Template talk:Bit and byte prefixes

From Wikipedia, the free encyclopedia
Jump to: navigation, search

"Note that the IEC names are defined only up to exbi-, corresponding to the SI prefix exa-. The two SI prefixes zetta- (1021) and yotta- (1024) have no corresponding IEC binary prefixes, though the obvious continuation would be zebi- (Zi = 270 = 10007 × 1.180 591 620 717 411 303 424) and yobi- (Yi = 280 = 10008 × 1.208 925 819 614 629 174 706 176)."

This standard was developed specifically to meet industry’s needs in data processing and data transmission. It eliminates confusion by setting out the prefixes and symbols for the binary, as opposed to decimal, multiples that most often apply in these fields. Zebi and yobi are the latest in this evolution. The term zebi means “two to the power 70”. The term yobi means “two to the power 80”. [1]

Byte[edit]

Shouldn't byte be on this list? --Domthedude001 18:26, 22 June 2006 (UTC)

Bit is the classification of a single digit 0 or 1, represented as a lowercase "b". Byte is the classification of eight bits to make a character, represented as an uppercase "B". — Preceding unsigned comment added by 173.170.1.32 (talk) 03:33, 11 May 2014 (UTC)

See also[edit]

Capital or small k for kilobyte[edit]

When represents kilo, use small k, ever, because the Capital is used for represents the Kelvin units. —Preceding unsigned comment added by 187.38.167.125 (talk) 22:08, 23 February 2011 (UTC)


I dare to say that in 95% of the cases the abbreviation KB means 1024 bytes and that in less than 5% of the publications this is spelled as kB. The expression kB to mean 1000 bytes is even rarer. This just means that in the context of bytes the SI small k is just irrelevant. −Woodstone 17:43, 20 February 2006 (UTC)

Agreed. This is explained in the kilobyte article that the abbreviation links to. — Omegatron 20:33, 20 February 2006 (UTC)

Ok. So the template needs to have small k, but the explanation can say KB is common practice. −Woodstone 20:41, 20 February 2006 (UTC)

Yeah, the template is showing the standards, not common practice, which is too ambiguous for a concise table, anyway. — Omegatron 21:44, 20 February 2006 (UTC)

There's a wierd marketing convention that RAM manufactures use 1GB=2^30, while disk manufacturers use 1GB=10^9. See gigabyte. (I wonder what flash memory makers do.) Might be worth a link back to gigabyte. --Nagle 23:42, 28 March 2006 (UTC)

Once a long time ago I got it in my head that kb/s, mb/s (both lowercase) was a good/clear convention for kilobits/megabits per second - but in the end it was just way clearer for all involved to spell out a few more letters -- kbits/mbits. 74.103.98.163 18:31, 18 January 2007 (UTC)

That's like saying "When represents mega, use small m, ever, because the Capital is used for represents the Molar units." or "When represents milli, use big M, ever, because the lower case is used for represents the metre units." JIMp talk·cont 05:48, 5 March 2012 (UTC)

Clarification to table[edit]

The table as it was was misleading as it suggested that, for example, Megabyte had only the SI definition, and didn't take into account common usage of the binary type. I have expanded the table to make it clear that Megabyte, kilobyte etc can mean both binary and SI, whereas the new Mebibyte definitions relate to the binary numbers only. I hope everyone agrees that this helps to clarify the situation, which in itself, is unclear! --Rebroad 12:21, 10 March 2006 (UTC)

Parenthesis[edit]

I think that the parenthesis as they are used now in the left hand side quantity column are confusing. It doesn't seem clear to me that the "(SI standard meaning)" part from the row above applies to the quantity section. What about something like:

Quantities of bytes
SI prefix quantities Binary prefix quantities
from IEC 60027-2


Name


Symbol
Value in
Popular
Usage
Value in
Standard
SI


Name


Symbol


Value
kilobyte kB 210 103 kibibyte KiB 210
megabyte MB 220 106 mebibyte MiB 220
gigabyte GB 230 109 gibibyte GiB 230
terabyte TB 240 1012 tebibyte TiB 240
petabyte PB 250 1015 pebibyte PiB 250
exabyte EB 260 1018 exbibyte EiB 260
zettabyte ZB 270 1021 zebibyte ZiB 270
yottabyte YB 280 1024 yobibyte YiB 280

This way each numerical value has the usage-style directly above it, which seems clearer to me. Downside is that the table is slightly bigger. Comments? --Jakohn 19:21, 21 June 2006 (UTC)

It's quite big. — Omegatron 19:43, 21 June 2006 (UTC)
Quantities of bytes
SI prefixes Binary prefixes
(IEC 60027-2)

Name
(Symbol)
Value in
Popular
Usage
Value in
Standard
SI

Name
(Symbol)


Value
kilobyte (kB) 210 103 kibibyte (KiB) 210
megabyte (MB) 220 106 mebibyte (MiB) 220
gigabyte (GB) 230 109 gibibyte (GiB) 230
terabyte (TB) 240 1012 tebibyte (TiB) 240
petabyte (PB) 250 1015 pebibyte (PiB) 250
exabyte (EB) 260 1018 exbibyte (EiB) 260
zettabyte (ZB) 270 1021 zebibyte (ZiB) 270
yottabyte (YB) 280 1024 yobibyte (YiB) 280
(Removed infobox class from both demos to make it easier to compare.) Shrunk a bit, now somewhat skinnier than current live version, just 2 lines longer. Seem better? --Jakohn 20:16, 21 June 2006 (UTC)
How 'bout something a little simpler than that, maybe something like this? — SheeEttin {T/C} 22:30, 21 June 2006 (UTC)
Quantities of bytes
Popular use
(SI standard meaning)
Binary prefix standards
from IEC 60027-2
Name Symbol Quantity Name Symbol Quantity
kilobyte kB 210 (103) kibibyte KiB 210
megabyte MB 220 (106) mebibyte MiB 220
gigabyte GB 230 (109) gibibyte GiB 230
terabyte TB 240 (1012) tebibyte TiB 240
petabyte PB 250 (1015) pebibyte PiB 250
exabyte EB 260 (1018) exbibyte EiB 260
zettabyte ZB 270 (1021) zebibyte ZiB 270
yottabyte YB 280 (1024) yobibyte YiB 280
Thats confusing as well because the parenthesis are ambiguous. In previous examples it was clear that they denoted a separate way of measuring, which is really what this whole chart is about. The purpose is to denote two different ways of measuring bytes. This version of the chart may be more compact, but it does not achieve the goal of making it easier to understand the difference between the two systems.12.135.134.146 00:37, 30 August 2006 (UTC)
I like this version. See Template:Quantities of bits and Template:Bit rates as well. Keep in mind that "popular use" mostly applies to bytes, and not either of those two. — Omegatron 00:42, 30 August 2006 (UTC)

No toya-[edit]

There's no toya-. Indeed, the claimed abbreviation "T" would conflict with "Tera". I found no citation for this claimed prefix, so I junked it. Dwheeler 00:41, 7 February 2007 (UTC)

No byte[edit]

To bring up an old ignored topic, is there any particular reason why byte is not on the list? --Dlevenstein 16:18, 1 April 2007 (UTC)

Because it's quantities of bytes. A byte isn't a quantity of itself, is it? (And don't get all "reflexive property" on me, either, you know what I mean.) — SheeEttin {T/C} 20:23, 3 April 2007 (UTC)

Vandalism[edit]

Nice to know somebody enjoys tampering with the template (x-ray specs for exabyte? Honestly...). --Aerodotus 01:47, 4 July 2007 (UTC)

1024 vs 1000[edit]

I'm glad the table finally recognizes the existence of the 1,024 as the byte growth factor in binary. However, I changed the inaccurate denotation of "Legacy use" in the table header to "Binary prefix." In point of fact, "legacy" in this context is definted as "Obsolete; of or pertaining to old or outdated computer hardware, software, or data that, while still functional, does not work well with up-to-date systems." (Random House Unabridged Dictionary, 2006).

