[mythtv-users] seagate giving refunds out

Sat Nov 3 07:41:51 UTC 2007

On 03/11/2007, Steve Hodge <stevehodge at gmail.com> wrote:
> On 11/3/07, Nick Morrott <knowledgejunkie at gmail.com> wrote:
> > On 03/11/2007, Steve Hodge <stevehodge at gmail.com> wrote:
> >
> > > If Joe Public don't know the difference they would surely assume that
> "giga" and "mega"
> > > would have the standard (decimal) meanings, would they not?
> >
> > Ask Joe Public what the giga- prefix means. Carry on until you find
> > someone who can answer. Ask Joe Public if they think all computer
> > products should measure storage capacity in the same way. Carry on
> > until you find someone who says no.
>
> Giga and mega a pretty widely understood. And consistency is certainly
> desirable. But it's not the harddisk manufacturers who are being
> inconsistent.

Outside the computing sphere, SI prefixes are pretty much set in
stone. Within computing, consistency is of paramount important, even
if it means being consistent with all other players in an industry
which uses SI prefixes consistently incorrectly. For a long time,
computing has used mega- and giga- to mean 2^20 and 2^30 and these are
generally accepted uses of these prefixes when referring to quantities
of bytes.

> > Computers use base 2. Hard drive manufacturers used to use base 2, but
> > now use base 10. I'm probably missing the point entirely ( the thread
> > on /. at
> http://hardware.slashdot.org/article.pl?sid=07/11/01/231250
> > echoes my points though),
>
> An appeal to Slashdot is hardly convincing I'm afraid :-)

It wasn't an appeal. More that I wasn't deluding myself by contrbuting
to the thread. At least I didn't post a link to Digg...

> > Suppose I have 8GB of RAM in my computer, and want to hibernate using
> > an available 8GB drive - can I do it? Why not?
>
> Because an "8 GB" RAM stick (or combo of sticks) does not hold 8GB. It holds
> 8GiB or 8.6 GB. The RAM manufacturer has inaccurately stated the size.
> And you can 't hibernate that amount of RAM to an 8GB drive whether drive
> manufacturers use GB or GiB because there is filesystem overhead to be taken
> into account.

I completely agree with your points. I'm not singling hard drive
manufacturers out in particular, but the general inconsistency that
exists, and that HDD manufacturers appear to be the most obvious 'odd
man out'. As capacities continue to grow, the discrepancies between
measuring in decimal bytes and binary bytes will continue to grow.

> > > > I think it's reasonable that Joe Public should be able to compare
> > > > apples with apples, and not have to worry about GiBs. I understand the
> > > > difference. I doubt my grandad does.
> > >
> > > GHz, Km. Your granddad would most likely expect 500GB to mean 500
> billion
> > > bytes. The only people who have a complaint here are people who are
> already
> >
> > I fail to see the connection between your two fundamentally different
> > units of measurement which are always base 10, and the topic of this
> > thread, which is using 2 different methods of quantitative measurement
> > (using base 2 and base 10) to represent the same fundamental quantity.
>
> 500GB is not a base 2 number, however you define "giga-". It's a base 10
> number. No product I've ever seen has ever quoted anything in base 2. What's
> your point?

Sorry for the poor explanation - here's a simple example, but I doubt
it'll add much:

512GB of data is generally understood to mean 2^39 bytes (i). 512GB of
HDD capacity is correctly (in SI terms) stated to mean 5.12 * 10^11
bytes (ii). Both capacities *could* be slapped on a storage product.
The latter correctly uses the giga- prefix to refer to 10^9, but uses
it inconsistently with respect to generally accepted terminology when
referring to computer storage capacity. Expressed in decimal, the
capacities are:

i) 549,755,813,888 bytes
ii) 512,000,000,000 bytes

There's a difference of 37,755,813,888 bytes (37.7GB, 35.2GiB), most
of which you'll notice as 'missing space' in the OS. FAT/FS overheads
aside, is it reasonable to expect to be able to store the data on the
HDD?

> My point is that giga-, mega-, kilo-, etc, have well know standard meanings.
> I don't see how you can dispute that. If you want to argue that it's
> bad/wrong/illegal for computer vendors to misuse those terms, fine. But it's
> not the drive manufacturers who are abusing the terminology.

I'm not disputing the meanings of the SI prefixes. I wish separate
binary units were in place and that all parties used them
consistently. The hard drive (and DVD-R) manufacturers use the
terminology correctly, but inconsistently with the rest of the
industry.

> > IANAL, I don't live in the land of the class action, and can only
> > admire its comedic legal system from afar, but I was under the
> > (obviously false) impression that a case still needed a valid reason
> > to go forward, where merit is in the hands of the judiciary, not the
> > plaintiff.
>
> The fact that a lawsuit has proceeded or even come to a particular
> conclusion is hardly proof that the right decision was made. In this case
> Seagate have settled so we can draw no conclusions as to their legal guilt
> or innocence.

Settlement always raise some eyebrows. It'll be interesting to see if
other manufacturers get hauled in to court in the future. They all
state their capacity policies (as did Seagate), and use the SI
predixes correctly and consistenly within the HDD industry. I guess it
comes down to reasonableness in the context of the computer industry
as a whole and  whether it generates consumer confusion.

> > I'm sure my Octogenarian granddad would just like to be able to buy a
> > hard drive and have its reported capacity in the OS agree with the
> > label on the disk. I'm guessing there's a fair bunch of other users
> > (geek and non-geek) who would agree.
>
> I don't disagree, but again, the drive manufacturer has used the appropriate
> terminology correctly. Besides, the OS available capacity is never going to
> match what's on the drive. Even before you format and lose a bunch of space
> to the file system you've lost the some space to partition tables.

Again I agree with you. I expect most people don't even care about the
technical reasons why capacities never exactly match, even if both
drive and OS report using the same scheme.

Cheers,
Nick

-- 
Nick Morrott

MythTV Official wiki:
http://mythtv.org/wiki/
MythTV users list archive:
http://www.gossamer-threads.com/lists/mythtv/users

"An investment in knowledge always pays the best interest." - Benjamin Franklin