Become a fan of Slashdot on Facebook


Forgot your password?

New Power-of-Two Prefixes? 289

EngrBohn writes "The August issue of IEEE Spectrum mentions a proposal by the International Electrotechnical Commission to introduce new prefixes for words that indicate powers-of-two (page 18 of the print issue). This would replace kilobytes (kB) with kibibytes (KiB), megabytes (MB) with mebibytes (MiB), gigabytes (gB) with gibibytes (GiB), and so on. The rationale is two-fold. First is to restore the integrity of the SI prefixes to meaning powers-of-ten. Second is to eliminate ambiguity over whether, for example, a megabyte is 10**6 bytes or 2**20 bytes. Think this is a non-issue? I noticed this morning that Iomega's 100MB Zip disks have a 10**8 byte capacity, and Maxtor also considers a megabyte to be 10**6 bytes. "
This discussion has been archived. No new comments can be posted.

New Power-of-Two Prefixes?

Comments Filter:
  • The real thing to do is just to use If this displays properly in your browser, congratulations. :)
  • Agh. I mean
  • I do networking research, and the 2^x/10^z question comes up all the time--this is because bandwidth is usually expressed in powers of 10, and file size is usually expressed in powers of 2. This would be fine if people said "100 10^6 b Ethernet", but they don't, they say "100 Mb Ethernet".

    As a result, we get a lot of headaches. A good paper or talk distinguishes between 2s and 10s, or better yet normalizes everything to the same base. A bad presentation doesn't distinguish anything, leaving you to figure out which units they're using where.

    The proposed names are unpleasant, but we could just consider them placeholders until we come up with something better. We definitely need some sort of word here.
  • by jd ( 1658 )
    Hey! A Zorkybyte is a well-defined unit!

    The units are:

    1024 Flatheadbytes to 1 Zorkybyte;

    1024 Zorkybytes to 1 Frobozzbyte;

    1024 Frobozzbyte to 1 Infocombyte

    These are understood by every grue under the sun!

  • So the new names will only really make sense to anyone whose taken courses in low level computer stuff.

    Or anyone whose taken a bike ride.
  • Kikibyte sounds like some Polynesian parrot.

    To a Filipino speaker (Tagalog, anyway), this term would describe a unique form of oral sex.

  • Does this mean that Y2K really is the year 2000? I've been telling everyone it's coming in 2048.

    Man.. I've got some serious catching up to do now.

  • Hard drive manufaturers have been using the right notion. They actually only provide you with a drive that if marked as 6 GB has only about 6 billion (US billion, mind you) bytes. This new standard would not change that. They would still only sell a 6GB drive and somewhere on the box it might have a translation into Gibibytes for those who care. The point of this notation change is for people who work with this stuff everyday on the job. If I have to refer to memory segements on byte boundries but refer to them in metric, base ten units I am refering to them incorrectly. This notation finally allows geeks to be precise about nitpicky, but important things. People who deal with this sort of thing everyday will be able to use this but the general populous won't care and probably will be confused when one of us correct them.

    Long Live the Gibibyte!
  • It's far more nefarious than that!

    The MiB wanted to make sure people didn't freak out if they see MiB on confidential documents, and they decided the Kids in Black and the Girls in Black needed the same protection!
  • I've a better idea. I propose that HD manufactuerers give away 200 Gb RAID arrays to everyone on Slashdot, unless they agree to conform to the computer definitions.

    Refusing to hand over the hard drives will be considered compliance with the terms of the agreement. If they -do- hand over the drives, nobody on slashdot is likely to care what convention they use.

  • I want my Kibis and Bits!
  • Yes, telecom bandwidth is always given in base 10. They consider it a straight pipe and measure in bits where 1 kb means 1,000 bits/sec. Throughput or information rate refers to the usable bandwidth, i.e. minus headers or other overhead. Compression adds further confustion to the issue. Telecom does not factor in the compression. If you have a 2:1 compression rate over a 56k line, then the telcom refers to it as a 56k line while the user will see a higher transfer rate.
  • This is a much bigger problem than the petty dispute about cracker/hacker terminology. (flames>/dev/null)

    The problem is that science requires specific meanings for measurement units. Mega means 10**6, not 2**20. No one may use the new units in popular press, but it'd be great for publication. There would be no doubt that 25 Kibibytes is 25 * 2**10 bytes.
    It's a shame that they didn't do this 20 years ago.

  • Yes, but this is in a country that
    can't decide wheter pounds means a
    unit of money to Force.