While the 1,024 specification does qualify as "old," the fact that it is still in dominant use in nearly all PC hardware and software (including operating systems like Windows XP/Vista and Unix and byte-specific software programs like Nero 7+), proves that it is in fact neither "obsolete" or "outdated." And the "does not work well with up-to-date systems" statement doesn't work since virtually all "up-to-date systems" are still being built on the 1,024 specification.

The only difference today is that most of these systems and hardware are not being *advertised* under this specification, even though they are being *built* under it. This is nothing more than a cheap marketing ploy to make their products appear to have more byte space than they actually do.

For example, single-layer DVD+/-R discs are commonly advertised at "4.7 GB". But if you put the disc into your drive, I'd challenge you to find a single piece of software or operating system that would recognize it as containing 4.7 GB of free space. If you try to burn 4.7 GB of data onto the disc, even with overburning enabled, you're likely to get an error message. This is because 4.7 / 1.024 (MB) / 1.024 (KB) / 1.024 (B) is approximately equal to 4.377; which, surely enough, is roughly how much space most operating systems and software would recognize on the disc (if you have a blank DVD+/-R disc, stick it in your drive and see for yourself if you don't believe me). Therefore, if anything, it is in fact the 1 GB = 1000 MB (etc) standard that "does not work well with up-to-date systems."

The reason I'm making such a big deal of this is because this marketing tactic already creates a lot of confusion among consumers. Look on Google and you'll find forum posts all over the place from people trying to figure out why their burning software won't let them burn 4.6 GB onto their 4.7 GB discs. This misleading practice therefore harms consumers, even those who are otherwise well-informed. If Wikipedia legitimizes this practice by, among other things, portraying the "1024" specification as "legacy", then it will only lead to further victimization of the consumer and make it more difficult for people to get the information needed to understand why this discrepancy exists. 76.28.188.174 (talk) 00:09, 17 December 2007 (UTC)

However common it may be to use K, M etc in conjunction with bytes to mean powers of 2, that does not make them binary prefixes, since in most other applications they are not binary. Perhaps a better term in the table header would be "binary usage". −Woodstone (talk) 09:06, 17 December 2007 (UTC)
The gigabyte is an ambiguous unit and the template makes that perfectly clear. If a round plug does not fit into a square hole it is not the fault of the plug - it is the fault of the person or organisation responsible for designing the interface, in this case the computer industry as a whole. If you can't see that I suggest you read the section entitled Consumer Confusion in the GB article. Thunderbird2 (talk) 16:56, 17 December 2007 (UTC)
I was the one who changed this from "historical use" to "legacy use", with the idea that it is not historical until it is in the past but active legacy systems abound. My preference would have been to automatically understand the binary meaning before either bit or byte and to forbid the decimal interpretation (what is an 8-bit byte doing with decimal prefixes, anyway?), as the IEEE did not too long ago, but IEC and several other groups, including the IEEE, have decided otherwise, so this seems like a lost cause, however much the confusion remains because decimal prefixes are not distinguishable from their other use as binary prefixes, and however nauseating the binary names (I don't see myself using them anytime soon). But to the point: I found it difficult to choose between historical, legacy, mixed, binary, deprecated, and maybe some others that I don't remember right now (entrenched?), and I thought legacy was the least inappropriate (my reference, Merriam-Webster's online dictionary, gives a more neutral definition than the one above). Actually, I hope the legacy use doesn't go away anytime soon. It is all very unfortunate, of course. RFST (talk) 05:07, 19 December 2007 (UTC)
The IEC prefixes have less than 1% usage in the real world. The table needs to make it clear that the use of KB, MB GB etc are used in the binary powers of two sense. The Binary usage is not sufficient for this purpose. If anything the prefixes KB, MB, GB etc in powers of two need to be shown first in the table and the IEC prefixes moved to a separate table because that reflects real world consensus. And it's not "legacy use" it's more like "common use" and actually "standard use" since the terms are defined by the JEDEC. Fnagaton 11:27, 27 January 2008 (UTC)

(Outdent) Apart from being innaccurate by missing the common use data the table is also too wide and makes the formatting on MOSNUM a bit messy. I propse a better table below:

Quantities of bytes
Common use
(Somtimes known as SI)
IEC 60027-2
Name Symbol Quantity Name Symbol Quantity
kilobyte KB (kB) 210 (103) kibibyte KiB 210
megabyte MB 220 (106) mebibyte MiB 220
gigabyte GB 230 (109) gibibyte GiB 230
terabyte TB 240 (1012) tebibyte TiB 240
petabyte PB 250 (1015) pebibyte PiB 250
exabyte EB 260 (1018) exbibyte EiB 260
zettabyte ZB 270 (1021) zebibyte ZiB 270
yottabyte YB 280 (1024) yobibyte YiB 280

This table includes the common use along with the decimal use making it clear which is which and reflects the consensus in the real world. It also includes the neologisms from IEC. Lastly it is narrower than the existing table. Fnagaton 11:44, 27 January 2008 (UTC)

This proposal makes the suggestion that the SI prefixes can have a binary meaning. It even suggests that powers of two might be equal to powers of 10. That is completely false and should not be encouraged. Some of the same letters (K, M, G) are sometimes used in binary meaning in specific contexts, but that should be kept clearly separate form their SI values. −Woodstone (talk) 12:14, 27 January 2008 (UTC)
OK Then how about:
Quantities of bytes
SI prefix quantities IEC 60027-2


Name


Symbol
Common
use
SI

Name


Symbol


Value
kilobyte KB/kB 210 103 kibibyte KiB 210
megabyte MB 220 106 mebibyte MiB 220
gigabyte GB 230 109 gibibyte GiB 230
terabyte TB 240 1012 tebibyte TiB 240
petabyte PB 250 1015 pebibyte PiB 250
exabyte EB 260 1018 exbibyte EiB 260
zettabyte ZB 270 1021 zebibyte ZiB 270
yottabyte YB 280 1024 yobibyte YiB 280

Fnagaton 12:24, 27 January 2008 (UTC)

How about using the existing version, but just scratching all the fluff:

Quantities of bytes
Value Name Value Name
Binary usage
SI prefixes   IEC prefixes
10001 kB kilobyte KB 10241 kibibyte KiB
10002 MB megabyte MB 10242 mebibyte MiB
10003 GB gigabyte GB 10243 gibibyte GiB
10004 TB terabyte TB 10244 tebibyte TiB
10005 PB petabyte PB 10245 pebibyte PiB
10006 EB exabyte EB 10246 exbibyte EiB
10007 ZB zettabyte ZB 10247 zebibyte ZiB
10008 YB yottabyte YB 10248 yobibyte YiB

