Brian Martinez ([info]cluebyfour) wrote,
@ 2008-05-02 00:04:00
Previous Entry  Add to memories!  Tell a Friend!  Next Entry
Entry tags:computers, courts, stupidity

how do you write "douchebag" in binary?
It would be disingenuous of me to say I hate lawyers; I'm married to one after all. But shit like this, involving a class-action suit against Creative Labs over the advertised capacity of their MP3 players, convinces me that some lawyers deserve to be shot in the face:

According to the settlement agreement, the lead plaintiffs, who filed their federal lawsuit in California, alleged that Creative had misled consumers by exaggerating the capacity of its MP3 players. The fraud allegation hinged mainly on two different definitions of gigabyte. According to the decimal definition (the only one I knew until today), a gigabyte is 1 billion (109) bytes. According to the binary definition, a gigabyte is 1,073,741,824 (230) bytes. While Creative used the decimal definition in its advertising, the settlement says, "certain computer operating systems report hard drive capacity using a binary definition." On those systems, a 20GB Creative Zen player would register as only 18.6GB or so, about 7 percent less than advertised.


Now it's true that hard drives use the decimal definition for a gigabyte when advertising their capacities. However, it's not "certain computer operating systems" that report hard drive capacity using the binary definition; it's all computer operating systems. Computer data is stored in binary format, so it can't report capacity in gigabytes using any other definition.

The rank stupidity of all this is that, whether you divide a hard drive's capacity by 1,000,000,000 or 1,037,741,824, it's still the same number of bytes. Only the divisor has changed. So Creative didn't exaggerate the capacity of their players at all; they simply failed to account for the fact that computers use the binary definition of a gigabyte. 20GB decimal = ~18.6GB binary. There's nothing misleading about this, except to consumers who are ignorant of how computers and data storage work. And I don't think Creative has an obligation to educate them; it just has to be truthful.

But in the hands of skilled trial lawyers, this inconsequential difference can be represented to a judge (who quite likely is also technologically ignorant) as a case of deceptive advertising, even fraud, and instead of wasting enormous time and money on defending themselves from this nonsense, Creative opted for a settlement. So if you own a Zen or other Creative MP3 player, you're entitled to buy a 1GB MP3 player at half-price, or 20% off any item ordered from Creative's online store.

For striking this courageous blow for consumers, the lawyers pocket $900,000. If it makes you feel better, in binary that would only be $878.9K.

Update: As [info]ilcylic demonstrates in a comment below, an OS (Linux in this case) can report hard drive capacity in decimal format if it so chooses.  But given that volatile memory storage will always be reported in binary, I'm not sure that resolves the confusion issue.


(Post a new comment)


[info]agentsteel53
2008-05-02 06:37 am UTC (link)
I disagree with you here.

t's not "certain computer operating systems" that report hard drive capacity using the binary definition; it's all computer operating systems.

I'm pretty computer-savvy and it was quite the surprise to realise that something sold as "320GB" is really only "289GB" of actual storage, because of an overloaded, and expediently milked, definition of "GB".

I consider this to be the "when in Rome" argument ... a gigabyte, in the world of computers, is 2^30 bytes - to claim it is 10^9 bytes is an error.

to provide an analogue to the discussion in another realm: imagine someone intentionally conflating the troy and the avoirdupois ounces as measures of gold for his own personal benefit. I would call that fraudulent, even if "everyone knows" that the general definition of ounce that is used to measure everything from carrots to cocaine (avoirdupois) is different from the one specialist use (troy) that is applicable to just one specific context.

Edited at 2008-05-02 06:38 am UTC

(Reply to this)(Thread)


[info]ilcylic
2008-05-02 02:22 pm UTC (link)
Clearly, the easiest solution for the manufacturers to avoid further lawsuits (though, it'll kinda suck for 97% of the consumers, but hey, what do the class-action lawyers care?) is to print drive sizes in bytes.

Seagate 320 (model number, not claim of size in any unit, SI or digital)
320,000,000,000 bytes capacity.

Your computer will still report it to you as "289 GB", but now you'll have to complain to your OS manufacturer about it.

