Snow Leopard: The new one gigabyte, now slimmer than before

Hey @harveylubin,

About the dozen thing:

Say you go to a new place and the dozen there is different. It's different because everybody in this place thinks that the dozen is 10, so they changed it. If people that use computers that use the old system think that a kilobyte is 1000 bytes, they're only 24 bytes off. This was OK when we were only using small hard drives. But if these people think that a petabyte (which is what we will be working with in the not-to-distant-future) is a quadrillion bytes, then they would be 125,899,906,842,624 bytes off, which is a hell of a lot. So if they change it to what people think it is, then everyone is right.

And FYI, @Anderson WOULD have felt ripped off prior to the change as most hard drive companies (including Apple) have used Base-10 measurements for a long time in products such as their time capsual and their computers, but not in their operating systems. So it all makes sense.

The only argument against this change that holds any water for me is the consistency issue. I don't want to have to figure out which system iTunes is using vs Finder vs a PC. So I'd like the option to revert to binary, unless the rest of the world comes along to decimal.

I don't buy the argument that Apple is ripping me off with this. My MacBook has the same amount of space it always did (actually about 6.5 GiB or 7 GB more now.)

I get that there's a level at which hard drive manufacturers are ripping us off, but on the other hand, hard drives just keep getting bigger and cheaper, so it's hard for me to complain. Back when drive manufacturers made the change, all my data fit on a 10 GiB drive, which cost $250.

Now I have 90 times the data. And I have a drive 93 times the size for half the money. If manufacturers were forced to size their drives using binary figures, would the extra 70 GB add about 7% to the cost of the drive? Probably. Would I pay an extra $10 to get a true TiB drive. Certainly. Would I have gotten a better deal? Not at all. Would the consistency and logic of the binary calculation appeal to my sensibilities? Definitely.

Apple wants people to see they really got their TB. It's silly, and it's pandering to the drive manufacturers, but it changes nothing about the size of my drive (or yours). Computers may work in binary, but the only reason I'd rather not use decimal to inventory my data is that everyone else does it differently, and because I'd like to see consistency between the way computers work and the way we describe them. But anyone who has the capacity to understand binary mathematics, can also convert binary numbers to decimal numbers.

There may be no amount of IEEE decision making that will ever make a drive base 10, but that doesn't mean it can't be measured that way. Any number can be measured in any base. The folks who built the roads in Europe may have used miles, and you might even find one somewhere that was built purposely to be exactly 100 miles long. That doesn't mean it can't be measured in kilometers, and it certainly wouldn't change it's length if you did.

I've updated to Snow Leopard, and I guess I'll see how much of a drag it is having my files measured differently than everybody else's. But whether you call it an extra 6.5 GiB or 7 GB, I have more space on my drive and hardly feel ripped off.

"Computer storage and addressing has NEVER been metric"

Storage certainly has. Read through some old hard drive specs from the dawn of computing. Even drum memories (precursor to hard drives) were measured in metric kilobytes. Measuring disks as multiples of 1024 is simply wrong.

"unless someone goes back to square-one and reinvents computers based on a system of 10s"

Your ignorance is showing. The first computers WERE based on a system of 10s. http://en.wikipedia.org/wiki/Decimal_computer

The "hard drive manufacturers are ripping us off!!!!" argument is about as dumb as it gets.

And your "dozen" analogy is totally wrong. Check out footnote #2: http://www.wdc.com/settlement/docs/document20.htm#_ftn2

Seems like wherever there's an extra buck to be made, someone will jump at it. The original idea behind kilobyte being 1024 bytes makes sense. It's the closest power of 2 to the power of 10. And it's remarkably close, being only 2.4% off the convenient 10^3 grouping that we use for large numbers. So, rather than making up a whole new set of prefixes -- which would have been an audacious move for the young science of computers -- they "borrowed" the ones that everyone was already familiar with. There's nothing wrong with that. It worked great for years. Until the marketing guys came along, and saw there was a buck to be made under the guise of making things easier for the consumer -- why give the consumer 1024 bytes for the kilobyte when you can give them 1000 and act like you're doing them a favor by "reducing confusion" or some-such guise for increasing gross margins? Except technical people know what a kilobyte is -- it's 1024 bytes. The cat's out of the bag. It's not going to be easy to get it back in -- just go to Google and type "bytes in a kilobyte" or look up kilobyte in the dictionary -- they seem to have missed the memo, as have most technical sources I've consulted in the past ten years. Kilobyte, Megabyte, Terabyte have been technically, and popularly defined, and the fact that the prefixes were borrowed from SI doesn't change the fact that they now have their own definitions as powers of 2^10.

Those who support this change are bringing up two points that have no logical basis.

First of all, for anyone to say that "It started with computer types INCORRECTLY using the previously word existing "kilo" to mean 1024." is laughable, and demonstrates a level of ignorance that is hard not to pity.

Computer storage and addressing has NEVER been metric, and never will be (unless someone goes back to square-one and reinvents computers based on a system of 10s... in which case we will have to trash all of our current computers, software, hard drives, etc.).

Second, the IEEE is not a holy spirit that sees and knows all. It is a group of people who make decisions on technical standards for marketing purposes. Sometimes their decisions make a lot of sense, but other times (such as in this instance) they don't!