Woodstone (talk) 12:27, 27 January 2008 (UTC)

It still doesn't specifically include the information from real world common use that kilobyte/KB etc are commonly used in powers 2. It ignores real world consensus. My proposal corrects that omission while yours does not. Fnagaton 12:30, 27 January 2008 (UTC)
It is not common usage; it is only usage in certain cases (mostly memory chips). So "binary usage" is a better reflection of reality. I have added to the kB right after the decimal numbers and made it KB to the right. −Woodstone (talk) 12:46, 27 January 2008 (UTC)
Both of your proposals improve on the structure of the existing template, which I find confusing. I prefer Woodstone's over Fnagaton's for two reasons:
  • I disagree with the heading "common use" for the binary sense of the SI prefixes, because the decimal one is just as common. What's wrong with "binary use"?
  • "K" is not an SI prefix.
Thunderbird2 (talk) 12:42, 27 January 2008 (UTC)


That proposal still does not make it clear that kilobyte can be 1024 bytes etc. As I've shown before kilobyte/megabyte etc is common use because >99% of real world sources use the terms whereas <1% use IEC terms. That is real world consensus for common. If you don't like the "SI prefixes" how about:
Quantities of bytes
Binary prefix quantities IEC 60027-2


Name


Symbol
Binary
use
SI

Name


Symbol


Value
kilobyte KB/kB 210 103 kibibyte KiB 210
megabyte MB 220 106 mebibyte MiB 220
gigabyte GB 230 109 gibibyte GiB 230
terabyte TB 240 1012 tebibyte TiB 240
petabyte PB 250 1015 pebibyte PiB 250
exabyte EB 260 1018 exbibyte EiB 260
zettabyte ZB 270 1021 zebibyte ZiB 270
yottabyte YB 280 1024 yobibyte YiB 280

Fnagaton 13:11, 27 January 2008 (UTC)


That one gives the impression that SI prefixes are binary. How about this?

Quantities of bytes
Common use quantities IEC 60027-2


Name


Symbol
Binary
use
Decimal use

Name


Symbol


Value
kilobyte KB/kB 210 103 kibibyte KiB 210
megabyte MB 220 106 mebibyte MiB 220
gigabyte GB 230 109 gibibyte GiB 230
terabyte TB 240 1012 tebibyte TiB 240
petabyte PB 250 1015 pebibyte PiB 250
exabyte EB 260 1018 exbibyte EiB 260
zettabyte ZB 270 1021 zebibyte ZiB 270
yottabyte YB 280 1024 yobibyte YiB 280

Thunderbird2 (talk) 13:39, 27 January 2008 (UTC)

That one gives the impression that SI prefixes are binary. How about this?

I'd prefer this for formatting reasons of the "Decimal use" being too wide:
Quantities of bytes
Common use quantities IEC 60027-2


Name


Symbol
Binary
use
Decimal
use


Name


Symbol


Value
kilobyte KB/kB 210 103 kibibyte KiB 210
megabyte MB 220 106 mebibyte MiB 220
gigabyte GB 230 109 gibibyte GiB 230
terabyte TB 240 1012 tebibyte TiB 240
petabyte PB 250 1015 pebibyte PiB 250
exabyte EB 260 1018 exbibyte EiB 260
zettabyte ZB 270 1021 zebibyte ZiB 270
yottabyte YB 280 1024 yobibyte YiB 280

Fnagaton 13:42, 27 January 2008 (UTC)

Looks good to me. What do others think? Thunderbird2 (talk) 14:04, 27 January 2008 (UTC)

How about this slightly condensed form:

Quantities of bytes
Common prefix Binary prefix
Name Symbol Decimal
SI
Binary
JEDEC
Name Symbol Binary
IEC
kilobyte KB/kB 103 210 kibibyte KiB 210
megabyte MB 106 220 mebibyte MiB 220
gigabyte GB 109 230 gibibyte GiB 230
terabyte TB 1012 240 tebibyte TiB 240
petabyte PB 1015 250 pebibyte PiB 250
exabyte EB 1018 260 exbibyte EiB 260
zettabyte ZB 1021 270 zebibyte ZiB 270
yottabyte YB 1024 280 yobibyte YiB 280

Woodstone (talk) 17:37, 27 January 2008 (UTC)

I'd prefer it without the "SI" and "JEDEC" column headers, i.e. my last version, because "Binary use" and "decimal use" is more immediately understood and then if the reader wants more info they can click on the links to find out more. Fnagaton 18:37, 27 January 2008 (UTC)
From the values in the table it's immedately clear which values are binary or decimal, so such column headers do no add any information. Note that the standards are still linked. −Woodstone (talk) 18:47, 27 January 2008 (UTC)
I agree with the Fnagaton that the headers "binary use" and "decimal use" are more helpful than "JEDEC" and "SI". Both versions are good though. Thunderbird2 (talk) 22:40, 27 January 2008 (UTC)
After the edit to the headers to add "Decimal" and "Binary" I am not that enamored with the "Special use" as it has all sorts of wishy-washy point-of-view about the word. How about "Uncommon use" instead? Since we all know they are not that commonly used. Or just change it to the "IEC 60027-2" I put before? Either one would be OK with me, but "special" isn't. Fnagaton 00:47, 28 January 2008 (UTC)
Like this example below:
Quantities of bytes
Common use IEC 60027-2
Name Symbol Decimal
SI
Binary
JEDEC
Name Symbol Binary
IEC
kilobyte KB/kB 103 210 kibibyte KiB 210
megabyte MB 106 220 mebibyte MiB 220
gigabyte GB 109 230 gibibyte GiB 230
terabyte TB 1012 240 tebibyte TiB 240
petabyte PB 1015 250 pebibyte PiB 250
exabyte EB 1018 260 exbibyte EiB 260
zettabyte ZB 1021 270 zebibyte ZiB 270
yottabyte YB 1024 280 yobibyte YiB 280
Hmm, looking at the example the table actually looks better than I expected it would. Fnagaton 00:50, 28 January 2008 (UTC)
The titles should be balanced, not one a word, the other a standard. Whay not use: "common prefix" and "binary prefix"? (changed in my version above). Common can both be understood in the sense of "used for both" and also of "often used". −Woodstone (talk) 11:05, 28 January 2008 (UTC)
OK. Fnagaton 17:52, 28 January 2008 (UTC)
IT prefixes
Common SI Industry IEC
Name Symbol Value Value Name Symbol Value
kilobyte kB, KB 10001 10241 kibibyte KiB 10241
megabyte MB 10002 10242 mebibyte MiB 10242
gigabyte GB 10003 10243 gibibyte GiB 10243
terabyte TB 10004 10244 tebibyte TiB 10244
petabyte PB 10005 10245 pebibyte PiB 10245
exabyte EB 10006 10246 exbibyte EiB 10246
zettabyte ZB 10007 10247 zebibyte ZiB 10247
yottabyte YB 10008 10248 yobibyte YiB 10248

<noinclude>

  • First optional parameter gives the unit name, defaults to “byte”.
  • Second optional parameter gives the unit symbol, defaults to first parameter or “B”.