(Reply to this)(Parent)(Thread)


[info]agentsteel53
2008-05-02 02:47 pm UTC (link)
Heh, nice way of passing the buck: "blame your OS".

I will have to look at some of my hard drives at home to see what they claim (I have a few boxes lying around) - if they just have "model number 320" or if they do actively state "320 gigabytes" and/or "320GB".

50+ years of computer history state that a gigabyte *is* 2^30 bytes. It's how it is defined. An inch *is* 2.54 centimeters. The fact that "giga" is an overloaded operator doesn't invalidate this definition. When it prefixes non-computer terms like "ton", it's 10^9. When it prefixes computer terms, it's 2^30. There is no ambiguity here - confusion, possibly, in the eye of an underinformed beholder, but nothing wrong with the definition itself.

(Reply to this)(Parent)


[info]cluebyfour
2008-05-02 03:36 pm UTC (link)
In your last paragraph, it would only be fraud if the mass of gold changed (2oz avoirdupois is less in mass than 2 troy oz). That's not the case with hard drives--you don't suddenly lose bytes because you used a binary definition. 20,000,000,000 is 20 billion bytes whether you write it as 20GB (decimal) or 18.6GB (binary).

If you deliberately represented it as 20GB (binary) when it fact it wasn't then yes, that would be a problem. But Creative never did that. In fact, two years before the lawsuit was filed, they had added clarification to the packaging that their measurement was based on 1 GB = 1,000,000,000 bytes. No deception, no injury to consumers. The lawyers in this case are just douchebags looking for an easy paycheck.

(Reply to this)(Parent)(Thread)


[info]agentsteel53
2008-05-02 03:51 pm UTC (link)
My concern is that "20GB (decimal)" is an invalid construct, because as far as I know, the prefix "giga", when modifying the term "bit" or "byte" is *defined* to be a binary measurement.

If one writes "20GB (decimal)" - with modifier in parentheses present - then I suppose that can be resolved to "20 x 10^9 bytes", but if one writes "20GB" - no modifier present - then that, by definition, resolves to "20 x 2^30 bytes".

If it *is* clarified on the packaging, then you are correct, that it is a matter of sloppy research on the consumers' part, but to me even that would be douchebaggy behaviour, similar to selling "one dozen hot dogs" and in fine print saying "we define a dozen to be ten".

(Reply to this)(Parent)(Thread)


[info]cluebyfour
2008-05-02 04:10 pm UTC (link)
Here, read this; it explains what I'm trying to say:

http://en.wikipedia.org/wiki/Binary_prefix#Consumer_confusion

I get what you're saying Jake, but there's about 30 years of precedent on this, and while consumers might get confused, it's not an issue of fraud or anything that requires a freakin' class-action suit to resolve.

(Reply to this)(Parent)(Thread)

(no subject) - [info]agentsteel53, 2008-05-02 04:13 pm UTC

[info]tcpip
2008-05-02 08:34 am UTC (link)

011001000110111101110101011000110110100001100101011000100110000101100111

Well, you asked :-)

(Reply to this)(Thread)


[info]cluebyfour
2008-05-02 03:37 pm UTC (link)
Are you sure? Can you spell check that?

(Reply to this)(Parent)(Thread)


[info]tcpip
2008-05-02 10:37 pm UTC (link)

http://www.theskull.com/javascript/ascii-binary.html

(Reply to this)(Parent)


[info]apotheon
2008-05-02 04:27 pm UTC (link)
What's the extra bit? Parity?

(Reply to this)(Parent)


[info]deadpansev
2008-05-02 12:05 pm UTC (link)
I think it is better if companies would use the binary definition, but having a lawsuit over it seems a bit absurd. What should have happened is a geek revolt where those who knew the difference start to demand that either the company start printing the binary capacity on the drives or they won't buy the drives. The most likely compromise in the face of customer pressure would have been creative and other companies printing both numbers on the package.