    Although, I suppose they don't use
    pounds as a unit of force anymore,
    because they are all metric'ed out.
  • Now go have your coffee . . . :-)

  • I've used 'marketing gigabytes' and 'real gigabytes' for a long time.

    I'm not sure which of the big disk drive manufacturers deliberately created this confusion, but they did this back somewhere around the 40MB drive time frame. Suddenly, simply by redefining their terms a little, they could have '42 megabyte' hard drives instead of '40 megabyte'. Anyone who actually knew anything despised the practice, but the manufacturers that did NOT accept the term quickly found themselves at a competitive disadvantage.

    I think it took less than six months for 'marketing gigabytes' to take over for the correct usage of the term. It has been causing problems ever since.

    Castlewood is the only new manufacturer I know of that seems to be using 'real gigabytes'. I have one of their ORB drives and you really do get 2.2GB (at least as far as Windows is concerned).

    These new words just aren't going to work. You must really understand them to use them properly, and how many people do YOU know in your daily life that understand base-10 versus base-2 notation?

    Anyone will understand 'marketing gigabytes' versus 'real gigabytes'. This usage makes the original lie obvious, and will help to correct the problem by gently reclaiming the correct word, instead of forcing a new one down people's throats.
  • The second means nothing to a computer, the byte does.
  • The second means nothing to a computer, the byte does mean something..
  • I definately like the idea - especially after the arguement I had with my husband on whether to partition 2000 MB for a 2GB partition, or 2048MB. But 'kibibytes'? 'mebibytes'? 'gibibytes'? Could they have chosen *anything* that would have been more confusing? I suppose I can understand the desire to keep things similar, but for the first few years, lay people are going to think you an idiot - or at least in possession of an odd stutter - if you use these terms to them. It would have been far better to simply use something that had a totally different sound/first letter. Like... I don't know. Snazzybyte. Fluptybyte. Zurzabyte. *Anything*.
  • This has been punted around the industry for a few years now. Read some back issues of the IEEE mags, especially the Technically Speaking column.

    This is a great idea, because it separates the two systems of ^10 and ^2. The only ones who will suffer in the long run are the marketing assholes who like to cheat in their specifications.

    Without a doubt, even if this system is adopted (and it will be, the debate has gone on for years, and is now tilting towards acceptance), it will be another decade or two until it reaches widespread use. But for a while, it will hilight the differences between leading edge geeks who like change, and unimaginative nerds who like things to stay the same (640 Kbytes is enough memory for anyone for ever).

    The only thing I would also like to see is some larger and smaller values, into the ranges of 2^-100 and 2^100 or even further. How much space will there be if the other story on 3D holographic storage turns out to be the next great thing? I would love to have a credit card sized 2^100 bytes of information, could keep all the world's pr0n and MP3s on it :-)

    the AC
  • I think for the binary version (preserve the Metric System for powers of 10!), we could use pig latin.
    2^10 bytes = 1 Ilokay Bytes (1 iB) = 1024 B = 1.024 kB
    2^20 bytes = 1 Egamay Bytes (1 eB)
    2^30 bytes = 1 Igagay Bytes (1 IB)
    2^40 bytes = 1 Erratay Bytes (1 EB)
  • I for one am fully receptive to this change! To bad it doesn't clear up the confusion where 1GB = 1000MB = 1000000kB but 1GiB = 1024MiB = 1048576kB i.e. the problem that you still can't divide/multiply by 1000 to go from unit to unit.
  • I'm just getting used to MegaHertz instead of
  • Why are there 1024MB in a GB of RAM, but only 1000MB in a GB of Hard Disk space, but there are 1024KB in one MB for either?

    Uh, there aren't. The HD manufacturers correctly use the definition 1GB = 10^9 bytes. Only "1.44MB" and "2.88MB" floppy disks uses the stupid 1MB = 1024000 bytes definition.

  • Actually, there are good reasons why a base 60 system makes sense. For one thing, b60 can be divided in more ways than b10, and even in more ways than 10^2.

    integer factors of 10: 1, 2, 4, 5, 10 integer factors of 100: 1, 2, 4, 5, 10, 15, 20, 25, 50 integer factor of 60: 1, 2, 3, 4, 5, 6, 10, 12, 15, 20, 30

    This makes it perfect for time. Saying I'll meet you in 5, 10, 15, 20, 30, 40, 45 minutes are all natural in b60, but 20 and 40 wouldn't be in b10. We tend to think b10 is the natural way to do things, but the only thing magical about it is that we have 10 fingers and toes. The sumerians were pretty brilliant, after all. Its the b10 system that doen't make sense. We can't even easily express 1/3 in it!

