• On BNET: 3 worst things about the iPhone 3G S
December 4, 2007 10:14 PM PST

Gigabytes vs. gibibytes class action suit nears end

by Declan McCullagh

A long-running California lawsuit over whether all megabytes and gigabytes are created equal may have reached its end on Friday.

The class action lawsuit against Kodak, Sandisk, Lexar Media, and other memory card makers alleges that the defendants intentionally misrepresented the capacity of their flash memory devices by using decimal definitions, in which a megabyte is 1,000,000 bytes. The suit says a binary definition is appropriate, meaning that one megabyte equals 1,048,576 bytes and that the memory card sizes were overstated by 4 percent to 5 percent.

When memory capacity was smaller, the difference didn't mean much. A decimal kilobyte, at 10^3=1,000 wasn't very different from 2^10=1,024.

But as capacity grows, the differences become more significant (technically, the ratio between the decimal and binary representations increases). This explains why your new terabyte drive isn't as capacious as you hoped it might be. A 10^12=1,000,000,000,000 decimal terabyte is roughly 10 percent smaller than the binary equivalent of 2^40=1,099,511,627,776. Here's more background on why computers work this way.

So the class action lawyers sued five flash memory card makers, alleging breach of contract, fraud, and violations of California's unfair competition laws.

In a strict standards sense, the companies were probably right, the decimal metric prefixes were accurate when applied to removable storage, and customers shouldn't have grown used to the near-equivalence. But the attorneys decided to settle and reached an agreement: some customers would get a 5 percent refund, while all would get a 10 percent discount from the companies' online stores. Another part of the settlement was to disclose that decimal prefixes were being used. (The deadline to make claims was December 20, 2006.)

The class action lawyers at the Pasadena firm of Kendrick & Nutley and the San Diego firm of Kendrick, Bonas & Nutley got rich, or at least richer: they got a check for $2.38 million.

Four people objected to the settlement and filed appeals roughly a year ago. One claim was that the 5 percent refund amounts to only a few dollars and was insignificant. Another was that the fees handed to the class action lawyers were too high.

But a three-judge panel of the California Court of Appeal's First District rejected those arguments on Friday and upheld the settlement.

At this point it's reasonable to note that there actually are terms that avoid all this confusion, and those include IEEE 1541 terms gibibyte (2^30 bytes = 1,073,741,824 bytes) and tebibyte (2^40 bytes = 1,099,511,627,776 bytes, or 1,024 gibibytes).

Unfortunately--or, perhaps for hard drive and flash drive manufacturers, fortunately--gibibyte and tebibyte still sound a little too silly to be taken seriously.

Declan McCullagh, CNET News' chief political correspondent, chronicles the intersection of politics and technology. He has covered politics, technology, and Washington, D.C., for more than a decade, which has turned him into an iconoclast and a skeptic of anyone who says, "We oughta have a new federal law against this." E-mail Declan.
advertisement
Click here!
Recent posts from News Blog
Neil Young Archives Blu-ray: Rip off?
Acronis revises survey results about backup habits
Acronis miscalculates data on users' bad backup habits
Flickr co-founder presses beta button
Comcast, Sony open retail store
Cox to try coaxing the Internet into submission
Was InfoWorld's CTO of the Year award a year late?
VMWare VI4 renamed to vSphere
Add a Comment (Log in or register) (11 Comments)
  • prev
  • 1
  • next
This was known along time ago
by Maelstorm December 5, 2007 12:46 AM PST
I remember tags on old Seagate ST220 hard drives (20 MEG) that said that 1 Megabyte = 1,000,000 bytes, and later drives that have both the megabyte and gigabyte factors, then just the gigabyte one. Because it's on the label of the drive itself, people were informed about it...if they bothered to read it. Most computer people should know that RAM is base 2 (since that is how computers count), and hard drives are base 10 (because of the IEEE standard). It was that way since the 1980's, and maybe even longer. Personally, in my opinion, I believe that this was a frivolous lawsuit that began with the ignorance of the consumer and was perpetuated by the greed of the trial lawyers.

I also want to point out the error in the article that this blog post refers to. That article states that computers read and write data to drives in 1024 byte blocks. This is not true. The smallest data transfer between any computer and hard drive is 512 bytes, which is the physical sector size of the drive. It has been this way since I can remember (1980's).
Reply to this comment
nice label..but..
by basraw December 5, 2007 6:15 AM PST
THEY DON'T USE IT ANYMORE

"I remember tags on old Seagate ST220 hard drives (20 MEG) that said that 1 Megabyte = 1,000,000 bytes, and later drives that have both the megabyte and gigabyte factors, then just the gigabyte one. Because it's on the label of the drive itself, people were informed about it...if they bothered to read it."
View reply
common issue with EVERY device
by Riquez-001 December 5, 2007 2:11 AM PST
Every single electronic storage device from memory cards to
computers to xBox's has this issue.

I agree it's misleading & as devices grow in size so does the discrepancy. For instance, I recently bought a "500GB" backup
drive - but hold on! The capacity is only 465GB. Thats 7% less
than advertised. & those missing 35GB? That's A LOT of storage.

