-
Important Announcement for the TrueNAS Community.
The TrueNAS Community has now been moved. This forum will now become READ-ONLY for historical purposes. Please feel free to join us on the new TrueNAS Community Forums
Capacity
- Thread starter kafar
- Start date
- Status
danb35
Hall of Famer
- Joined
- Aug 16, 2011
- Messages
- 15,504
I know you know this, but for @kafar's benefit, if anyone's "lying", it's been computers for decades. "Kilo-", "mega-", "giga-", and "tera-" mean things--10^3, 10^6, 10^9, and 10^12, respectively. Computers have been misusing them to refer to 2^10, 2^20, 2^30, and 2^40. A while ago, hard drive manufacturers realized they could make their hard drives sound bigger, and still not technically be lying, if they used GB and TB correctly (i.e., to refer to base-10 values, not base-2). More recently, someone invented the GiB and TiB terminology to refer to the base-2 capacities, but it doesn't seem to have caught on very much.
Almost all operating systems, including FreeBSD (and thus FreeNAS), use the incorrect "computer" meaning--1 TB = 2^40 (or
1099511627776) bytes. Drive manufacturers use the correct, SI meaning of 1 TB = 10^12 (or 1000000000000) bytes. And hence most of the discrepancy between the stated disk capacity and the pool capacity.
jgreco
Resident Grinch
- Joined
- May 29, 2011
- Messages
- 18,680
That's basically only valid if you take the SI definitions as the only possible correct definitions. We'd been using the IEEE 100 nomenclature for many years - with the hard drive manufacturers cheerfully using them as well until they decided to "switch".
So when a unit comes to have an agreed meaning in the industry, because "kilobyte" always meant 1024 to ANYONE in the industry, and that's even documented in standards documentation produced by the industry (IEEE 100)... and then an attempt to rewrite that for the marketing benefit of hard drive manufacturers (and specifically shortchanging consumers in the process) ... I have a hard time seeing that as anything but deceptive. Western Digital and Seagate have both been sued on this front and chose to settle rather than try to defend the "SI" unit argument, so even the manufacturers don't believe their own BS.
I stand by what I said: it's hard drive manufacturers lying.
So when a unit comes to have an agreed meaning in the industry, because "kilobyte" always meant 1024 to ANYONE in the industry, and that's even documented in standards documentation produced by the industry (IEEE 100)... and then an attempt to rewrite that for the marketing benefit of hard drive manufacturers (and specifically shortchanging consumers in the process) ... I have a hard time seeing that as anything but deceptive. Western Digital and Seagate have both been sued on this front and chose to settle rather than try to defend the "SI" unit argument, so even the manufacturers don't believe their own BS.
I stand by what I said: it's hard drive manufacturers lying.
I'm sorry jgreco, but you are flat out wrong. Drive manufacturers aren't lying, a large number of consumers are just ignorant. Just because we used to do things one way is not a valid excuse to continue perpetuating a faulty system (see: metric system in almost every country except the US, Liberia, Burma, and the UK). Add to this the fact that with each scale we go up (K->M->G->T), the difference between the binary and SI prefixed values gets larger and larger, the "old school" usage makes less and less sense. This is only going to get worse as data scales continue to grow. There is a perfectly valid representation for both the binary and SI prefixes, i.e. GiB vs GB, recommended by IEC and NIST. Insisting on using the latter when you mean the former is only going to drag this issue on longer.
Heh.
SirMaster
Patron
- Joined
- Mar 19, 2014
- Messages
- 241
I also don't believe HDD manufacturers are lying. Kilo can't mean anything else except 1000. Whoever tried to call a kilo 1024 was just an idiot in the first place. I'm a Software Engineer and I work in the industry and this is how we have always referred to it.
Both NIST and IEEE have said for many many years that kilo = 1000 and kibi = 1024.
http://en.wikipedia.org/wiki/IEEE_1541-2002
http://www.physics.nist.gov/cuu/Units/binary.html
The only one who is lying is Microsoft in Windows where they call base 2 data counts by the base 10 name. But is anyone really surprised that Microsoft is not following the standard?
Maybe there was a period about 10 years ago or so when you could make a claim that HDD manufacturers were lying, but they have not been any time in the last several years. Regardless of how it was over a decade ago, this is how it is today which is what matters when talking about systems and software today.
Both NIST and IEEE have said for many many years that kilo = 1000 and kibi = 1024.
http://en.wikipedia.org/wiki/IEEE_1541-2002
http://www.physics.nist.gov/cuu/Units/binary.html
The only one who is lying is Microsoft in Windows where they call base 2 data counts by the base 10 name. But is anyone really surprised that Microsoft is not following the standard?
Maybe there was a period about 10 years ago or so when you could make a claim that HDD manufacturers were lying, but they have not been any time in the last several years. Regardless of how it was over a decade ago, this is how it is today which is what matters when talking about systems and software today.
- Joined
- Feb 15, 2014
- Messages
- 20,194
What OS doesn't call 2^10 "kilo"?
SirMaster
Patron
- Joined
- Mar 19, 2014
- Messages
- 241
Linux, BSD, Solaris, OSX...
Go ahead on your FreeNAS. Try "df -BGB" vs "df -BGiB".
They clearly know the difference between GB=Gigabyte and GiB=Gibibyte while Windows labels the data GB but the count clearly represents GiBs.
Windows is the only one I know that uses the wrong labels.
Last edited:
DJ9
Contributor
- Joined
- Sep 20, 2013
- Messages
- 183
jgreco
Resident Grinch
- Joined
- May 29, 2011
- Messages
- 18,680
Ooo. I'm a software engineer too, and I also work in the industry, and we still refer to 2^N. As a matter of fact, so do memory manufacturers, CPU manufacturers, most OS releases, and a vast majority of the industry. I take it by combining "we have always referred to it" and "period about 10 years ago" to mean that you're relatively new to the profession. Some of us have been working with computers much longer than that.
If you can say that it was true years ago, then obviously something changed. My opinion hasn't changed. I can't have been correct back then and suddenly I'm not correct now. Drive manufacturers are trying to change the units, and it isn't for any purpose other than to make their drives look better while underdelivering on capacity.
What's funny is that just a few decades ago a byte wasn't even defined as 8 bits! Things have changed over time, and I agree that the recent "argument" that has been manufactured by the hard drive manufacturers is their way of selling a smaller product while claiming its bigger.
- Status