    Besides, working in different base systems promotes flexibility of mind, grasshopper.
  • I have a better idea.

    Instead of kilobytes, call them kibobytes. Think James Parry will object?

    Gawwwwwd... Looks like these people have too much free time on their hands. Why don't they just spend their free time trying to invent warp drive, or something? Leave this kind of stuff to Jay Leno, or David Letterman.

  • I still don't know if 20 kib/s is kibibits or kibibytes. Ditto for kiB/s, KiB/s, and all other possible capitalizations.
  • Are they going to enforce when I use "that" or "which" too?

    Languages change and evolve on their own, and making changes to technical terms that have already made it into common household usage is ridiculous and implausible. Just another waste of time, money and effort. They probably been working on this plan since we had 32 "kibibytes" of RAM.

    TheGeek []

  • Indeed. It's way to late to change the terminology. BUT, It's not too late for manufacturers to have to come clean about them ripping us off.

    If [insert drive vendor here] were to start slapping a label on their 21474836480 byte disks that said "20 TRUE Gigabytes of capacity!" with a little inforgraphic on how the others are ripping you off, we could be very sure that the rest of the vendors would soon follow suit. This doesn't address the fact that we're talking about powers of 2 not powers of ten but it will at least be a start.

  • when I had that old M6502-based Kim-1 with 1 KB of RAM...
  • In realistic terms, the whole kilo-mega-giga-bytes thing doesn't make any sense except that it looks similar to the base 10 system we all know. How often are things actually close to 1 kilo/meg/gig? Let's see, the 8088 had a 20 bit address so it could access 1 megabyte (2^20). Other than that you usually have 2, 4, 8, 2^4 (16), 32, 64, 128, 2^8 (256), 512, 1024, etc... with emphasis on the more 2's in the picture, eg. 2^2^2^2 = 65536 bytes/colors/gold pieces whatever. The confusion is because of looking at things in base 10 in the first place. You need to start looking in something binary friendly like hexadecimal. 0x10000 different colors! I don't have 128 MB of RAM = 134217728 bytes... I have 0x8000000 bytes of RAM. If I get another 8 sneezybytes I'll have 0x10 sneezybytes! Or maybe I could just add this 4 sneezybyte DIMM here and then I'd have C sneezybytes.
  • me and my friends have been calling them "metric megs" and "binary megs" for a while. although with frequent use, one has became commonly called "metric megs bleegh" :)

    I can't wait till i start seeing the disk drives saying gibibytes and confusing people. We shall finally know who exactly is ripping us off. I know i hate it when i get only 60 megs of memory.
  • We can't get the general public to stop saying "hacker" when they mean "cracker". How can we expect them to get this?
  • I'm all in favor of the changes.

    When you buy a hard drive that is 20GB, how big is it?

    20,000,000,000 bytes? (20*10**9)
    21,474,836,480 bytes? (20*2**30)

    Note that the difference is over 1.4 billion bytes. (Over a Gigabyte, by any definition).

    To me, this is a significant difference, and a flaw in the terminology. As sizes grow, the difference between the standard metric prefix definitions (where mega=1 million and giga=1 billion) will grow exponentially from the actual terms we are using.

    I'd also like to see people become more aware of these things and more conscious of using them. There is a lot of terminology confusion in the market.

    Modems, historically, have been labelled in 'bit' transmission rates, rather than bytes. It is important to know the difference between 10MB/sec and 10Mb/sec; 10MB/sec = 80Mb/sec

    Also, capital M, please. You have 128MB, not 128mb. Small m means 'milli', or one thousandth. Capital M means 'mega'. The metric system is well defined, but these small abuses diminish its worth.
  • I have always understood kilobyte to be 1024 bites, megabyte to be 1024^2 bytes, etc; and when I realized that diskmakers were misusing the words I actively considered trying to sue the diskmakers for false advertizing (still think they should be sued, I just don't have the money to do it)

    Kilobyte, etc should be strictly defined as NOT powers of 10! That would make the problem go away (poof)


    Another interesting thing that some group should do is come up with a system for speaking in hexideciaml. If we could speak in hex, then we could think in hex, and one of the biggest kludges of all time (actively using the decimal system) would go away.