</noinclude> The symbol ‘B’ is actually incompatible with the SI– it already is used for the bel –, ‘byte’ is suggested instead, the French ‘o’ for octet would work too. — Christoph Päper 16:30, 17 March 2008 (UTC)

But we're not blindly following SI here ;) and Wikipedia uses terms found in the sources relevant to the topic. In the real world (and defined by the JEDEC who produce the standard relevant to the topic) "B" can be byte. Fnagaton 16:53, 17 March 2008 (UTC)
I thought you would like that observation, because in consequence it gives something like the following tables. — Christoph Päper 12:56, 18 March 2008 (UTC)
Prefixes for byte
Decimal Binary
Common SI Industry IEC
Name Symbol Value Symbol Value Name Symbol Value
kilobyte kbyte 10001 KB 10241 kibibyte KiB 10241
megabyte Mbyte 10002 MB 10242 mebibyte MiB 10242
gigabyte Gbyte 10003 GB 10243 gibibyte GiB 10243
terabyte Tbyte 10004 TB 10244 tebibyte TiB 10244
petabyte Pbyte 10005 PB 10245 pebibyte PiB 10245
exabyte Ebyte 10006 EB 10246 exbibyte EiB 10246
zettabyte Zbyte 10007 ZB 10247 zebibyte ZiB 10247
yottabyte Ybyte 10008 YB 10248 yobibyte YiB 10248
Prefixes for byte
Decimal SI Name Binary Industry IEC
10001 kbyte kilobyte 10241 KB KiB kibibyte
10002 Mbyte megabyte 10242 MB MiB mebibyte
10003 Gbyte gigabyte 10243 GB GiB gibibyte
10004 Tbyte terabyte 10244 TB TiB tebibyte
10005 Pbyte petabyte 10245 PB PiB pebibyte
10006 Ebyte exabyte 10246 EB EiB exbibyte
10007 Zbyte zettabyte 10247 ZB ZiB zebibyte
10008 Ybyte yottabyte 10248 YB YiB yobibyte
IT prefixes for bit and byte
Decimal SI Binary Industry IEC
10001 k kilo 10241 K kilo Ki kibi
10002 M mega 10242 M mega Mi mebi
10003 G giga 10243 G giga Gi gibi
10004 T tera 10244 T tera Ti tebi
10005 P peta 10245 P peta Pi pebi
10006 E exa 10246 E exa Ei exbi
10007 Z zetta 10247 Z zetta Zi zebi
10008 Y yotta 10248 Y yotta Yi yobi
IT prefixes for bit and byte
n Decimal: 1000n Binary: 1024n
SI Industry IEC
1 k kilo K kilo Ki kibi
2 M mega M mega Mi mebi
3 G giga G giga Gi gibi
4 T tera T tera Ti tebi
5 P peta P peta Pi pebi
6 E exa E exa Ei exbi
7 Z zetta Z zetta Zi zebi
8 Y yotta Y yotta Yi yobi

I'm the one who started this "1024 vs 1000" section with the opening argument, and to be honest this is the first time I've checked it out since then. I just wanted to say, after reading all the subsequent discussion, I am extremely pleased with the thought and effort you all put into fixing this. I also wanted to say that I really like the final table format you guys came up with! It's accurate, concise, and aesthetically compatible.

So yeah, I'm very satisfied with it now. Thank you all, and well done! =)

71.231.87.53 (talk) 08:43, 5 April 2008 (UTC)

Thanks for that. It was icily silent after the last change. Now we might as well kill (redirect) the sister template:quantities of bits. −Woodstone (talk) 06:58, 6 April 2008 (UTC)
For the record, this is the version (made by Woodstone) that was agreed in talk.
Quantities of bytes
Common prefix Binary prefix
Name Symbol Decimal
SI
Binary
JEDEC
Name Symbol Binary
IEC
kilobyte KB/kB 103 210 kibibyte KiB 210
megabyte MB 106 220 mebibyte MiB 220
gigabyte GB 109 230 gibibyte GiB 230
terabyte TB 1012 240 tebibyte TiB 240
petabyte PB 1015 250 pebibyte PiB 250
exabyte EB 1018 260 exbibyte EiB 260
zettabyte ZB 1021 270 zebibyte ZiB 270
yottabyte YB 1024 280 yobibyte YiB 280
Which then got shortened to this version, again by Woodstone:
Prefixes for bit and byte
Decimal
Value SI
10001 k kilo
10002 M mega
10003 G giga
10004 T tera
10005 P peta
10006 E exa
10007 Z zetta
10008 Y yotta
Binary
Value JEDEC IEC
10241 K kilo Ki kibi
10242 M mega Mi mebi
10243 G giga Gi gibi
10244 T tera Ti tebi
10245 P peta Pi pebi
10246 E exa Ei exbi
10247 Z zetta Zi zebi
10248 Y yotta Yi yobi
Fnagaton 17:07, 16 April 2008 (UTC)

The kilo- prefix[edit]

The current version of the template indicates "KB" as a valid form of kilobyte. This is not very accurate, because KB is only accepted by JEDEC. The organization which regulates the SI only accepts "k" as a prefix for "kilo-" (standard prefixes). The SI is a regulation mandatory by the law in lots of countries and used worldwide while JEDEC is only a the standardization part of a trade group. I think the text "KB/kB" should be shortened to just "kB" to be terse while still correct. --RoberPL (talk) 17:55, 25 March 2008 (UTC)

The practice of using K to mean 210 is even frowned upon by JEDEC’s own standard with the words
  • The definitions of kilo, giga, and mega based on powers of two are included only to reflect common usage. IEEE/ASTM SI 10-1997 states "This practice frequently leads to confusion and is deprecated." Further confusion results from the popular use of a "megabyte" consisting of 1 024 000 bytes to define the capacity of the familiar "1.44-MB" diskette. An alternative system is found in Amendment 2 to IEC 60027-2: Letter symbols to be used in electrical technology – Part 2
Thunderbird2 (talk) 18:51, 25 March 2008 (UTC)
It isn't "frowned upon" by the JEDEC. It clearly states in the standard "kilo (K) [note uppercase K] (as a prefix to units of semiconductor storage capacity): A multiplier equal to 1024 (210)." Quoting the notes that are after the standard definition without including the standard definition is taking them out of context. Fnagaton 19:32, 25 March 2008 (UTC)
I wasn't trying to misrepresent what is said there. Here's the complete definition, for anyone interested.
  • mega (M) (as a prefix to units of semiconductor storage capacity): A multiplier equal to 1 048 576 (220 or K2 , where K = 1024).
  • NOTE 1 Contrast with the SI prefix mega (M) equal to 106 , as in a 1-Mb/s data transfer rate, which is equal to 1 000 000 bits per second.
  • NOTE 2 The definitions of kilo, giga, and mega based on powers of two are included only to reflect common usage. IEEE/ASTM SI 10-1997 states “This practice frequently leads to confusion and is deprecated.” Further confusion results from the popular use of a “megabyte” consisting of 1 024 000 bytes to define the capacity of the familiar “1.44-MB” diskette.
I was quoting from NOTE 2. Thunderbird2 (talk) 08:52, 26 March 2008 (UTC)
Seems like more of a glossary definition than an endorsement. They're not saying "do this if you want to be JEDEC-compliant"; they're just saying "this is what some people do". — Omegatron 23:29, 26 March 2008 (UTC)
From page 2 of the standards document: "No claims to be in conformance with this standard may be made unless all requirements stated in the standard are met." So logically if a memory company wanted to state they are in conformance with the JEDEC standard then they would need to use KB/MB/GB in the binary sense. Fnagaton 23:38, 26 March 2008 (UTC)
But I'm looking at that document right now, and they include both the "kilo-" style written out and the "kibi-" style in a table, with a note that "kilo-" is "included only to reflect common usage". So which of these is to be followed in order to be considered in conformance? Has anyone written them to ask for a clarification?
Are there any other standards or "official" documents that state that the "KB = 1024" style is preferred? I thought I read that there was an old (possible superseded) IEEE standard that did. — Omegatron 00:08, 27 March 2008 (UTC)
The notes attached to the standard JEDEC definitions for kilo/mega/giga are there for comparison (for example "Contrast with the...") and are not expected to be seen for conformance with the JEDEC standard. You'll note the terms kibi/mebi/gibi do not have their own entries and only exist in the notes. IEEE 100 defined kilo = 1024. User:Swtpc6800/Binary Fnagaton 00:24, 27 March 2008 (UTC)
kibi/mebi/gibi do have their own entries in JEP120-A, though, which also has the same "conformance" clause and a link to the other document. This doesn't seem to be as clearly defined as you make it out to be.
Does IEEE 1541-2002 supersede IEEE 100-2000 or do they conflict? Maybe our headings should be "IEEE 100" and "IEEE 1541" instead of "JEDEC" and "IEC"? — Omegatron 01:32, 28 March 2008 (UTC)
JEP120-A does not have its own entries for kibi etc. The only mention for kibi is to look at the note for kilo. The only mention of mebi is to look at the note for mega. This is exactly the same as JESD100B.01 which is what I linked earlier. Also JEP120-A published May 2000, the standard document I linked (JESD100B.01) published December 2002. So yes it is as clearly defined as I said it was above. Fnagaton 12:03, 28 March 2008 (UTC)

