GB vs GiB

[JPG]

In a brief off topic discussion in another thread the subject of what a gigabyte is ensued.

As it turns out after some shallow research, since 1999 a gigabyte is officially "1,000,000,000" bytes.

In the interest of easier typing and mental assimulation, I will continue this with Kilobytes instead of Gigabytes.

To many of us computer heads a kilobyte is 1024 bytes.

To others(storage equipment folks) a kilobyte is 1,000 bytes.

This leads to confusion(and law suits) so in at the end of the last millenium(sp?) the IEC decided to eliminate the confusion.

The result was the definition of 1024 Bytes as KiB. The "i" comes from bInary as opposed to decimal.

I have purposly refrained from using the non=abbreviated version of binary kilobytes. It simply looks too weird!

The consistency comes from K,M,G,T... always referring to a decimal power of 10. 10 to the x where x= 3,6,9,12 ...

The Ki,Mi,Gi,Ti, always refers to a binary power of 2. 2 to the y where y = 10,20,30,40...

Before my scant research I was in that group who thought storage folks were taking advantage of most common(normal) folk who did not know beans about bits/bytes etc..

BTW them storage folks were using the decimal norm before the binary norm was imagined.

So it is that I am now firmly entrenched with the decimal version of KB. It will take some time to get used to KiB. I still am having trouble with Hertz rather that cps.

[jsburger]

In a brief off topic discussion in another thread the subject of what a gigabyte is ensued.

As it turns out after some shallow research, since 1999 a gigabyte is officially "1,000,000,000" bytes.

In the interest of easier typing and mental assimulation, I will continue this with Kilobytes instead of Gigabytes.

To many of us computer heads a kilobyte is 1024 bytes.

To others(storage equipment folks) a kilobyte is 1,000 bytes.

This leads to confusion(and law suits) so in at the end of the last millenium(sp?) the IEC decided to eliminate the confusion.

The result was the definition of 1024 Bytes as KiB. The "i" comes from bInary as opposed to decimal.

I have purposly refrained from using the non=abbreviated version of binary kilobytes. It simply looks too weird!

The consistency comes from K,M,G,T... always referring to a decimal power of 10. 10 to the x where x= 3,6,9,12 ...

The Ki,Mi,Gi,Ti, always refers to a binary power of 2. 2 to the y where y = 10,20,30,40...

Before my scant research I was in that group who thought storage folks were taking advantage of most common(normal) folk who did not know beans about bits/bytes etc..

BTW them storage folks were using the decimal norm before the binary norm was imagined.

So it is that I am now firmly entrenched with the decimal version of KB. It will take some time to get used to KiB. I still am having trouble with Hertz rather that cps.

[wrmnfzy]

Thank you, I think.

[edma194]

As it turns out after some shallow research, since 1999 a gigabyte is officially "1,000,000,000" bytes.

Which official decided this? And are we going to finally get the official value of 'pi' as 3?

[JPG]

International Electrotechnical Commission

[DLB]

I like the clarity, but I hate the approach. Changing the definition of an existing term and creating new terminology for the old definition? Are you kidding? That's like changing a foot to be 10 inches and inventing a new term for something 12 inches long. It does not seem to me that this new definition is widely adopted. Notably, it is not adopted by MS in the OS and FS. I just stuck a flash card (allegedly 64GB) in my Win10PC, went to FM and right-clicked properties. I see: :Capacity 63,831,015,424 bytes 59.4 GB" It does not mention GiB, but I'm confident that if it did this would also be 59.4. It does not say 63.8 G_anything. Suggesting that neither MS nor I got the memo when GB was redefined. Or we ignored it. So if I want to meet the forum requirement for 3MiB and I'm using the file system on my computer I have to know that 3MB is the same thing, neither mean 3 million bytes in context and both are based on the binary convention popular in the computer science field.

Plus GiB is already taken by the military, the worldwide experts on abbreviations and acronyms. When fighter aircraft commonly had two crewmembers, it stands for Guy in Back.

For those people that would care, we had a reliable system in 1997 because we knew that what KB, MB, or GB meant depended on context. Now we have an unreliable system because it depends on whether the user accepts and uses the new definition for the old terminology. Perhaps it would have been better if the new terminology mapped to the non-standard definition. The "i" could come from somewhere in the word decimal instead of somewhere in the word binary.

These well-meaning folk should work on the ambiguous angular measurement 'mil.' Here we have one unit of measure with two completely different definitions further confused by the fact that they are close together. But 1/6400 of a circle does not equal one milliradian, they are just close enough to create confusion or worse. (Don't ask me how I know. It was expensive and time-consuming, but no one was hurt.)

- David

[edma194]

So thanks to the courageous efforts of the IEC any use of abbreviations like GB are now more uncertain than they were before.

[JPG]

The IEC did not 'redefine' anything. Kilo, Mega, Giga. . . have been defined as "decimal" prefixes for god knows how long.

It was the use of those prefixes for binary stuff that created the confusion.

AS usual this will trickle down to the marketing types eventually. Since they thrive on using potentially misinterpreted terms to their advantage . . . . . . .

Now about them circles/angles.

6400??? Whazzat? A more precise number is 6283.185307... = 1000 mil / radian /360 = 0.01745... radians/degree

A circle consists of 2π radians. / 360 = 0.01745... radians/degree.

They be the same thing IF the 6400 'approximation' is NOT used.

[BuckeyeDennis]

As an old computerhead myself, I would have sworn on a stack of bibles that a kB was 1024 bytes, or 2^10. Etc. for MB, GB and such. I had no clue that this had changed until I read this thread.

But I’ve gotta say that I think the IEC did the right thing. After all, is a kilometer (km) 1024 meters? I think not.

Better to aggravate us old computer farts now, and get it over with once and for all. Otherwise, the lovely pristine SI unit system would eventually devolve into something as ugly as the imperial system, and the various bastardizations thereof. Such as the lbm, or pounds-mass. Anyone here for cubits? And although “feet” is still a well-understood unit of measure, there are also “hands”, “knots”, and various other units that once made perfect sense to the users thereof.

My vote is for unambiguous, consistent standards. And in this case, retrospectively, I think the binary crowd played fast and loose with unit prefixes that had previously and consistently been defined in the decimal system.

[DLB]

6400??? Whazzat? A more precise number is 6283.185307... = 1000 mil / radian /360 = 0.01745... radians/degree

The military sometimes (or used to?) uses the Grad or Gradian system, 400 grads in a circle. In that system an 'angular mil' usually 'mil,' is 1/16 grad, or 1/6400 of a circle.

- David