  • Actually, the visual difference is in the spacing between the number and the unit abbreviation. Meaning that
    1MB and
    1 MB
    is not the same thing. This is ALREADY an IEEE standard. However it is kind of a dumb standard because it is not very obvious and most people do not expect the difference to be in the spacing.
  • When people talk about bandwidth, they often refer to Gigabits of information. Everyone's chins drop to the floor because they just heard the word Giga and they picture their 4 Gigabyte drive squeezing through a network line in seconds...

    Gibi/kibi/mibi bits??? It's bad enough already. Plus those words do suck to pronounce.
  • You can pick us programmers when we start speaking in code, but things like "!=" aren't especially more interesting than things like ":-)".

    It's part of the net culture, so deal with it =P

  • Oh, now FluptyByte I like! How can we campaign to get this accepted?


  • No.

    Bad Idea(tm).

    Bits are one thing that you can't have fractions of.

  • No.

    != is not a C-ism : it's just looks rathar a lot like 'does not equal' sign, which the C operator happens to make use of as well.

  • christ i feel dumb now.
    maybe i should refrain from posting before my coffee...
  • With binary units might it be best to stick with binary?

    Certain base numbersystems are best for certain tasks, base 10 isn't especialy good for anything.

    As soon as I come up with a convincing way to speak in hexidecimal or octal I'll no longer use base 10 as my primary number system.

    I may even have to do a writeup and attempt to get it posted as a news story here.

  • Okay, kids. I can understand why they'd want to differentiate between terms, but can't they pick something that sounds better?

    Geeks already have enough stuttering problems. We don't need any more hard-to-say terms mucking with our ability to avoid sounding like a 2-year-old. "I'd like 512 Mebb-mebibibib-eeebibytes of RAM, and a 21 Giga, er, googa, I mean, giBIBBIbyte hard drive. I'm building a Lihnooks, um, Lig NUX, uh, LeeNix baux." Yeah.

    On the other hand, I once heard someone stutter in the coolest possible way, while saying "check it", or rather "ch-ch-ch-ch-ch-ch-ch-check it" (with lightning speed!). But I'm way off-topic now.

  • But aren't giga, tera, etc. derived from words meaning giant, monster, and not any power of 10?
  • Nice suggestions. I don't think we'd need to invent new logic to build computers to work in base 10, would we? I mean, the Eniac calculated in base ten.
  • How 'bout this: we already say "hey, I've got 128 megs of RAM" or "that's a 2 gig drive." So instead of changing decimal prefixes to something silly sounding, why don't we just "officialize" the current jargon like so:
    1 kilobyte = 1024 bytes
    1 meg = 1024 * 1024 bytes
    1 gig = 1024 * 1024 * 1024 bytes

    Ok, so we may need to revisit kilo and tera, but I like this better....
  • 10 megabit Ethernet refers to 2^20 bits (not bytes)

    Wrong. 10 megabit Ethernet is 10*10^6 bits/s. In communication engineering, kilo is 10^3, mega is 10^6, giga is 10^9. Rates are specified as bits/s or symbols/s.

  • Of course, by my suggestion, you *can* do both. Memory is "32MB" while a Zip disk could be 100 mB, to distinguish a "true" binary-based "mega" from a decimal-based one. The key here is to use whatever system makes the most sense. In specifying memory size, maybe binary makes sense. In displaying a file size while showing only the most significant digits, decimal makes more sense.

    Uh... of course, people actually used to SI/Metric might read mB as milliBytes as lowercase "m" is the prefix abbreviation for milli (1/1000). Not that a millibyte makes any sense, but it adds another level of possible confusion. At least KiB et al are different enough to survive capitalisation mangling.


  • several people have correctly pointed out that "m" and "M" already have different meanings in SI units...silly me, I totally forgot that (if I still did any EE stuff I probably wouldn't have :-) )

    So, maybe Knuth does have the answer (no real surprise there). Use "K" for 1000, "KK" for 1024. And so forth.

    Dunno. S'long as I don't get confused when reading stuff on the screen, I'd be happy.
  • They'd say you lose 1.475GB :)
  • I'm all for engineering notation here (scientific notation with the exponent forced to multiples of 3 - makes real world transformations much less error-prone, which, after all is why it was developed.)