JEDEC[edit]

I've written to JEDEC, and they confirm that the list of definitions in JESD100 is not a requirement. In fact, they specifically say that "definitions shall not include requirements", and "no one is obligated to use our definitions even if they use the terms". They list these terms in this document only to present the facts of how they are commonly used, and the fact that they are deprecated. So I think it is quite misleading to use "JEDEC" as a heading for this section.

"IEEE 100" might be an appropriate heading, but I haven't read it and don't know much about it myself. — Omegatron (talk) 23:44, 9 April 2008 (UTC)

For the record I wrote to the JEDEC as well and your statement is incorrect. The definitions they show in their standard are not deprecated as you claim. Also you misrepresent the facts because to claim compliance with the JEDEC standards KB/MB/GB have to be used in powers of two sizes. IEC prefixes are not defined as part of the standard either, they are mentioned as a footnote. The fact is the JEDEC standard says "No claims to be in conformance with this standard may be made unless all requirements stated in the standard are met.". That disproves what you just claimed. Fnagaton 13:55, 10 April 2008 (UTC)


I wrote to JEDEC and I got a response from both the chairman and a senior member of JC-10. I'm sorry, but I'm going to trust their interpretation of their own document more than yours.

Their position is quite clear. A summary of the response I got:

  • The "No claims to be in conformance..." clause is a "general notice" that appears in every JEDEC document. The JESD100 document simply contains definitions of terms used in other documents.
  • It is "obvious" that the dictionary definitions are not requirements. "One of our precepts for definitions is that definitions shall not contain requirements."
  • "no one is obligated to use our definitions even if they use the terms"
  • To say that a product was "in compliance with JESD100" would be "meaningless".
  • "The standard meanings of the terms kilo, mega, giga, etc are defined in the SI system of measurement by international agreement to be powers of 10."
  • These terms were "borrowed" for computer memory usage, 'where 1024 was called kilo with an error of "only" 2.4%'. (Note past tense)
  • This document just presents the fact that they are commonly used in a different manner from the standard, and that this usage is officially deprecated.
  • A manufacturer might choose to obligate themselves to use these definitions, if they say "as described in JESD100", but this would be purely voluntary, and they could say this while using either the KB = 1024 definitions or the KiB = 1024 definitions (but not the M = 1,024,000 definition, for instance). "It could apply to either the deprecated powers-of-two system or the kibi-tebi system", since, from their perspective, both are defined in this document. It is not "just" a footnote.

So no, JEDEC is a poor heading for this column, since they define both conventions in the same document, which is not a requirement of a standard anyway. Please stop misrepresenting this. — Omegatron (talk) 00:12, 11 April 2008 (UTC)

Even without the correspondence received by Omegatron, the JEDEC document itself includes the phrase "The definitions of kilo, giga, and mega based on powers of two are included only to reflect common usage". That sounds to me like clear attempt to acknowledge that use without actually endorsing it. If Fnagaton has evidence supporting his view I suggest he produces it. If he can't do this I support Omegatron's position. Thunderbird2 (talk) 06:00, 11 April 2008 (UTC)
The KB/MB/GB prefixes are in a document that has the text "Terms, Definitions, and Letter Symbols for Microcomputers". Therefore it is correct to say the terms are defined by the JEDEC. To try to claim those prefixes are "not endorsed" is point of view and is contrary to the facts. Omegatron is not a reliable source. Since Omegatron's claims are not from a reliable source then this means what he writes cannot be verified. Therefore trying to use his claims in articles (or templates used in articles) is in violation of Wikipedia:No original research. Use of his claims in articles (or templates that are included in articles) is also in violation of Wikipedia:Neutral point of view. Since his claims do not meet the criteria for inclusion in Wikipedia articles (or templates that are included in articles) then they should be disregarded. A case in point is Omegatron's recent edit warring to include material that is unsourced and entirely his own point of view, the reason he gave for one of the edits "i've talked to JEDEC directly about this and the interpretation here is misleading" violates these Wikipedia policies. Just to make it absolutely clear Omegatron, Wikipedia does not publish original thought: all material in Wikipedia must be attributable to a reliable, published source.. If you want to put in articles that "the JEDEC do not endorse KB/MB/GB" then find reliable sources that say so. Also if you want to put in articles "the JEDEC have deprecated KB/MB/GB" then you must also supply reliable sources that state that. Fnagaton 08:51, 11 April 2008 (UTC)
You've never had a reliable source in the first place. You're just repeating your personal interpretation of the document, which I have shown to be wrong. From JEDEC's point of view, they are defining both the binary prefixes and the IEC prefixes (not just including one as a "footnote"). From their perspective, both are equally valid, and neither are a required part of the standard. You'll have to come up with a reliable source that JEDEC "endorses" one over the other, or you cannot continue putting this notion into articles and templates. — Omegatron (talk) 15:06, 13 April 2008 (UTC)
You are wrong because: 1) The reliable source is the document itself and the words contained within that document. 2) You have not shown anything to be wrong, what you have done is post your own personal opinion and interpretation. What you have done is violate Wikipedia:No original research. 3) You have the whole "endorses" thing the wrong way around because it is clear the JEDEC document defines the terms. What you have to do is provide proof that they do not define and therefore not endorse these terms. Fnagaton 15:19, 13 April 2008 (UTC)
You see Omegatron it isn't your change to remove JEDEC from the table that I disagree with, in actual fact if you look at my comment from earlier on this page "I'd prefer it without the "SI" and "JEDEC" column headers, .... Fnagaton 18:37, 27 January 2008 (UTC)" I actually agree with not having those headers in the table. The fact is my personal opinion and your personal do not mean the table can be changed to what you prefer because as the discussion above shows the consensus was to include JEDEC in the heading. What I specifically disagree with is you coming along and making a change without talking about it first with the other involved editors because we all helped with the consensus for the change. This means what I want you to do is follow procedure and that means talking about these things first, rather than forcing your point of view onto something. Fnagaton 16:18, 13 April 2008 (UTC)


You're completely right that the emails I got are not reliable sources. I never said they were. But this is irrelevant. I'm not adding things to articles and citing those emails as my sources. I'm removing claims from articles that don't have reliable sources, and which I now know to be untrue based on first-hand knowledge. "The burden of evidence lies with the editor who adds or restores material. .. Any material lacking a reliable source may be removed."