What the IEEE is doing with this decision, at the pressuring of the hard drive manufacturers, is to change the long-standing, reality-based binary system of sizing hard drives (to what they have been getting away with for the past 10 years) so that they can make more money by selling their products at inflated, unrealistic sizes.

If you are supporting this and enjoy being ripped off by buying a "1TB" hard drive that is really only 930 GB in size, then it's hard to understand your point of view.

As much as it is an IEEE supported standard, instigated by the hard drive manufacturers, it is still not based on anything approaching reality.

The reality is that computers (and hard drives) always have been binary systems, NOT base-10. Saying that you are doing this to go metric and/or be less confusing to people also is unrealistic.

Hard drives are made up of sectors/blocks of 512 bytes (2 to the 9th bytes) or 1024 bytes (2 to the 10th bytes) (see http://en.wikipedia.org/wiki/Cylinder-head-sector). These are the smallest units that make up a hard drive. There is no such thing as a 1000 byte (10 to the 3rd) sector.

It is therefore IMPOSSIBLE for hard drive manufacturers to make a hard drive that actually uses their own measuring of "1TB" (1,000,000,000,000 bytes). They are stuck in reality having to make hard drives that use the binary system, NOT the base-10 system!

This is why an advertised "1 TB" hard drive is 1,000,204,886,016 bytes in actual size (not 1,000,000,000,000 bytes). There is no amount of IEEE decision making that can ever make a hard drive base-10. Repeat... It is IMPOSSIBLE.

When you buy a "1TB" hard drive you are only getting a bit more than 90% of a real 1 TB.

And just remember that the gap between what the hard drive manufacturers sell you and the actual number of bytes you think you are buying will widen as time goes by.

With each progression to a new order of measurement, (MegaBytes, GigaBytes, TeraBytes, PetaBytes, ExaBytes, etc.) that difference gets bigger and bigger; meaning more savings for the hard drive manufacturers and the less you are getting when you buy a hard drive.

For example the difference between a real kilobyte (1024 bytes) and 1000 bytes is only a 2.4% difference. But the difference at the TeraByte level (1,099,511,627,776 bytes vs. 1,000,000,000,000 bytes) jumps way up to a much larger 10% loss! And this loss will get greater and greater exponentially into the future.

In other words: Reality Bytes.

In any case, it's still nothing more than a marketing tactic.

Take the iPod Touch. You know, an "8GB" iPod Touch. has 8GB of storage. So why the heck does "Settings > General > About" show only 7GB "available"? And my 16GB model only shows 14,5GB available. OK, so maybe it's the OS that takes up the extra space, but then tell us that.

I think its funny that many readers are blaming the hard drive manufactorers or even Apple for calculating Disk sizes incorrectly. It's true that before 1999, file sizes were calculated as multiples of 2 since thats the way computers work. But the SI prefixes are way older than IT and have always been multiples of 10. So in reality, Apple is finally fixing this missuse of SI prefixes. Personally, I have wanted this feature for years.

Of course, computers will continue calculating in binary, they just convert the numbers before dispaying. For humans calculating in decimal is much easier than calculating in binary and after all, ease of use is what the graphical user Interface is all about.

I really am getting fed up of all this dumbing down.

Computers have always used binary and therefore the units used to respresent higher values have always been based on powers of 2. I've owned computers since 1983 and my KBs have always been 1024 bytes. I have programmed in Machine Code and using a 1000 byte KB would have screwed things up so badly.

Until computers start thinking in decimal, these binary values should be preserved.

Hard drive manufacturers only started using the decimal values so they could advertise a higher capacity for their drives.

I am one of the people who won't be upgrading to Snow Leopard until Apple fix this.

"It's only in recent years that hard drive manufacturers colluded with one another to change the way it sized hard drive capacities."

Not so.
This all started with computer types INCORRECTLY using the previously existing word "kilo" to mean 1024. Kilo meant one thousand well over two hundred years ago (1795). As in kilogram and kilometre.

Then the computer types went on to incorrectly use "mega", which already was in use in the metric scale, as in megahertz, megawatt, megaton, etc. Similarly giga was already in use in the metric scale, as in gigahertz, etc.

So the correct use of these words was for powers of ten - LONG before they started to get used in computer language ---- long before computers even existed. Their use in the metric system is, in fact, the already existing and correct usage.

The mebi etc. have been suggested. Yes, they have not been officially adopted. But "the IEEE Standards Board decided that IEEE standards will use the conventional, internationally adopted, definitions of the SI prefixes. Mega will mean 1 000 000". The use of kilo, mega and giga to mean 1024's always was incorrect. The IEC and IEEE are simply trying to clarify this mistaken and confusing usage by emphasizing the pre-existing usage in the world-wide metric system.

"the reality is that a gigabyte has always been 1024 megabytes"
Indeed - but it never should have been - BECAUSE GIGA ALREADY HAD A MEANING - as part of the metric system.

See also

1. Summary - http://physics.nist.gov/cuu/Units/binary.html

2. Worldwide metric (SI) prefixes (kilo, mega, etc.) http://en.wikipedia.org/wiki/SI_prefix.

2. The world's standard of measurement, "The International System of Units" at http://en.wikipedia.org/wiki/International_System_of_Units. Scroll down to read "units".

It is very odd throwing a 4.19gb file onto a fat32 usb drive.. lol