    The *correct* meaning of "Giga" is completely unambiguous: it's 1 x 10^9. Why is it that the traditional CS "computer twit" types insist on the sorts of ridiculous abbreviations and approximations that the rest of us know will sooner or later land them in trouble? 2^10 != 10^3, and it never has. This is just plain lazy usage - the terms "kilo", "mega", etc. should NEVER have been twisted in this way. Let's just toss base-2 nomenclature for good NOW while we can.

    e.g.: Why is Y2K an abbreviation? Isn't that the tinking that got us into this mess in the first place? (I'm all for at least five digit year fields, or maybe ints, or maybe even long ints... [grin])
  • As was brought up a long time ago, this system brings rise to "Gibs" being the spoken short form of Gibibytes... Let ut not forget

    Kibs, Mibs, Tibs, ...
    I think this brings new meaning to "Kibble and bits and bits and bits".
  • This makes things easier, and clearer. Now those hard drive companies can't try to trick the not so bright...
  • There's absolutely nothing in science that requires this. The only thing that matters is that people know what is meant. If we all accept a pite (314159 bytes) to be the new unit of storage space and work with that, then everything will be just fine.
  • by Anonymous Coward on Tuesday August 10, @07:46AM EDT (#) why is it so obvious the binary measures need new names? they are far more universal measures than the decimal ones, so why not come up with new names for them instead! MB = 1024*1024, like it should be

    This is a joke right? The other prefixes have been in use for at least 100 years in scientific literature and is used by people everywhere but the US.

    The change your proposing is sort of like changing the C standard to conform to say microsoft's visual c's quirks .

  • Glytch sez:

    Chibi-bits?! Ugh! I can just imagine a mutie-freak HD with pink hair being installed in my machine...

    Never mind that...would Macintoshes label Chibi-bytes as Sammy-bytes? Would TurboLinux start describing things in terms of Kerbybytes (making it REALLY confusing if you happened to be running Kerberos)? Would we have sama-bytes and sensei-bytes and chan-bytes? :)

    OK, so it's blatantly obvious that I watch more anime than is generally regarded as healthy. :) I'll also note that, as a rule, I genreally can't watch more than about five minutes of Sailor Moon without bleeding eardrums and/or tooth decay setting in, and I prefer shows that take the piss of magical princess shows (like Magical Girl Pretty Sammy--actually a Tenchi Muyo spinoff--or Card Captor Sakura (yeah, it's magical princess, but not tooth-decay inducing)...).

    ObSlashdot: Yes, there are computer puns/references above. Kerberos (aka Kerby-chan) in Card Captor Sakura appears as a winged stuffed lion thingie (a very kawaii stuffed lion thingie) whilst Pretty Sammy is in general a pretty savage satire of Sailor Moon in general and (in episode 2) has an extremely wicked funny parody of Microsoft :) (Which is even funnier when you realise the two best-supported OS's for Kanji are MacOS and TurboLinux, and the Japanese version of Win95/98 blows goats even worse than the American version--to the point many Japanese consider it literally unusable...:)

  • Usage, not standards or flamewars, will decide which words survive. Use the best word for the job. Is your goal to communicate or to obfuscate? Are you talking to nerds or avrage joes?

    • megabyte
      • precise? no (useful for obfuscation/marketing)
      • pronounceable? yes
      • recognized by masses? 60%
      • understood by masses? 30%
      • understood by /. readers? 100%
    • million bytes
      • precise? yes
      • pronounceable? yes
      • recognized by masses? 100%
      • understood by masses? 50%
      • understood by /. readers? 100%
    • mebibyte
      • precise? yes
      • pronounceable? mebi
      • recognized by masses? 0%
      • understood by masses? 0%
      • understood by /. readers? 80% (now)
    • zorkybyte, prollybyte, mephthobyte, and others invented here
      • precise? no
      • pronounceable? some
      • recognized by masses? 0%
      • understood by masses? 0%
      • understood by /. readers? 0.01%
  • The prefix Giga is really supposed to be pronounced 'Jiga', like Doc and Marty in 'Back to the Future' ( 1 POINT 21 GIGAWATTS?!?! WHAT THE HELL IS A GIGAWATT! ), and people don't pronounce _that_ correctly -- I mean, who wants a 10 'Jig' Hard Drive.. :) Compared to this, the new prefixes look like they WANT us to mis-pronounce them...
  • Your comment is correct for bits. But, all during my career in the computer field (22 years and counting), BYTES -- not bits -- has ALWAYS been defined in terms of base-2 values, e.g.:

    1 * 2^10 = 1 Kilobyte
    1 * 2^20 = 1 Megabyte
    1 * 2^30 = 1 Gigabyte
    1 * 2^40 = 1 Terabyte
    1 * 2^50 = 1 Petabyte
    1 * 2^60 = 1 Exabyte
    ...and so on...