You've never had any reliable source that says "JEDEC endorses a particular unit convention as part of their standards"; you're just interpreting their document that way.

Any interpretation of primary source material requires a reliable secondary source for that interpretation. To the extent that part of an article relies on a primary source, it should make no analytic, synthetic, interpretive, explanatory, or evaluative claims about the information found in the primary source.

This applies to the Binary prefix article and any other articles you've been inserting this into, as well. — Omegatron (talk) 18:14, 13 April 2008 (UTC)

You're wrong, it is relevant. Did you, or did you not, try to cite your alledged "talk" with the JEDEC [2] in order to make the edit to try to imply the JEDEC said the units are deprecated? You did, don't try to deny it, the diff comment speaks for itself. Your edit is facutally incorrect because it isn't the JEDEC saying the units are deprecated, the actual quote comes from "IEEE/ASTM SI 10-1997" instead. You are trying to remove material where reliable sources have been cited and the reasons you have been giving are your own personal opinions. Your "first hand knowledge" is still not good enough reason to remove that material. This applies to your edits on other topics. By the way your point about the word "endorses" is a straw man logical fallacy since you are the one who used that word first ( 23:29, 26 March 2008 (UTC) ) on this page. Don't try to change the point by using the word "endorse" because my position has been to use the word "definition". The word "definition" is precise since it comes from the definition documents and not open to interpretation. Your other edit removes yet mor efactual content again using your personal opinion. You are wrong to do so because:
  • It is a fact the JEDEC publish computer memory standards.
  • It is a fact the memory standard document JESD100B.01 is titled "JEDEC STANDARD".
  • It is a fact the same document says "Terms, Definitions, and Letter Symbols for Microcomputers, Microprocessors, and Memory Integrated Circuits".
  • It is a fact the same document includes sections for each of the terms they define.
  • It is a fact the same document has sections for each term they are defining, with the term being defined in bold text.
  • It is a fact the terms kilo, mega, giga are listed in their own sections.
  • It is a fact the terms kibi, mebi and gibi do not have their own sections and are not highlighted in bold text, rather they are included as footnotes.
  • It is therefore correct to say the JEDEC define kilo, mega, giga.
If the terms kibi, mebi and gibi did have their own sections then your argument would have some merit, but the fact is they do not. Fnagaton 18:24, 13 April 2008 (UTC)


Actually Omegatron, here is one for you. If you are still going to contend the JEDEC do not define these terms in their standards definition document then you have to prove the SI really does define kilobyte and the IEC really does define kibibyte. Provide the exact cites from reliable sources that state this. If you cannot provide the exact cites that means you must remove the terms from the pages on SI and IEC. Fnagaton 19:19, 13 April 2008 (UTC)


I note Omegatron's recent attempt to again push his point of view and unsubstantiated original research into the template instead of coming to some sort of agreement in talk first. Fnagaton 16:20, 18 April 2008 (UTC)
  • Omegatron. What is this with your facination with IEC and JEDEC? Your attempts to simply have it “Omegatron’s way” have the effect of promoting units of measure that simply didn’t catch on with the rest of the English-speaking world. Our job is to communicate to a given readership with minimal confusion so they can learn about a subject and are primed as well as possible to learn more about the subject in their readings elsewhere. What you are doing is attempting to hijack Wikipedia and use it as your personal forum to promote change (in violation of WP:SOAP WP:NEO, and WP:V) because the IEC prefixes are a good idea. That is not the purpose of any encyclopedia. It’s time to give it up and allow other editors to make Wikipedia ‘go with the flow’ of the rest of the English-speaking world regarding binary prefixes. Greg L (talk) 17:39, 18 April 2008 (UTC)
It's about time you stop harassing Omegatron and your smear campaign against the IEC. --217.87.60.234 (talk) 18:00, 18 April 2008 (UTC)
  • 217.87, if you are suggesting that I’ve been “stalking” Omegatron, or have been hounding him beyond reason for what could be considered as acceptable debate here on Wikipedia, you are mistaken. If you check my contributions, you’ll find it’s been quite some time that I ever stumbled across his antics and bothered to offer my 2¢. If you allege otherwise, bring on your evidence. Or are you just implying that I have no right to weigh in on this issue? If that, you just wasted 1648 kibles ‘n bits of Wikipedia disk storage space with your above post. As for “a smear campaign against the IEC”, that’s one of the most nonsensical things I’ve ever seen written down this week. Are you trying to advance the cause of the IEC prefixes or are you really on my side of this issue, slyly providing me with poor arguments on purpose?!? Besides, Omegatron is a big boy and is an administrator at that; I think he can defend his behavior on his own. Greg L (talk) 18:43, 18 April 2008 (UTC)
I don't know why you mention "stalking" but there's probably some good reason. Anyway, I'm certainly not the only here, who considers your behaviour towards Omegatron as unacceptable. I could care less whether he's an admin. I didn't even know that before you pointed it out some time ago. My opinion is simply that whenever someone is treated that way we must not keep watching but speak up. This has nothing to do with siding with any position. I don't know what a kible is but I guess you just can't refrain from trying to ridicule others when you're running out of valid arguments. --217.87.60.234 (talk) 19:15, 18 April 2008 (UTC)
  • Fine, your stated position for defending Omegatron is a valid position and I respect it. That doesn’t mean I agree that he 1) needs it, and 2) doesn’t deserve to be challenged for his position on this issue. As my “1648 kibles ‘n bits rub on a silly unit of measure that all agree the typical Wikipedia reader is unfamiliar with (and with regard to “valid arguments”), I’ve got plenty of both. Screwing up templates to give the appearance of “Oh… this is the way it’s really done” when the rest of the world doesn’t do it that way, and justifying it because it’s a proposal from some an important commission, doesn’t mean squat when it comes to using them in an encyclopedia. It still amounts to WP:SOAP and has the effect of making Wikipedia the only damn place that routinely uses them in the real world. That’s pure B.S. and these proponents of using them have to get real. Greg L (talk) 19:31, 18 April 2008 (UTC)
  • Greg talks a lot of sense 217.87.x.x you'd be better off listening to him. Greg (and others) challenging a position with reasonable argument is not harassment, instead what you are doing 217.87.x.x is more like harassment because you are throwing around baseless accusations. If Omegatron doesn't want his position challenged then he's chosen the wrong place to write it. He has the option, like all of us, to take a break whenever he likes or to go and try to edit something else if he doesn't want the stress of arguing a minority point of view. Fnagaton 22:46, 18 April 2008 (UTC)
I didn't make any baseless accusations whatsoever. My base is this. Omegatron has already stated reasons for removing the word "JEDEC" from the table. By the way, I talk a lot of sense, too, but you probably won't listen to me anyway, will you? --217.87.60.234 (talk) 00:31, 19 April 2008 (UTC)
Posting a random link to someone's talk history when there isn't anything in that history to support your accusation is a good example of a baseless accusation. Omegatron's "reasons" are original research, are not reliable sources and also his claims are the exact opposite to what I know on the subject. Also his "reasons" are also contrary to the facts as presented in the JEDEC standards documents and supporting reliable sources on the subject. This has already been shown up above and on other talk pages on this subject. Fnagaton 09:31, 19 April 2008 (UTC)
Fnagaton, let me ask you a question. "The moon is not made of cheese." Do you agree with this statement? --217.87.60.234 (talk) 13:11, 19 April 2008 (UTC)
I agree this is not “Omegatron’s way” or his soapbox. To remove the JEDEC text is distorting the facts.DavidPaulHamilton (talk) 23:34, 18 April 2008 (UTC)
No, it is not. The table existed before someone edited the "JEDEC" to it. Further, JEDEC has neither invented the convention in question nor are they an authority. They even say so themselves. --217.87.60.234 (talk) 00:25, 19 April 2008 (UTC)
It was a group agreement to have JEDEC in the table. Then Omegatron tries to make changes on his own. DavidPaulHamilton (talk) 07:27, 19 April 2008 (UTC)
It's misleading to use JEDEC as a header alongside IEC (or SI) as if they were somehow comparable. JEDEC are not the originator of the binary use, and their definition falls little short of apologising for it. There must be something better. Any sugestions? Thunderbird2 (talk) 20:33, 23 April 2008 (UTC)
How about "old" and "new" instead of "JEDEC" and "IEC"? --217.87.102.163 (talk) 20:50, 23 April 2008 (UTC)
Using JEDEC is not misleading. It is misleading to mention SI since they are not the originator of the metric system. Also it is misleading to mention IEC because their prefixes are hardly used. the JEDEC are the standards organisation for this prefix use and it is only your point of view about their definition.DavidPaulHamilton (talk) 11:49, 24 April 2008 (UTC)

