Why does Explorer use the term KB instead of KiB?


Although the International Electronic Commission established the term kibibyte for 1024 bytes, with the abbreviation KiB, Windows Explorer continues to use the abbreviation KB. Why doesn't Explorer get with the program?

Because nobody else is on the program either.

If you look around you, you'll find that nobody (to within experimental error) uses the terms kibibyte and KiB. When you buy computer memory, the amount is specified in megabytes and gigabytes, not mebibytes and gibibytes. The storage capacity printed on your blank CD is indicated in megabytes. Every document on the Internet (to within experimental error) which talks about memory and storage uses the terms kilobyte/KB, megabyte/MB, gigabyte/GB, etc. You have to go out of your way to find people who use the terms kibibyte/KiB, mebibyte/MiB, gibibyte/GiB, etc.

In other words, the entire computing industry has ignored the guidance of the IEC.

Explorer is just following existing practice. Everybody (to within experimental error) refers to 1024 bytes as a kilobyte, not a kibibyte. If Explorer were to switch to the term kibibyte, it would merely be showing users information in a form they cannot understand, and for what purpose? So you can feel superior because you know what that term means and other people don't.

For an explanation of other storage units, you can consult this helpful chart from xkcd.

Comments (74)
  1. ERock says:

    Never underestimate the burning need to feel superior to others.

    Space really should be described in bits since 8-bits-in-a-byte is platform specific. :)

  2. TJ says:

    It seems to me that the only people in the entire computing industry who don’t mean 1024 bytes when they say "kilobyte" are the hard disk manufacturers, though USB flash drive manufacturers are starting to use 1000-byte kilobytes as well (my old 128MB and 512MB jumpdrives were actually exactly 128MiB and 512MiB, my new 32 GB one is actually only 29.8GiB). I used to have a stack of Seagate 810MB hard disks from around 1994-ish that someone gave me and they were all exactly 810MiB. I never understood why they switched.

  3. Someone You Know says:

    @TJ

    I always assumed they switched so that they could charge you the same price and not give you as much for it.

  4. Adam V says:

    I like the Kelly-Bootle Standard Unit of 1012: compromise between two sides by choosing something that’s worse than both. Success!

  5. Aaron Lewis says:

    One standard versus years of practice saying it the other way… Yeah, I think the IEC lost this one.  I’m sure they’ve accepted it by now.

    It’s nowhere near as annoying as hearing people regularly use "there’s" when talking about multiple subjects.  For example, "There’s three of them."

    I guess, if you want to find ways to alienate yourself from the general populace, it’s easy enough to do without finding a standards document.

  6. Neil (SM) says:

    There’s also the concern that saying "gibibyte" or "mebibyte" out loud will make you sound like a doofus. Perhaps one more reason it never caught on.

  7. zzz says:

    When I first heard these kibi mebi KiB MiB terms I simply thought that they sound and look ridiculous. Beyond that I haven’t pondered about the technicalities since those are irrelevant if it’s obvious no one is going to use these in the first place.

    Maybe the kids who wRiT3 ‘leet’ have some appreciation for the KiB and MiB, so when they grow up … they’ll come to realise how dumb that looks.

  8. Florian W. says:

    Hey Cool!

    So everyone can say, Microsoft didn’t do it, so I don’t have to do it as well.

    But who must start with the switch, so Microsoft is following? Apple?

  9. Cheong says:

    Now, to make things a bit more puzzling for customers, the ISPs used kilobits for measuring network speed.

    I guess I can call my 250GB harddisk 2 terabits to make it sound bigger. :P (And you guessed it, there is also kibibit that can be used…)

  10. TJ says:

    Let’s just end all the confusion and have some standards-keeping organization declare that KB MUST mean 1024 bytes, MB MUST mean 1024 KB, etc, and then pretend KiB and MiB and such never existed. And then we’ll all gang up on those who use their own weird conversions (hard disk manufacturers) and force them to follow the new standard in their labeling. And then the next time I buy a hard drive it’ll be a few dozen gigabytes bigger. (Or it’ll just be labeled 938.4653 GB)

  11. Karellen says:

    "Why doesn’t Explorer get with the program? Because nobody else is on the program either."

    That is not entirely correct. I just opened up the standard file manager on my system – Dolphin (v4.2.2) – and file sizes there are listed in KiB and MiB.