    It's only the hard drive makers, starting with the IBM PC/XT, that screwed with the definition to inflate their X-byte capacity claims.

    I think the confusion here is whether one is talking about bits or bytes. The engineers are talking bits; the programmers bytes.
  • Very simple, very logical, and very consistant with the spirit of the original meanings.

    The hard disk industry has been around for decades longer than the PC industry. The metric system has been around since the French revolution. The prefixes are from a language that's millenia old. Now, tell me again how hard disk manufacturers are being inconsistent with the original meanings?
  • beh writes: Personally, I'd vote for the SI honouring the computer science business and officially declaring, that with regards to computing the prefixes are based on powers of 2, thereby forcing companies like Maxtor and the like to change to this way.

    I'd like to point out that companies like Maxtor and other harddisk manufacturers are part of the computer industry. The entire problem with using powers of 2 vs powers of 10 is the inconsistency. If powers of 2 were used always in a computer context, there would not have been a proposal for kibi and friends.

    --- Abigail
    How many vibrations/sec is a MHz?

  • 00mB, according to convention, would represent 100 milliBytes...

    ...or about 1/10 byte. Since 1024 is about 1000, and 1/8 is about 1/10, I declare that 100mB = 1b.
  • Seriously, is anyone here gonna buy a HD measured in gibibytes? Sounds like something a kid would talk about.

    "Yeah, I got a 50 gibi HD yesterday"

    Or maybe you'd say 50 gib. Are these people BeeGees fans or something?
  • Draconian writes:4 GB (decimal) sounds better than 3.7 GB (power-2). All the confusion is just because of marketing gurus trying to make their product look good. Thank god the memory chip companies don't follow this approach to advertising.

    Yeah, good old chip companies. Does that mean my 400MHz chip is really doing 419430400Hz?

    --- Abigail

  • Hard drive makers.

    Ever since the invention of the term "byte," it's been understood that it's powers of two, at least by the technical community.

    By using a word construct like [prefix]byte, you know it's power of two. Of course, since the general population is full of clueless lusers, the hard drive companies started using the terms such as megabyte and using the terms under their SI meanings... powers of ten. Why? Because the numbers are smaller, and the public will assume it's "computer" numbers, and not "metric" numbers. Of course, when people complain the hard drive manufacturers can say "but 'mega' means million, not one million, forty-eight thousand, five hundred seventy-six."

    I counterpropose that we make hard drive manufacturers adhere to the 2^10 kilo system.

  • Please somebody tell me that the Teletubbies havn't invaded computer jargon...

    More to the point, which inspiring person came up with theses names... kibibytes... OK computer jargon doesn't sound good at the best of times but even so... kibibytes... kiddybites, the new, tasty snack from Haribo...

    Why not invent a new word from bytes ? I'm not known for my imagination, but woulnd't something like kilobets or kilobats (going along the lines of B-i-T-s B-y-T-e-s...
  • As an Electrical/Computer, I readily admit that IEEE has made some REALLY stupid standardizations in the past. Look up what they use for and and or gates in terms of sucks...

  • I agree! AFAIK a byte is not an SI unit. So adding kilo to doesn't change that fact. So we have accepted that 1 kilobyte = 1024 bytes and 1 megabyte = 1024 kilobytes. It is the hard drive companies that are in the wrong. If they want thier "industry standard" to be something other than what is accepted usage then they should change thier nomenclature. some other folks here have suggested some possible words. Mine are maybebytes(mbB), mightbebytes(MbB) and bugbites(oucH).
  • This reminds me of when the Police (the band, not the blueshirts) put out a song called "Ah Doo Doo Doo, Ah Da Da Da" specifically because they thought it would be funny to hear a bunch of the boss jocks try to say it and still sound cool. It will take a lot longer to get popular simply because a lot of alpha geeks out there wouldn't be caught dead talking about their code in terms of kibis & bits.

    Another problem I have with this system is that it is still tied to the base10 numbering system, setting the markers around every three tens place. Admittedly, this is the way we think about them currently, but let's apply a little Sapir Whorf [], eh?