(Reply to this)(Thread)


[info]cluebyfour
2008-05-02 03:41 pm UTC (link)
It's cleaner of course to say "200GB" instead of "192.7GB" on the box. I think just stating clearly what they're basing it on is enough. But you're right, adding both number would probably resolve the issue permanently.

I notice that in Windows Explorer, if you right-click the drive and bring up properties, it will display the total number of bytes on the disk, and then report it as gigabytes (using the 2^30 format).

(Reply to this)(Parent)


[info]flainn
2008-05-02 02:25 pm UTC (link)
This is one reason why there has been a push, in the last ten years, to use GB (gigabyte) to mean 10^9 bytes, and GiB (gibibyte) to mean 2^30 bytes. However, it hasn't really caught on, and GB continues to mean one or the other or both, depending on the situation.

The really bad thing about this is that it sets precedent; all drive manufacturers are equally culpable in this regard, although I believe some at least try to educate customers as to the difference.

(Reply to this)(Thread)


[info]agentsteel53
2008-05-02 02:54 pm UTC (link)
I thought that "the situation" was clearly defined, though. "Giga" as a prefix represents 2^30 for a particular set of suffixes: "bit", "byte", and some others I am not aware of, but I'm sure I could find out... and 10^9 for the rest. Is there any suffix for which it is sufficiently ambiguously defined as to be a Real Problem, as opposed to someone not R'ing TFM?

(Reply to this)(Parent)(Thread)


[info]flainn
2008-05-02 03:29 pm UTC (link)
According to the SI system, "kilo," "mega," "giga," etc, only refer to base-ten numbers. Thus giga should always mean 10^9.

However, due to sloppiness on just about everyone's part (mine included), it's been used interchangeably to mean 10^9 or 2^30. At least, in the circles I frequent.

About ten years ago when I was starting my CS program, "they" (I'm not sure whether this was SI or just the ACM) started pushing for the use of KiB, MiB, and GiB (and others, though I haven't seen them used) to denote the binary equivalents of each. Had they been consistently used by people and drive manufacturers, we wouldn't have this problem; but frankly saying "kibibyte" and "gibibyte" just sounds silly.

So no, as far as I know, "giga" should always mean 10^9, regardless of the suffix; and "gibi" should always mean 2^30, regardless of suffix. But note I said should, not does.

(Reply to this)(Parent)(Thread)

(no subject) - [info]agentsteel53, 2008-05-02 03:45 pm UTC
(no subject) - [info]flainn, 2008-05-02 03:50 pm UTC
(no subject) - [info]agentsteel53, 2008-05-02 04:07 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 04:20 pm UTC
(no subject) - [info]agentsteel53, 2008-05-02 04:22 pm UTC
(no subject) - [info]apotheon, 2008-05-02 04:42 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 04:13 pm UTC
(no subject) - [info]apotheon, 2008-05-02 04:44 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 05:06 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 05:12 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 05:24 pm UTC
(no subject) - [info]apotheon, 2008-05-02 05:44 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 03:58 pm UTC
(no subject) - [info]agentsteel53, 2008-05-02 04:02 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 04:15 pm UTC
(no subject) - [info]agentsteel53, 2008-05-02 04:19 pm UTC
(no subject) - [info]apotheon, 2008-05-02 04:57 pm UTC
(no subject) - [info]agentsteel53, 2008-05-02 05:01 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 04:14 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 04:15 pm UTC
(no subject) - [info]apotheon, 2008-05-02 04:59 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 04:12 pm UTC
(no subject) - [info]agentsteel53, 2008-05-02 04:17 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 04:23 pm UTC
(no subject) - [info]agentsteel53, 2008-05-02 04:25 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 04:26 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 04:45 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 04:50 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 05:15 pm UTC
(no subject) - [info]apotheon, 2008-05-02 05:25 pm UTC
(no subject) - [info]apotheon, 2008-05-02 05:11 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 05:21 pm UTC
(no subject) - [info]apotheon, 2008-05-02 05:27 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 04:24 pm UTC
(no subject) - [info]agentsteel53, 2008-05-02 04:33 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 04:46 pm UTC
(no subject) - [info]agentsteel53, 2008-05-02 04:49 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 04:59 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 05:01 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 05:17 pm UTC
(no subject) - [info]apotheon, 2008-05-02 05:39 pm UTC
(no subject) - [info]apotheon, 2008-05-02 05:30 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 07:22 pm UTC
(no subject) - [info]apotheon, 2008-05-02 07:25 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 07:27 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 07:29 pm UTC
(no subject) - [info]apotheon, 2008-05-02 07:36 pm UTC

