it is that a byte is NOT A DECIMAL SYSTEM VALUE.

It is invalid to use byte within a purely decimal system value. If the HDD manufacturers would care to call it something other than byte, something that exists in the decimal system instead, let them go right ahead and do it, but then they'd need to actually tell us the actual value in bytes too because it is without question the relevant value in the context of data storage.

RE: The common mistake

By Some1ne on 6/29/2006 1:55:37 PM , Rating: 2

I disagree, a byte is just 8 bits, and really there's no "system" associated with how you decide you want to cound groups of 8 bits. You can count them in groups of 1000, or groups of 1024, or groups of 69 if you really wanted, and no approach is any more or less valid than the others.

It would still be nice if the counting were consistent between the OS and the drive manufacturers though...if all the operating systems around define a "GB" as 1024^3 bytes, then that's how HDD manufacturer's should define it too. Maybe instead they should just provide a raw count of total bytes available on the drive, and not use labels like "MB"/"GB"/"TB", and then people who prefer to count their bytes in multiples of 1000 can say "well that one's a 100 GB drive", and people who prefer to count their bytes the same way a computer does can look at the same thing, and say "well that one's a 92.4 GB drive".

Parent

RE: The common mistake

By TomZ on 6/29/2006 3:51:56 PM , Rating: 2

WinXP at least displays the sizes in the way that you suggest - total number of bytes, plus GB.

Parent

RE: The common mistake

By Tyler 86 on 6/29/2006 4:52:21 PM , Rating: 2

So does Linux, and Mac, and .. well.. the only thing that I know of offhand with an easily changable storage representation display would probably be Azureus, but even then it's 'easily' changable, hidden within under the interface section of the selection tree in options.

And it's not 'memory verus hdd', it's just 'binary storage'.

Parent

So the problem is with OS's not HDD Makers.

By Xajel on 6/29/2006 1:17:05 AM , Rating: 2

The HDD makers used the standards as well as all those kilos, megas, and gigas things...

So Which OS do we plame for this ??? What's the first OS that started to use the 1024 thing ??

RE: So the problem is with OS's not HDD Makers.

By DOSGuy on 6/29/2006 7:29:51 AM , Rating: 2

Which OS started this? Every OS ever, because that's how a kilobyte/megabyte/gigabyte has been defined since the beginning of computing.

Parent

RE: So the problem is with OS's not HDD Makers.

By Tyler 86 on 6/29/2006 5:00:24 PM , Rating: 2

Lest we forget, even before OSs, there were one thousand twenty four bytes in a kilobyte. Data storage came with computers not operating systems.

Parent

So the capacity stays falsely listed?

By paperfist on 6/30/2006 11:21:58 AM , Rating: 2

I didn't see anywhere that I read that the actual 'problem' was going to be resolved only that money and great software was going to be given out instead of fixing the real issue...

Or am I blind? :)

RE: So the capacity stays falsely listed?

By Tyler 86 on 7/1/2006 11:48:38 AM , Rating: 2

quote:
So the capacity stays falsely listed?

Yep, believe so.

Parent

RE: So the capacity stays falsely listed?

By Randalllind on 7/6/2006 11:45:34 AM , Rating: 2

If 1gb is 1024mb and not 1000mb then the hard drive makers should not get away with saying 1000mb.

Why can't anyone agree on anything any more? 500gb drives are 465gb sucks that should not be allowed!!

Everyone hates Standards?

Parent

Lawyers & manufacturers who make false claims are SCUM !

By cornfedone on 6/28/06, Rating: 0

RE: Lawyers & manufacturers who make false claims are SCUM !

By kasey01 on 6/28/2006 6:18:39 PM , Rating: 2

What false claims are you talking about in this lawsuit?

Parent

RE: Lawyers & manufacturers who make false claims are SCUM !

By Tyler 86 on 7/3/2006 5:02:26 PM , Rating: 2

He's referring to WD & Seagate specificly, as well as 'and all', whether they're making false claims or not, AND the lawyers, whether they are making false claims or not, as in the long run, no useful valuable compensation is gained by the consumer.

Parent

200,047,001,600 bytes

By Muirgheasa on 7/6/2006 8:18:11 AM , Rating: 2

200,047,001,600 bytes. That's the total capacity of my secondary HDD, which was sold to me as a 200GB drive. That gives me a total storage capacity of 186GB, but does that mean that I'm being ripped off? Or is Microsoft simply misleading me by quoting my hard drive as being smaller then it is? Of course not! That's how technology has always worked! If you feel ripped off - well, ignorance isn't an excuse in a court of law, is it?

RE: 200,047,001,600 bytes

By doggie96 on 7/6/2006 9:05:44 PM , Rating: 2

No, but what if your new Geforce 7900GTX came with only 500mbs of RAM? You would call Nvidia and they would say "were sorry, it is half a gig of vram?"

Parent

One per customer?

By DOSGuy on 6/28/2006 10:39:14 PM , Rating: 3

I bought four or five Western Digital hard drives during that timeframe, so I get to download the same piece of software four or five times? I also advised the purchase of several hundred Western Digital hard drives for the school board. Maybe I should spend a weekend matching a WD serial number with a student's name and give 300 of the kids some free software? It seems to me that the remuneration from a settlement should be per purchase instead of per customer. They're not offering any money, and nobody needs five copies of the same program, so Western Digitals biggest and most loyal customers get squat out of this. I admit that it's a dubious lawsuit, but it isn't even a proportionally just settlement!

New capacity labeling...

By Vysion on 6/29/2006 3:10:58 AM , Rating: 3

Hard drive makers will probably start labeling their storage capacity like this... "100 GB ~ 93.1 GiB"

This will prove that they were not doing anything wrong, and make the binary prefixes more common.

byte me

By rqle on 6/28/2006 11:52:36 PM , Rating: 2

WTH , once bought 10 WD 120GB/160GB harddrive, those guys place 200GB HD in the containers instead, can i sue too...

Not the only place...

By Keeir on 6/29/2006 2:03:54 AM , Rating: 2

Unfortunately, this is not the only place of measurement confusion...

WD and other Hard Driver manufactures are in the "wrong" here because they are producing products for a particular market where the term "Giga-byte" has a specific meaning...

Similar to how a troy ounce/pound is used to describe quanitities of gold rather than an english ounce/pound.

It sucks, but it's common

By GreenEnvt on 6/29/2006 9:25:40 AM , Rating: 2

This practice is not just in the computer industry.
Go to your local home depot and get a 2x4, notice it's not actually a 2x4, but more a 3.5x1.5.

What is the lesser of two evils, changing all documentation for all systems that currently advertise megabytes/kilobytes, etc? Or just accept it?

Class action

By rudder on 6/29/2006 10:24:53 AM , Rating: 2

You all are putting to much thought into this. Class action lawsuits are a business. There his a whole mess of lawyers just looking for a lottery ticket as they figure out what they can sue for. Stop arguing over what constitutes a gigabyte.

Same thing happened with the floppy drive and Toshiba being sued in class action. Lawyers are the ones who benefit.

Some class actions suits are indeed about the consumer, but the majority just make a law firm a ton of cash.