    Oh, you meant "nobody else with market share greater 0.01%" :-)

  12. While we’re on the subject, how come this stupid speedometer in my car shows m/h instead of km/h?

  13. nathan_works says:

    The MiB would be quite amused by this discussion.. Watch out for the black ford’s coming for you now..

  14. LionsPhil says:

    Yeah, Linux file managers have boldy strode forward as the ones to go first on this…and, frankly, I hate it. The terms are stupid and the extra "i" all over the place just adds visual noise. I’m quite glad that Apple and Microsoft are ignoring it.

    (Raymond DID say "to within experimental error". As far as end-users go, Linux /is/ an error. Heh heh.)

  15. Erzengel says:

    Maurits: Really? I live in Southern California and have an American car. Yet my speedometer has both m/h and km/h. Granted the km/h is smaller and under the m/h, but it’s there.

    Karellen: Did you not notice "within experimental error" Raymond put for people just like you? The amount of people who use kibi, mebi, gibi, etc, are negligible.

    Raymond:

    Just playing devil’s advocate.

    I would say using KiB over KB is more of historical reasons, much like a lot of things you blog about. Everyone knows KB, and it takes a lot of effort to overcome momentum. Nearly no one is going to try to overcome the momentum because it’s very costly. The only people who could make a significant enough force is Microsoft. If Microsoft doesn’t do it, no one else will.

    While I personally don’t think there’s much confusion about Kilobyte (generally) meaning 1024 bytes, historically non-standardized units served for significant confusion. The differences between various countries Feet and Inches caused much confusion. As such, using Kilo to mean different things is probably a bad thing.

    Also, Kibi sounding funny reminds me of how everyone reacted to the naming of the Wii. But that turned out very well, so if Kibi were given a chance to enter normal vernacular it would quickly become less funny and more useable.

    That said, I don’t really think it’s necessary.

  16. Me says:

    Way to go Microsoft. These unit names are just ridiculous.

    I don’t get why the driver manufacturers have to use the 10 based units. They gain nothing by having the numbers sound higher, because all their competitors do it, too. Why not just use the term that makes more sense, then?

  17. I think the use of KiB and MiB is ridiculous (it is all over the Wikipedia, so there isn’t hardly any day I don’t see it). It’s true that the difference in the definition of an inch has caused troubles in the past. But now we aren’t talking about half a dozen different definitions of KB. There is only one (storage media manufacturers apart), and it is worth 1024 bytes. Stop. I think there is no possible confusion. And if somebody is going to think that a Kilometer is 1024 meters long because of that, then s/he doesn’t understand the difference between different magnitudes, and will have bigger troubles solving a Physics problem.

    IMHO, the solution would be enforcing the current standard on those who break the rule (manufacturers of hard drives, flash chips and optical disks), instead of making everyone else pay for the mistake of a few.

  18. Leo Davidson says:

    "(it is all over the Wikipedia, so there isn’t hardly any day I don’t see it)."

    Yes, it seems to have taken hold at Wikipedia for some reason.

    The Mebibyte defintion on Urban Dictionary really made me laugh:

    "A new, weird and unnecessary unit of data capacity, created by some idiots of IEC and aggressively advertised on Wikipedia."

  19. someone else. says:

    “I don’t get why the driver manufacturers have to use the 10 based units. They gain nothing by having the numbers sound higher, because all their competitors do it, too. Why not just use the term that makes more sense, then?”

    You answered that yourself: All the competitors do it. So if DriveTech uses binary units and SuperStor uses decimal, SuperStor would have an advantage. So DriveTech won’t use binary units.

    For a similar reason, magazines (at least computer magazines) are often dated higher, i. e. the July issue is numbered 8 … or even 9. Or, taking the confusion one step further, the issue coming out in July is named August and numbered 9.

  20. me says:

    Ouch… I see that there was another posting from someone called "me". The last 3 postings of "me" (all with lowercase ‘m’), including this one, or really from me. The other "Me" (with uppercase "M") is not from me. ;)

  21. someone else says:

    @me

    I don’t think there is an SI unit for storage capacity. Therefore, it is impossible to use SI units. And if no SI units, why SI prefixes? This sound as far fetched as a kiloyard.