I don't see how it can be misleading to refer to SI prefixes as "SI prefixes", so I will ignore that comment. Regarding JEDEC, a phrase such as "included only to reflect common usage" clearly indicates something less than strong support. How do you suggest I read it otherwise? If the BIPM were to use similar wording in their documentation of the SI (and sometimes they do) that would indicate to me that a unit or symbol was either on its way out or not yet fully accepted. Why is this situation any different?

It seems to me that JEDEC has gained the right to head this column by default, just because they are the only organisation left that falls short of disassociating itself completely from this ambiguous and confusing practice.Thunderbird2 (talk) 14:29, 24 April 2008 (UTC)

You wrote "a phrase such as "included only to reflect common usage" clearly indicates something less than strong support" - That is your personal opinion and is not supported by the facts because the JEDEC publish their documents using KB, MB, GB all in powers of two sizes when they are talking about semi-condictor storage capacity. The JEDEC are listed on the column heading because the fact is they define these terms in powers of two sizes and they are the internationally accepted standards organisation dealing with this topic. Fnagaton 15:39, 24 April 2008 (UTC)
I suggest we change "JEDEC" to "Fnagaton". Otherwise, someone should contact JEDEC and ask them whether they want to associated with this. JEDEC didn't invent the convention, JEDEC already mentions and explains the new prefixes from IEC 60027-2. I don't see how it's valid to give the impression there was a dispute between JEDEC and IEC. Putting JEDEC into the header is completely arbitrary. Even "Microsoft" would be more appropriate because they are one of the few big players who completely refuse these new prefixes and they are also largely responsible for "infecting" normal users with this convention. --217.87.83.146 (talk) 17:16, 24 April 2008 (UTC)
Just because you do not have a user account it does not give you freedom to be uncivil and try to troll other editors.DavidPaulHamilton (talk) 20:00, 24 April 2008 (UTC)
A sock like you should be more careful with his statements. --217.87.83.146 (talk) 21:39, 24 April 2008 (UTC)
It's worth noting at this point that user 217.88... has been range blocked for a week for repeated vandalism. Fnagaton 02:13, 25 April 2008 (UTC)

replacement header for JEDEC[edit]

I've just realised that JEDEC define kilo-, mega- and giga- but no further. The table gives the misleading impression that binary definitions of tera-, exa- and so on are also sanctioned (or at least condoned) by JEDEC, when they are not. My suggestion? Use FOLDOC instead. Thunderbird2 (talk) 23:24, 5 May 2008 (UTC)

Have a look at my edit above for the proposed table at "13:42, 27 January 2008 (UTC)" and then your following comment at "14:04, 27 January 2008 (UTC)". ;) Fnagaton 23:58, 5 May 2008 (UTC)
I'm happy with the changes so far.DavidPaulHamilton (talk) 14:53, 6 May 2008 (UTC)

Is FOLDOC really an "authority", though? What's wrong with IEEE 100? — Omegatron (talk) 23:01, 11 May 2008 (UTC)

My impression is that IEEE 100 is superseded by later IEEE standards following IEC and NIST. FOLDOC is the only reputable body I know that continues to endorse binary use of SI prefixes. Thunderbird2 (talk) 23:12, 11 May 2008 (UTC)

I had the same impression, but thought it would be a more "official" definition. But I looked, and IEEE 1000 actually defines it both ways, as well:

megabyte Either 1 000 000 bytes or 220 bytes. Notes: 1. The user of these terms shall specify the applicable usage. If the usage is 210 or 1024 bytes, or multiples thereof, then note 2 below shall also be included with the definition. 2. As used in IEEE Std 610.10-1994, the terms kilobyte (kB) means 210 or 1024 bytes, megabyte (MB) means 1024 kilobytes, and gigabyte (GB) means 1024 megabytes.

Onto IEEE 610.10.  :) — Omegatron (talk) 23:16, 11 May 2008 (UTC)

IEEE 610.10 also lists both the decimal and binary definitions. — Omegatron (talk) 23:20, 11 May 2008 (UTC)

I'm going to cast my vote against FOLDOC. It's Denis' personal pet project. He cites no references to support his definitions and employs no review process, which he is certainly free to do. FOLDOC is essentially Denis' personal dictionary and since he appears to be a pretty knowledgable person on the subject, any single person would probably be well served by using his dictionary as their source. But, FOLDOC is no more "official" than his personal opinion. For the mega vs mebi battle, the template needs a FAR better source. --JJLatWiki (talk) 22:59, 15 May 2008 (UTC)
The point is that none of us can think of a better source than FOLDOC. JEDEC was used until we realised that they do not define prefixes larger than giga-. What alternative source do you suggest for (binary) definitions of tera-, peta- ... yotta-? Thunderbird2 (talk) 09:07, 16 May 2008 (UTC)
I understand the desire. In my opinion, if a standards body has not defined it, then the source could be general purpose dictionary of the type that has various committees with appropriate experts and philologists and an editorial review process. Unfortunately, the mainstream dictionaries are still reflecting the prevailing common usage of megabyte and ignoring the international and national standards bodies. I say the fields that have no authoritative supporting source should be left blank. It's unfortunate, but since the infobox information should be the most authoritative, its sources should be the most defensible. --JJLatWiki (talk) 19:40, 16 May 2008 (UTC)
Golly - hadn't thought of leaving any of it blank. Are you you suggesting that we return to JEDEC and define only the first 3 prefixes, or do you mean the header should be left blank? Thunderbird2 (talk) 20:37, 16 May 2008 (UTC)
I'm not necessarily getting in the ring for JEDEC. To me, their authority applies more to solid state devices and, more specifically, their members than to the broader electronics industry. But, if they're the only standards or standardizations body with the desired definitions, then yes, put JEDEC instead of FOLDOC on top and leave the holes in the definition as holes in the table. Maybe an asterisk in the empty boxes that says, "No authoritative definition". Or, include the missing JEDEC definitions but with an asterisk that says, "By extrapolation. Not explicitly defined." --JJLatWiki (talk) 23:23, 16 May 2008 (UTC)