Manufacturers just need to nip this trend in the bud, just say
465GB, Ill still buy the drive, but I wont be angry or confused.
Reply to this comment
Re: common issue with EVERY device
by jaximflash December 5, 2007 4:28 AM PST
I totally 100% agree!
My iPod....
by mlauzon December 5, 2007 8:11 AM PST
I also agree, as my iPod is a 30GB model, but it is actually 27.6GB...there is no way the firmware is 2.4GB in size; so where are the missing 2.4GB?!
I hope they change it
by jaximflash December 5, 2007 4:27 AM PST
This may be a well known issue within the technical circle, but to non-nerds, it's a new things. I think manufacturers should state the number of gigabytes according to the customers' expectations. If a customer is expecting 750 GB from a 750GB harddrive, then they should get 750GB, not 700GB. If they are given 700GB instead, then the manufacturer should change the packaging to reflect this. The manufacturer may be "technically" correct is their assertion, but it doesn't matter when a non-nerd is expecting something else.
Reply to this comment
Change....
by mlauzon December 5, 2007 8:12 AM PST
Well, the price needs to change as well, why pay for a 750GB HDD when you only are getting 700GB?!
MICROSOFT WINDOWS REPORTS ONE UNIT BUT LABELS IT WITH ANOTHER
by krosavcheg December 5, 2007 7:22 AM PST
The IEC, IEEE, CPIM, and NIST define Giga (G) as 1,000,000,000 or one billion.

The same standards organizations define Gibi (Gi) as 1,073,741,824.

As such, by standard definitions, 1TB hard drives are in fact 1000GB, or 1000 gigabytes.

"Windows Explorer" does not report capacity this way. Other operating systems don't get it wrong like Windows does. Windows WRONGLY reports GiB as GB (and MiB for MB and so on). So when you see that your 500GB drive only has 465GB in Windows, it is because it is actually 465GiB, not 465GB like Windows says. And 465GiB is equal to 500GB. So the only misrepresentation according to accepted international standards happened in Microsoft Windows software. Windows Explorer just labeled the number wrong and confused you.

So maybe people should be suing Microsoft instead.

This mistake is also made in some semiconductor memory circles like the system RAM industry. Their memory is built in base 2 terms (gibibytes) but they label their packages with base 10 terms (gigabytes). So they are actually giving you more than they say they are. They continue to use broken terminology because they don't want to be the odd men out by marketing their stuff as gibi-this and mebi-that.
Reply to this comment
That's just the way it is.
by Crunchy Doodle December 5, 2007 10:20 AM PST
Modern flat-panel monitors and TV sets carry accurate sizing information. You can actually measure 37" on an advertised 37" HDTV set. But you could never do that with glass CRT TV sets. Does anyone remember how a box for a TV set had two sets of size information, one for the USA and one for Canada? How could a 27" TV be 29" in another country? And you couldn't measure 27" or 29" anywhere.

So the marketers of memory cards and sticks are using the same tortured logic used for disk drives. So what? A 2GB SD card from one vendor is likely to be very close in formatted capacity to another vendors 2GB SD card. Does it really matter that they mislead the public a little on the precise sizing? Does the typical memory device user have such tight requirements that a few missing KB, or even GB really mean anything? I don't think so.

Let the vendors have their little game. They seem to need it. I know what I'm really getting when I buy a 2GB SD card or an 80GB photo archival player. Nobody being fooled - not in reality.
Reply to this comment
Yet another example of USA legal 'experts' having no idea!
by Scott.Walker December 10, 2007 5:02 PM PST
This is just ridiculous and another example where those responsible for writing the laws have no idea what they're talking about - is it any wonder that "The internet is not just something you dump something on. It's not a big truck. It's, it's a series of tubes!"? (Sen. Ted Stevens on 'Net Neutrality', 2006)

When a device is sold as a gigabyte device, then it's size should be interpreted by the consumer as stated, not assumed to be something totally different, such as a gibibyte device. The consumer does not have the argument 'oh, I'm too stupid to know the difference, I thought I was getting X, and that's what I should have got', and if this is the angle they take, what the hell are they doing arguing the point in the first place!?
There's a reason why there's the Giga Binary prefix gibi, the Mega Binary prefix mebi etc. - it's because they ARE different. A mega-anything is a million, a giga-anything is a billion of them, I don't care if they're electronic devices or electrons themselves, a giga is an SI prefix which means 10^3 and that's why the guys in the know (ANSI/IEEE) came up with specific units such as KiB, MiB, etc. to address this issue. I can't believe that the people driving the US legal system are so clueless to their own standards (ANSI), not to mention international standards (IEEE).

But then again, USA's space exploration specialists did manage to slam a multimillion dollar probe (that's US$10^6 not US$2^20) into Mars because they forgot they were using imperial units like feet, gallons, (and probably kilobytes of 1024B) when they really should have been using metric units like, oh, I don't know, ...the rest of the planet!

Get with the times USA. You use a metric currency system which seems to drive everything you do (this law suit proves that!), why not update the rest of your units to SI before somebody over there realises that every time a megagallon of oil gets shipped to them from the middle east, they're not actually getting the 1,048,576 gallons of oil they expected to get... You just know that's gonna start a war!
Reply to this comment
(11 Comments)
  • prev
  • 1
  • next
advertisement

Making sense of Windows 7 upgrades

faq The basics and the fine print on Microsoft's options for those eyeing the next operating system from Redmond.
• Full Windows 7 coverage

Road Trip 2009: Big Sky Country

CNET News reporter Daniel Terdiman takes his car full of gadgets to the Rockies and the Great Plains in search of tech, science, nature, and more.
• America's Fortress: Cheyenne Mountain

About News Blog

Recent posts on technology, trends, and more.

Add this feed to your online news reader

advertisement
advertisement

Inside CNET News

Scroll Left Scroll Right