  22. Duke of New York says:

    "then we’ll all gang up on those who use their own weird conversions (hard disk manufacturers)"

    It’s not their own, nor weird, if it follows the SI conventions. Hard disk manufacturers use SI-correct units because hard disks don’t have to decode address bits. The number of bits on the disk is determined by geometry, not circuitry.

    Give things a little thought before assuming that they’re just weird for weirdness’s sake.

  23. me says:

    It can actually get quite confusing at times, with the general rule of thumb being measurements of speed are in powers of 10 and measurements of storage are in powers of 2, unless, of course, you’re a storage manufacturer who uses powers of 10 to mislead consumers.

    But, as an example, say you’re a network engineer running some tests and you transfer a 10MB (as in 10 x 2^20) file over a 80Mbit connection (as in 80 * 10^6). The result is you’ll be measuring a file expressed in powers of 2 in speeds measured in powers of 10.

    Needless to say, when you find these kinds of situations, things can get confusing if accuracy is important while trying to intepret the results measured in two entirely different bases.

  24. porter says:

    More confusing than the 1024 != 1000, is that harddisk/cdrom etc manufacturers give us the unformatted size of the media.

  25. me says:

    @someone else:

    You are mistaken: There is an SI unit for storage capacity: It is "1"! Thus, 5 bit/s is 5 1/s.

  26. Leo Davidson says:

    @me: That’s true, networking stuff often uses 1000s.

    At the same time, most software displays upload/download speeds in kilobytes (1024) per second, at least if we’re talking about consumer stuff  (e.g. FTP clients). (I can’t speak for network-admin tools as I very rarely use them.)

    I realise the reported connection speed in the tray^H^H^H^Htaskbar notification area is a an exception. I don’t think redefining what the units mean in some existing tools will resolve that unfortunate ambiguity, though. It’ll just add to the confusion.

    I guess you have that issue with software which displays advertised information about storage/network hardware. IMO it’s less important with storage because some of the size is "lost" due to formatting anyway.

    I remember early versions of the popular online game distribution client whose name begins with S used to report speed in kilobits/second. There were lots of complaints as people found it confusing and irritating that they had to convert numbers in their head to work out how fast things were going in units they understood. I think the client was eventually changed to display things the same as everything else. (I know for sure they promised to change it, but then didn’t for ages, and I think they eventually did. Might be wrong about the last part.)

  27. someone else says:

    @me

    So drive manufacturers etc would have to write a capacity of one Gigabyte (decimal) as “1 G”?

  28. Nick says:

    @Leo

    It’s all over Wikipedia for the same reason that the CE/BCE crap is.  A bunch of editors with superiority complexes can’t help changing things for the sake of change.

    @Raymond

    An XKCD reference on the ONT.  My life is now complete!

  29. Duke of New York says:

    Still waiting for someone to explain why using the multipliers as defined in SI (and required by law) proves some kind of intent to mislead…

  30. ChrisMcB says:

    If you are going to figure out how fast you can send a file over a network you have to do some math no matter what units the size of the file is displayed in. There is overhead involved with the network. You can’t say, ohh this network is 100 bits/s and my file is 10,000 bits so it will take ten seconds. Cause your 10,000 bits aren’t the only ones being used.

    Personally I am less concerned with Mega meaning 1000 vs 1024 (whats that a 2% difference?) as I am with B meaning bit or byte (sure some people use b for bit and B for byte, but not everyone) Now you are talking almost a magnitude of difference.

  31. s_tec says:

    Somebody is missing a big opportunity here. If software vendors would switch to the real base-10 SI units, their customers would see an apparent 4%-7% increase in capacity. Think of the competitive advantage!

    Besides, the base-2 units are only useful for memory chips with power-of-two address decoders, not for disk drives, network hardware, or anything else. Base-2 is also worthless from a consumer’s point of view. Does it really matter if a memory stick has 34GB instead of 32GB? Numbers like 16, 32, 64, 128, 512, etc. are already pretty meaningless, so switching to 17, 34, 68, 137, etc. wouldn’t be any more meaningless for the average person.

  32. microbe says:

    I worked at a storage company (there is recently acquisition news around it) and we certainly used KiB in our product.

    Please, not everyone is in a windows world.