[info]cluebyfour
2008-05-02 03:42 pm UTC (link)
And computer OEMs do it too, of course. They also tend to have that "1 GB = 1,000,000,000 bytes" disclaimer all over the place.

(Reply to this)(Parent)


[info]apotheon
2008-05-02 04:39 pm UTC (link)
That cheesy construction "gibi" annoys the hell out of me. I'd have been happier if they went with GB (gigabyte) and FGB (faux gigabyte) -- or just to overload "GB/gigabyte". I refuse to say "gibibyte" with a straight face.

(Reply to this)(Parent)


[info]mama_gryphon
2008-05-02 04:28 pm UTC (link)
Freaking geeks! :P

It's a good lawsuit because, obviously, there are people who care about it. I couldn't care less about it personally. Frankly, I have better things to do like defend my client's and their children from their pedophilia ex's. However, if people care this much, then one of them takes it to a lawyer and the lawyer researches and finds more people willing to sign on, then how are they douchebags for prosecuting the case? (Yes, it's a lovely run-on sentence.) All you geeks seem to care how the heck the thing is defined. You'll sit and argue over it, etc. So why shouldn't a lawyer file a suit to 'protect the consumer interest' in this case?

(Reply to this)(Thread)


[info]cluebyfour
2008-05-02 04:31 pm UTC (link)
Because there doesn't seem to be any evidence that the consumer has been screwed.

And sure we're geeks about it, but the definition of a "gigabyte" is the very issue on which the lawsuit is based. Which I think is silly.

(Reply to this)(Parent)(Thread)


[info]mama_gryphon
2008-05-02 04:38 pm UTC (link)
Of course, the consumer got screwed. The consumer was told they could get 1 Gb (or whatever) on that damn MP3 player and they can't. They bought the damn things because they wanted to load up 1 Gb of whatever. They can't upload that much. And, it sounds like, it's not just off by a byte or two. Maybe that 30-50 bytes of space is the difference between having a tower shooter day and chilling at Starbucks for someone. It may be silly for you but it pissed someone off and a lawyer did something about it. Not every lawsuit filed has overwhelmingly compelling reasons, but some do effect a portion of society and a portion of that portion cares. Now are there better things the court can spend time on? Yes, but then the judges wouldn't get enough sleep and they'd be cranky.

(Reply to this)(Parent)(Thread)

(no subject) - [info]ilcylic, 2008-05-02 04:49 pm UTC
(no subject) - [info]mama_gryphon, 2008-05-02 04:53 pm UTC
(no subject) - [info]ilcylic, 2008-05-02 05:13 pm UTC
(no subject) - [info]apotheon, 2008-05-02 06:02 pm UTC
(no subject) - [info]apotheon, 2008-05-02 05:52 pm UTC
(no subject) - [info]mama_gryphon, 2008-05-02 06:51 pm UTC
(no subject) - [info]cluebyfour, 2008-05-02 06:58 pm UTC
(no subject) - [info]mama_gryphon, 2008-05-02 07:13 pm UTC
(no subject) - [info]apotheon, 2008-05-02 07:16 pm UTC
(no subject) - [info]mama_gryphon, 2008-05-02 07:35 pm UTC
(no subject) - [info]apotheon, 2008-05-02 07:58 pm UTC
(no subject) - [info]mama_gryphon, 2008-05-02 08:40 pm UTC
(no subject) - [info]apotheon, 2008-05-02 10:39 pm UTC

[info]apotheon
2008-05-02 04:35 pm UTC (link)
However, it's not "certain computer operating systems" that report hard drive capacity using the binary definition; it's all computer operating systems.