  • trying repeating the newly proposed terms to yourself. They don't exactly roll off the tongue, do they? I'm not suggesting the currently accepted terms are smoothly enunciated, but they're much easier than "MEH-BIH-BTYE". What would the shortened form be anyway, Meb? c'mon. Perhaps it's just that I (as well as the rest of us) am used to the current terms and their shortenings (meg, gig)...I am of the opinion vendors should standardize upon the classic definition of 2^x bytes instead of 10^x.
  • after more than 15 years of common usage,
    the vocabulary is certainly not going to change...

    Just always assume that the capacity is
    expressed in billions of bytes when buying
    a hard drive. And don't worry that some
    company might be at a disadvantage compared
    to others because it uses the correct definition
    of 'Gb': none of them do.
  • Hey, a good use of a M$-style License Agreement!
  • Nice try, but I imagine /. is out of his jurisdiction. I doubt kibo will reply.


    (hey, if you get a response from him, i want one too)
  • I'd like to add that the other side of this is, many people can't figure out the difference between a KiloByte, and a MegaByte let alone a Kilobyte and kibibyte or a MegaByte and a Mebibyte. These new notations will simply confuse the matter more for many many people.

    I do understand the methodology, but I'll also say that different names would have been better. I mean they aren't even easy to pronounce...

    My two cents worth anyway.

  • Instead of changing the recognized power-of-two nomenclature, simply introduce a couple new units *for use with hard drives*

    1 Kilobyte would still be 1024 bytes, etc.

    The new units:
    Weaselbyte (WB) - 1,000,000,000 bytes (The HD industry Gig)
    Slimybyte (SB) - 1,000,000,000,000 bytes (When drives get that big)

    Who knows - the HD manufacturers might get sick enough of having to say "The new standard in storage - the RonCo Flame Muffin 7200 RPM Hard Drive - Features Capacity up to 100 Weaselbytes!" that they would start reporting sizes like normal people.

  • Not that it really affects the discussion much, but this is news that has been floating around for a long while.

    the Rapidly Changing Face of Computing did an article on this ibibytes_A_New_HighTech

    Which has a link to the original source 03.htm#Information Technology

    Not that I really mind too much, but this dates back to March of this year!

    Apart from that, I think the names sound pretty odd - don't you? I agree with Jeff Harrow from RCFOC who says that a Kibibyte sounds like a type of dogfood.

    Oh well, like most things of this nature, I guess nobody will really ever use them except to show off their knowledge. (That'll be me then!)

  • I'd hate to see what would happen if we switche units again: K, then KK...What next?
  • Instead of redefining the prefixes, of which there are many, noone has considered redefining the byte itself. My proposal is the "bite", a unit consisting of 8.192 bits. In that case, 1 kibibyte equals 1 kilobite, 1 Mebibyte equals 1 Megabite, etc. The standard abbreviation for a bite should be "B".

    Or, in a table: (bad formatting courtesy of
    Old New My system
    --- --- ---------
    1kB 1kiB 1kB
    1MB 1MiB 1MB
    1GB 1GiB 1GB
    1TB 1TiB 1TB
    1PB 1PiB 1PB
    1EX 1EiB 1EB

    I think the advantages are obvious.

    --- Abigail

  • Knuth's idea for MMB and GGB, along with this guys idea for "long" and "short" is great. We, the slashdot community, need to start using them EVERY DAY. That is how things change, and we can make it happen

  • Why must committees break perfectly good, working standards?

    The current standard is not a 'perfectly good, working standard. The current standard leaves a lot of room for miscommunication and misunderstanding.

    Computers use zero-based counting are usually binary - which means byte-boundaries occur in powers of two.

    This is fine as long as you are only looking at it from a 'computer' centric viewpoint. However, there is a lot of confusion when the 'computer' viewpoint interfaces with other viewpoints, e.g. the telecommunications industry.

    How long does a 56M file take to transfer over a 56K link, assuming no overhead? The first confusing factor is that computers tend to use bytes while telecomm always uses bits. The second confusing factor is that telecomm always uses the base 10 definition of k, M, G while the computers use the base 2 definition.

    If the average idi^H^H^Hperson can't understand this simple fact, Steve Jobs would love to sell them their next computer.

    The inet-access mailing list, a mailling list for ISPs has many flame wars that boil down to a misunderstanding of the k, M, G definitions. These people are not your average idi^H^H^Hperson. Very few on the list are Mac/Window users.

  • The prefixes isn't really anything new. This was talked about almost a year ago, even though I don't know if it reached Slashdot and the slashdot community.
  • 1 bit = bit
    8 bits = byte
    1024 bytes= 'K' (kay?)
    1024 K = Meg
    1024 Meg= Gig
    And let SI terminology stay the way it's always been. Of course, I don't know what abreviation we'll use for terrabyte. ter? I vote 'T'.