  33. Eric B says:

    I think Microsoft already had the best solution:

    C:>Dir

    939 File(s) 10,468,918,236 bytes

     6 Dir(s)  242,108,547,072 bytes free

    See? No confusion

  34. bratch says:

    @microbe – that’s because the storage company is within the experimental error segment.  

    This isn’t ^H^H^Hsomething limited to ^H^H^Windows.  Even other^H^H^H^H^H^H comments appear to be from non-Windows platforms, or at least using other^H^H^H^H^Hdifferent browsers/editors.  

    It just sounds awkward, almost like saying, "I binged your company and found this …"

  35. John says:

    Eric: Why do you have 900+ files totaling ~10GB in the root of your drive?

  36. Warll says:

    Well I myself like standards, not just computer standards though.  If Kilo means one thousand everywhere else why should I have to learn that some fools decided that it should really mean 1024 for computers.

  37. Michael G says:

    @someone else:

    Non-SI unit, SI-prefix, common usage: "megaton"

  38. And then there’s those 1.44 MB floppies which are neither 1.44 megabytes nor 1.44 mebibytes — they’re 1440 kibibytes, or 1.44 * 1000 * 1024 bytes.

  39. Jonathan says:

    None of this would’ve happened if computers had been developed in east-Asia, where the primary unit is 10^4 rather than 10^3. Nobody would’ve thunk to 10000 with 8192 or 16384.

  40. n says:

    An FYI:

            10^X     2^X      Diff   % Diff As named  

     0  0 1.00E+00 1.00E+00 0.00E+00  0.00% Byte      

     3 10 1.00E+03 1.02E+03 2.40E+01  2.34% K          

     6 20 1.00E+06 1.05E+06 4.86E+04  4.63% M          

     9 30 1.00E+09 1.07E+09 7.37E+07  6.87% G          

    12 40 1.00E+12 1.10E+12 9.95E+10  9.05% T          

    15 50 1.00E+15 1.13E+15 1.26E+14 11.18% P          

    18 60 1.00E+18 1.15E+18 1.53E+17 13.26%

    21 70 1.00E+21 1.18E+21 1.81E+20 15.30%

    24 80 1.00E+24 1.21E+24 2.09E+23 17.28%

    27 90 1.00E+27 1.24E+27 2.38E+26 19.22%

  41. Random User 43789 says:

    TJ:

    The problem there is that tera-, giga-, mega-, kilo-, etc. are all SI prefixes, that correlate to powers of 10. Somewhere along the way, some "geniuses" decided to use kilo- to represent not 10^3 (1000), but rather 2^10 (1024), merely because "it was close". From there, the die was cast and the larger prefixes were mangled as their time came.

    Redefining the terms solves noth^h^h^h^h^h very little, as there would *still* be a divide between "computer" units and all other units of measure.

  42. Ives van der Flaas says:

    Well, K isn’t an SI prefix. k is.

    Why doesn’t IE use kB?

  43. Ives van der Flaas says:

    Ofcourse I meant Explorer.exe instead of IE.

  44. slq says:

    <blockquote>It’s nowhere near as annoying as hearing people regularly use "there’s" when talking about multiple subjects.  For example, "There’s three of them."

    I guess, if you want to find ways to alienate yourself from the general populace, it’s easy enough to do without finding a standards document.</blockquote>

    Ahh, the sweet taste of irony.

  45. configurator says:

    http://www.google.fr/search?q=1+kilobyte+in+byte

    I think that settles it. A kilobyte is officially 1024 bytes.

  46. Leo Davidson says:

    If the new proposal had defined two new suffix sets, one for the 1000 stuff and another for the 1024 stuff, then I might have been in favour of it.

    Instead the proposal simply seeks to replace one ambiguous system with another ambiguous system. It doesn’t reduce complexity or confusion; it adds to it. The real agenda seems to be to reclaim the SI units at any cost in complete ignorance of established practise.

    Sure, "10 MiB" is never ambiguous. It might not be understood by many people, and it might be ugly and unnecessary but it isn’t ambiguous.

    "10 MB" on a computer screen was never ambiguous, either, though. It always meant 10 * 1024 * 1024 on a computer screen.

    If this proposal became widely accepted then nobody would know if a program displaying "10 MB" meant "10 * 1024 * 1024" (as almost all do currently) or "10 * 1000 * 1000 and by the way I was written by a dork with a superiority complex."