Well . . . when I use the df -h command, that's correct. When I use the df -H command, I get a "decimal GB" total. I guess it depends on how you look at it, really.

Creative could have won this one without breaking a sweat -- if its lawyers and technical consultants to the legal team actually knew anything about the technical side of things. I'm sure that with a corporation as large as Creative, it has a stable of lawyers on retainer, obviating the need for any particular legal expenditures except during specific, key parts of the litigative process. As such, this really shouldn't have cost Creative Labs all that much money to fight.

I suspect it was just the fear of an unfavorable judgment that caused the company to settle -- and that's an unwarranted fear if you have anyone with half a technical head on his shoulders consulting for their defense. Worst-case scenario, the judgment goes against Creative, gets overturned almost immediately on appeal, and the judge returning an unfavorable ruling looks like an unlettered ass.

(Reply to this)(Thread)

part two
[info]apotheon
2008-05-02 07:59 pm UTC (link)
(continuation . . .)

Just because you don't see what their point is doesn't mean they don't have one. Just because you think it's silly doesn't mean there aren't good grounds for filing suit.


Who says I don't see their point? I just think their point is bullshit, and someone with more technical knowledge than them should have mentioned this to them somewhere along the way (someone retained by their lawyers to consult on the matter, if nobody else).

Just because you think it's silly doesn't mean there aren't good grounds for filing suit.


I get the impression you're equating "has legal grounds to do so" with "should do so". I make very clear distinctions between "legal sanction" and "ethical right", personally.

Just because Creative got sued doesn't mean their drop in the proverbial bucket is going to make a large dent in the insurance or business world.


Now you're assigning notions to me that I never expressed. Please stop.

Even if you have a copy of the Complaint, you would still not have copies of the evidence to be presented, etc. You're going off of your knowledge and what's reported. People too easily make assumptions based on that.


I already said I made an assumption, and that my previous statements were predicated upon that assumption. If you're going to continue to disagree with me, please do so on matters where we still disagree. Thrashing me over a point I've already conceded does nobody any fucking good.

The bottom line from a business and legal perspective is if it's certified for a class action then Creative was right in reaching a settlement and getting rid of it.


. . . or Creative did a piss-poor job of opposing that certification. Either way, the actual settlement decision was a business decision -- you're right about that.

Creative does get a win out of it because the settlement makes people shut up and move on instead of beating the horse dead twenty feet into the ground for three to five years.


A win would have been if the case was never filed, or never certified for class action, because someone realized the stupidity of the basis for the suit (assuming what I know is the relevant facts). This is a loss that's less significant than losing in court would have been. Again, it's a matter of relative levels of getting screwed, not of screwed/not-screwed.

(Reply to this)(Parent)(Thread)

Re: part two
[info]mama_gryphon
2008-05-02 08:45 pm UTC (link)
I get the impression you're equating "has legal grounds to do so" with "should do so". I make very clear distinctions between "legal sanction" and "ethical right", personally.


Just because Creative got sued doesn't mean their drop in the proverbial bucket is going to make a large dent in the insurance or business world.


Now you're assigning notions to me that I never expressed. Please stop.


Your "clear distinctions" were not made clear in the comment I read.

You stated there would be reprecussions in the insurance industry because of this case. If you were making a larger generalization I didn't read it that way.

As for attributing notions -- I thought I was having a discussion and you threw a few out yourself.

(Reply to this)(Parent)(Thread)

Re: part two - [info]apotheon, 2008-05-02 10:41 pm UTC

[info]apotheon
2008-05-02 08:00 pm UTC (link)
Oh, for crying out loud -- LJ somehow broke the connection between my reply and its parent.

(Reply to this)(Parent)(Thread)


[info]deadpansev
2008-05-02 10:31 pm UTC (link)
I had this happen to me, and discovered that the refried style I am using does this and it is known bug. The fix is to turn off the setting to keep your journal style on comment pages.

(Reply to this)(Parent)(Thread)

(no subject) - [info]apotheon, 2008-05-02 10:42 pm UTC

Create an Account
Forgot your login?
Login w/ OpenID
English • Español • Deutsch • Русский…