    M'KAY? :-P

  • 10**6 is almost 50k(47.someodd) less than 2**20... So on a 1 gig drive they save themselves 47,437k, or almost 50 meg. Doesn't sound like alot these days, but that means we're losing 1 gig on every 20 gig drive we buy...
  • I agree with you completely. While I understand what that did and why they did it, kibibyte just sounds so stupid! Just say it out loud. They can come up with something that has as much meaning but doesn't make you feel like a 2nd grader when you say it. Could just call it Kbytes or kibyes or something. The whole *bibytes thing's gotta go.

    The only reason I keep my ms-dos partition is so I can mount it like the b*tch it is.

  • Why do we need to conform to SI when Kilobyte and Megabyte doesn't have prefixes anyway.

    These are just single words with the following definitions: Kilobyte = 1024 bytes; Megabytes = 1024 Kilobytes.

    We don't try to disassemble "re-ally" (to go into the ally again?), so why should we assume Giga-bytes.

    //yeah, whatever
  • Damn clicking on the wrong button. Here is the URL again and this time clickable. []

  • Clearly unnecessary. There is an established convention that k means 1000, except when talking about bytes. Then it means 1024. And so on for M and G and T...

    All we need here is a way to correct certain harddisk manufacturers. Using laws against bad advertising perhaps. Those that need to know that k=1024 in some cases have no problems with it. Those too ignorant to grasp it don't need to, and they couldn't care less either.
  • Yes, in Catalan is the same way. And the US way is not used in Spain either, as fas as I know.
  • Let's go ahead and point out the obvious.

    This whole thing is a plot perpetrated by the "MiB" led by an attractive Will Smith look-alike. The plot is to "gib" people they think are too smart.

    hmmm...noisy cricket as a Quake III weapon. Interesting.

  • My ls, "ls (GNU fileutils) 3.16", doesn't have that option. I tried before I posted of course. Strange that fileutils-3.16 would have it for df but not ls. Guess it's time to upgrade to 4.0.
  • Why on earth would anyone consider this? This is a function of the engineering community being driven by corporate interests. Why would anyone need some new definition? Not to mention ones with ridiculous names. A megabyte is already defined, it's just that certain companies purposely misinterpret it to their advantage.
    Mega = 10^6
    Byte=8 bits
    MegaByte= 8E6 Bits
    wow! that was difficult............
  • The names were so dumb, I reread the posting carefully with my joke detectors on high alert. I finally concluded it was serious. Reminds me of the "tongue troops" in France and Quebec, protecting the purity of their mother tongue by discouraging certain words (e.g. "weekend") that invaded from another language.
  • by Masem ( 1171 ) on Tuesday August 10, 1999 @03:20AM (#1755206)
    Anyone else read that as "Kibobyte"? We really
    want a computer term that closely related to
    Kibo?? ;-)
  • -h, --human-readable == print sizes in human readable format (e.g., 1K 234M 2G)

    -H, --si == likewise, but use powers of 1000 not 1024

    GNU fileutils 4.0, November 1998

    Well, in Linux anyhow, I notice FreeBSD uses 4th Berkeley Distribution, May 8, 1995, and my IRIX boxes are totally out of date, I can't tell what OSF1 is using, but it doesn't support -h or -H either.

  • Have, for a LONG time, "cheated" on the megabyte thing, using powers of ten rather than powers of two. It allows them to sell less disk space for more money, and still convince people they're getting a bargain.

    Personally, I don't think there's anything confusing about megabyte, gigabyte, etc. It's 2^0, 2^10, 2^20, 2^30, etc. Very simple, very logical, and very consistant with the spirit of the original meanings.

    Kikibyte sounds like some Polynesian parrot.

  • by Signal 11 ( 7608 ) on Tuesday August 10, 1999 @04:33AM (#1755284)
    Maybebytes: It could be a byte. Then again, it might not.

    Gibytes: How many bodies are on the floor after you get done playing Quake.

    Kilibytes: What you call somebody who has contributed to the Gibytes of another player.

  • Believe me, if they can, they will. They'll round off as they always have, and all we'll have gained is a more obscure, longer to write, name.

"The number of Unix installations has grown to 10, with more expected." -- The Unix Programmer's Manual, 2nd Edition, June, 1972