    It’s a stupid, stupid idea and I’m glad Microsoft and just about everyone else are ignoring it.

    (I don’t dispute that it’s unfortunate we ended up with the current system but you can’t just wish it away and try to convert people to a conflicting system. Alternative system? Maybe. Conflicting system? Geddouddaheya!)

  47. someone else says:

    @Michael G

    A megaton (no matter if you mean mass or energy equivalent) is not an SI unit.

  48. me says:

    The KiB and MiB and so on might sound (and look ridiculous). However, almost all countries which use the SI have some low that *only* the SI units and the SI prefixes can be used. If you do not use them, you can get problems.

    In Germany, for example, you are not allowed to use anything but what the SI defines (with some exceptions, but there are not that many). If you use something else, you can get a "Abmahnung" (is it "call for order" in English?) from a lawyer where you have to pay money to him.

    This sounds ridiculous for KiB vs. KB? Indeed. However, lawyers always state that the HD manufacturers (and the network providers, which also use K=1000 in most cases) are the only ones doing it right. I have not yet heard about a lawyer doing a "Abmahnung" for K vs. Ki, or M vs. Mi. However, they already have done so for companies who try to use 17" LCDs. The inch is not an SI unit, and thus, they have to say 43,18 cm LCD. You see, when they can get money, lawyers start doing stupid things.

    Thus, I am waiting for the first one to create an Abmahnung because of this.

    Just my EUR 0.02.

  49. me says:

    @Leo Davidson:

    ‘"10 MB" on a computer screen was never ambiguous, either, though. It always meant 10 * 1024 * 1024 on a computer screen.’

    The classic 10 MB Ethernet has 10 Megabit/s: 10 * 1000 * 1000 bit/s.

    ISDN has 64 KB/s: 64000 bit/s, not 65536 bit/s.

    Thus, it was always ambiguous, but almost noone noticed, because the difference was very small. With MB and GB, the difference is much higher, so people find out that there is a difference.

    Just my EUR 0.02.

  50. Miral says:

    @someone else:

    It is an SI-derived unit though.  Which is close enough.

    @Aaron Lewis:

    "There’re" might be more correct when talking about plural subjects, but it sounds much more awkward than "there’s".  So it’s not surprising.

  51. Joel says:

    My theory on "there’s" was that it really just introduces a statement that some thing or things or situation(s) exist.  It’s an indivisible unit when used that way.  It’s also easy to start off a sentence with "there’s" without thinking about the nature of what "there" is.

    Also, try to reconcile this one: "there’s a lot of people in here" or "there’re a lot of people in here".  No matter what, you have surface number disagreement.  We decided a while ago that "a lot" and friends take the number of their objects despite being grammatically singular.  The reverse can happen with "there’s".

  52. I like the Kelly-Bootle Standard Unit of 1012: compromise between two sides by choosing something that’s worse than both. Success!

    "Should array indices start at 0 or 1? My compromise of 0.5 was rejected without, I thought, proper consideration".  (Stan Kelly-Bootle)

  53. Daniel says:

    Raymond,

    I’m sure you do not hear this often enough.  You are_the_shit.

    Underscores even.  Word,

    Thanks,

    Daniel

  54. Worf says:

    Firstly, anyone who blindly says "all computer units are in powers of two" brings up great ambiguity. First, kilo is 1000, always has been, always will be. Ditto on all other SI prefixes. We can blame early computing pioneers for this.

    The primary problem is there is enormous areas where computing and non-computing terms overlap, and having kilo`mean two different things is like having two different meanings for mile, gallons, ounces, etc. Oh wait, we have statute miles, nautical miles, US gallons, imperial gallons, troy ounces, avoirdupois ounces. Alas, they meet to great confusion oftentimes (does Europe really get better gas mileage? Or does the use of the larger imperial gallon skew things?). So why are we afraid of "kilo" and "binary kilo"?

    Next, I can’t believe the stupidity of some of the people here. You have a 2GB thumbdrive. Guess what! It has 2GiB of flash memory in it! But you lose memory in the management – flash translation tables (huge – these map logical sectors to physical flash blocks), bad block tables and bad blocks (up to 2% may be bad when new), spare sectors, etc. And in reality, every 512 bytes of NAND flash (used in all flash storage media – SSDs, memory cards, thumbdrives, etc) you have 16 bytes used for management as well (notably ECC data because there are numeous correctable errors), so your 2GB thumbdrive has, in reality, more than 2GiB in storage – it’s just a good chunk is chewed up by management information.