Prefixes for bit and byte
Decimal
Value SI
10001 k kilo-
10002 M mega-
10003 G giga-
10004 T tera-
10005 P peta-
10006 E exa-
10007 Z zetta-
10008 Y yotta-
Binary
Value JEDEC IEC
10241 K kilo- Ki kibi-
10242 M mega- Mi mebi-
10243 G giga- Gi gibi-
10244 Ti tebi-
10245 Pi pebi-
10246 Ei exbi-
10247 Zi zebi-
10248 Yi yobi-


Leaving it with blanks gives this box on the right. It seems sensible to me. What do others think? Thunderbird2 (talk) 09:10, 17 May 2008 (UTC)

Looks like a good idea, but better reverse the last two columns then. −Woodstone (talk) 18:49, 17 May 2008 (UTC)


Prefixes for bit and byte
Decimal
Value SI
10001 k kilo-
10002 M mega-
10003 G giga-
10004 T tera-
10005 P peta-
10006 E exa-
10007 Z zetta-
10008 Y yotta-
Binary
Value IEC JEDEC
10241 Ki kibi- K kilo-
10242 Mi mebi- M mega-
10243 Gi gibi- G giga-
10244 Ti tebi-
10245 Pi pebi-
10246 Ei exbi-
10247 Zi zebi-
10248 Yi yobi-

Oh right. That stretches my formatting skills to the limit ...

How about this? Thunderbird2 (talk) 19:35, 17 May 2008 (UTC)

I support that with JEDEC in either position. I think both formats reflect the positions of the most relevant standards and standardizations bodies. I don't know protocol in templates, but should there be direct links in the template to supporting citations? --JJLatWiki (talk) 02:43, 18 May 2008 (UTC)
I believe this is better than what we had, and no one has expressed opposition to it, so I have updated the template accordingly. I'm not sure how to deal with citations either, so in that respect I'm leaving it how it was. Thanks for your contribution. Thunderbird2 (talk) 20:53, 18 May 2008 (UTC)


The problem with JEDEC is that they define both the K=1024 and the Ki=1024 conventions in the same document, and using it as a heading implies that they endorse that particular one alone. I'm fine with FOLDOC in the header. — Omegatron (talk) 21:24, 25 May 2008 (UTC)

I am comfortable with either JEDEC (with the first 3 prefixes only) or FOLDOC (with the full set). But JJLatWiki makes a good point about the different status of the various organisations. For example, if you put FOLDOC side by side with IEC, the implication is that they have comparable status. How do we avoid giving that impression? Thunderbird2 (talk) 22:03, 27 May 2008 (UTC)
I don't know. I'd just give it a neutral header like "binary prefix" like it used to have, and link to the article that explains that convention. — Omegatron (talk) 00:35, 28 May 2008 (UTC)
I'm not opposed to removing the JEDEC column entirely per se, but not in favor of FOLDOC. I'm afraid, however, that if a "K=1024" column is not shown, the "K=1024" cabal (Note tongue-in-cheek) will protest. If there is any notable standards or standardizations organization out there who either defines or endorses "1024=K", without the ambiguity of also endorsing "1024=Ki", I would favor it. But at present, it seems everyone may be just waiting to see if this "Kibi-thing" catches on before they make a commitment. Maybe JEDEC's duality is their interim move to follow the IEC. Frankly, I don't understand why the ITU or TIA hasn't endorsed the "Kibi" since they're also strong users of the "K=1000" definition. So, is there any way to explain why there is a JEDEC column showing only "K=1024" with just a link to another WP page? Is that "self-referencing"? --JJLatWiki (talk) 16:34, 28 May 2008 (UTC)

For citations, we could include a superscripted link, like IEC[3]Omegatron (talk) 21:27, 25 May 2008 (UTC)

good idea Thunderbird2 (talk) 22:05, 27 May 2008 (UTC)
agreed --JJLatWiki (talk) 16:34, 28 May 2008 (UTC)

brontobytes[edit]

i've provided multiple sources for brontobytes. why are some eidtors fighting to keep this out? one said it is "non standard", but this page isn't the "IEEE standard quantities of bytes" page. do you guys have a better name for the quantity larger than yotta? if not, i'm going to put in bronto. 71.112.130.129 (talk) 06:39, 9 March 2009 (UTC)

The brontobyte is a fake unit, used informally by some to signify "a lot of bytes", similar to the use of zillion. It does not belong in a table specifying precise and well defined quantities. For more information about the level of vaguary, see also the article bronto. −Woodstone (talk) 10:01, 9 March 2009 (UTC)

moved from template[edit]

{{Tfd|Quantities of bytes|Data unit templates}}

Rename?[edit]

The contents of this template have been arrived at through consensus, but the name appears to be a historical accident. Recently User:Cybercobra has sought to change the template to fit the name and the analogy with Template:Quantities of bits (and perhaps even to refashion it as a navigation template), but that is not what is intended by the "users" of the template, most of whom use it as an informative table in the article, using a different template for navigation. So, rather than changing the template to fit the name, we should probably change the name to fit the template. How about "Template:Binary prefixes", once that one is deleted? Shreevatsa (talk) 01:02, 30 September 2009 (UTC)

I hadn't seen your comment (only your edit summary), but I went ahead and moved the template. --Cybercobra (talk) 01:18, 30 September 2009 (UTC)

Again "tera" in JEDEC[edit]

The recent template contains a "tera" in the JEDEC-section, which was added a few days ago. As discussed above, 100B.01 does not seem to define "tera" in a binary manner. Since no source was provided for the amendment and the user did not participate the above discussion, this is possibly a misapprehension. The preveious version seems to apply to a consensus, and therefore, I dare to preliminarily delete that term until further evidence could be given, ok? --GlaMax (talk) 10:38, 1 October 2009 (UTC)

Decimal and Binary[edit]

This is very misleading seeing decimal and binary in the column headers and then base 1000 and base 1024 below. Both of these are derived from their base units: 103 and 210--base ten (decimal) and base two (binary), respectively. I would highly recommend reverting the table to its original state in this regard because basing everything on kilobyte and kibibyte instead of byte (the unifying unit) makes no sense. FordGT90Concept (talk) 01:51, 28 April 2011 (UTC)

It makes sense in that the exponent progression is simple and clear. Calling the units binary and decimal like this is pretty normal, I don't find it too confusing. But I personally don't care either way. --Cybercobra (talk) 03:55, 28 April 2011 (UTC)