    I use the IEC prefixes because I strive for accuracy, not ambiguity. After all, didn’t we lose a spacecraft over SI vs. imperial? When I report information, I make it clear – e.g. "allocated 13000000 bytes (~12MiB)" (because a simple right-shift does it, versus dividing by 1000000).

    Oh, BTW, most mobile carriers charge data in SI units. Plus the over the air headers and trailers. When they say 5 cents a kilobyte, they mean 5 cents per 1000 bytes. Or around 900-ish bytes at the IP transport level, or 800-850 payload bytes. And 7.2Mbps HSDPA throughput is 7,200,000 bits/second, or 900kB/sec at the physical layer.

    (And face it, Joe average user thinks a kilobyte is 1000 bytes, mega is a million, etc. Give them a 12,582,912 byte file and they will say 12.5 megabytes.)

  55. Drak says:

    @Warll and others:

    You forget that when these sizes were thought up it was all JARGON. It was not meant for the general populace. The people who used it knew what was meant.

    Then marketing got wind of a way to abuse the terminology to make money…

  56. anony.muos says:

    But why does Explorer deliberately hide the sizes in Vista if nothing is selected and why does it not show free space on the details pane or status bar? Why are they empty? Was it a performance issue Microsoft couldn’t tweak? Why does Explorer hide sizes on the details pane for 15+ files  and again for 15+n files? Fix this extreme annoyances before RTM please. Why does it remove the ability for shell extension to display dynamic information in columns even if the user prefers to? Why have a status bar if all it shows is number of items selected, absolutely nothing else? Isn’t that info on the details "pain" as well? Why, oh why did Microsoft hide sizes beginning with Vista? It is for this reason that I will stay with XPMCE.

  57. Gavin Greig says:

    Perhaps a large vendor needs to set an example – like Microsoft? I’m sure there would be brickbats, but it might still be the "right" thing to do.

    I have some sympathy with the "it works OK as it is" argument, and I’m not sure that, put on the spot, I would make a different decision. Perhaps it’s more important than it at first appears though – by continuing to use KB to mean 1024 bytes, everyone who does so is undermining an international scientific standard (SI prefixes), the use of which extends far beyond computing.

  58. Ian says:

    ‘me’ correctly identified that bytes are a dimensioness quantity and there is no need for an SI unit for them. You wouldn’t talk about kilo-oranges or deca-biscuit tins.

    I personally have no problem with kB because everyone knows what it means.

    If you want to be strictly correct, kiB is really no better than kB, and just as odd as kibi food blenders.

  59. kB

    The Vista paint.exe says "Cm" in the file attributes dialog.  I have to admit that I filed a bug.

    Windows 7 RC paint.exe? "Centimeters".

  60. Brian Tkatch says:

    In other words, the entire computing industry has ignored the guidance of the IEC.

    The way i see it it was the IEC that ignored the guidance from the people.

  61. jeremy says:

    "And face it, Joe average user thinks a kilobyte is 1000 bytes, mega is a million, etc."

    I’m sorry, but the "average Joe" does not have any idea what kilo, meta, etc mean.  I know, I teach physics.  My gods, they do not know and not even god could change that for most of them.

    The people who do know understand there’s a difference.  And of all the people to complain about operator overloading, it’s the CS people!  Come on.  size(decimal) has a different NATURAL definition than size(binary).  Log_10(number) gives us order of magnitude for a number that’s "naturally" decimal and kLog_2(number) gives us the order of magnitude for a number that’s "naturally" binary!  Since we want 2^10 to be "kilo" instead of "10 giga" k=3/10.

    3/10 Log_2(2^20) = 6 = mega

    so 2^20 bytes = 1 MB.

    See!  Natural!

    Changing the *natural* size of a unit system to be base 10 is just perverse.

  62. Roman says:

    I’m an experimental error! I like the binary prefixes and consistency (and hate traditions).

    Plus I agree that for the general populace a multiplier of 1000 makes much more sense than 1024, and as such the decimal kilobyte should probably get more use than it does now.

  63. Anonymous Coward says:

    Erm… Explorer uses kB, with a small k, as it should as far as I’m concerned.

    (A while ago I bought a pack of empty DVD’s and I still feel a slight sting from the moment I realized Philips had lied about the capacity of the things.)

  64. spork says:

    The problem is "kibibytes" just sounds dumb to english speakers (and probably speakers of several other languages).  Bibbity bobbity bibytes.

    I recognise that it’s less than ideal to use decimal prefixes for base two quantities, but of all the names to choose, why the hell choose "kibi, mebi, gibi, tebi" etc. ?

    Retarded sounding, all of them.

    Consider:

    kidrabyte, medrabyte, gidrabyte, tedrabyte

    kismobyte, mesmobyte, gismobyte, tesmobyte…

  65. 640k says:

    Shocking news! Microsoft is not following standards.

  66. Common Sense says:

    It’d be nice if Windows Explorer offered an option to display file sizes with either binary (KiB, MiB, etc) or decimal (KB, MB, etc) prefixes. Use any of those as you choose. If we can specify a time format, why we can’t specify this one too.

  67. porter says:

    >  If we can specify a time format, why we can’t specify this one too.

    (a) With Open Source everything is configurable. :)

    (b) Because nobody would be able to agree on the acceptance criteria for testing.

  68. MIke Annonson says:

    "’me’ correctly identified that bytes are a dimensioness quantity and there is no need for an SI unit for them. You wouldn’t talk about kilo-oranges or deca-biscuit tins."

    It gets really silly when you consider what a byte is: 8 bits.  Another completely not-10-base container.  Then entire system is alien to natural human counting systems.  Still, appropriating the SI style prefixes seems real stupid these days.  It’s rather hard to know which base something is being measured it, having to remember lots of history to know how to convert back and forth.  I do like the whole kidra-, medra-, gidra- idea; it doesn’t sound like baby gibberish while helping to add definition.

  69. To MIke Annonson says:

    If you don’t like your kidra, medra system – don’t use it (not to mention – don’t even offer it here). If you consider 8 bit a silly container – let me remind you that in many cases one byte (8 bit) represents one character in your document. If you file has, let say, 250 bytes, it means it contains 250 characters (including white spaces and next line characters). I don’t see anything silly about it. Programmers often use hex or octets. And I do not find it silly either. And, finally, SI is not a unit as you mentioned. It’s a standard that puts in order mess created by recent misusing of the common decimal prefixes (x1000) with non decimal meanings (x1024). ‘Kilo’ as prefix is always meant x1000, nothing more and nothing less. SI standard makes it possible to use both decimal and binary prefixes without any confusion and ambiguity.

  70. Herohtar says:

    "But why does Explorer deliberately hide the sizes in Vista if nothing is selected and why does it not show free space on the details pane or status bar? Why are they empty? Was it a performance issue Microsoft couldn’t tweak? Why does Explorer hide sizes on the details pane for 15+ files  and again for 15+n files? Fix this extreme annoyances before RTM please. Why does it remove the ability for shell extension to display dynamic information in columns even if the user prefers to? Why have a status bar if all it shows is number of items selected, absolutely nothing else? Isn’t that info on the details "pain" as well? Why, oh why did Microsoft hide sizes beginning with Vista? It is for this reason that I will stay with XPMCE."

    You actually can view the file size of a group of 15+ selected items, but you have to click the "Show more details" that appears in the details pane. My guess is that they decided it was a hit to performance, especially since the Vista UI eats up so many resources already, so instead of fixing their real problem they just disabled that.

  71. Mike Annonson says:

    I have no clue who to address this reply to because the person who posted it completely omitted their own name …

    "If you don’t like your kidra, medra system – don’t use it (not to mention – don’t even offer it here)."

    I cannot fathom what you are complaining about here, but I did not put forth those prefixes; spork did, further up this thread.

    "If you consider 8 bit a silly container – let me remind you that in many cases one byte (8 bit) represents one character in your document. If you file has, let say, 250 bytes, it means it contains 250 characters (including white spaces and next line characters). I don’t see anything silly about it. Programmers often use hex or octets. And I do not find it silly either."

    I did not say that.  At all.  I said it was silly to take a decimal counting system and graft it onto an octal one, while not even maintaining decimal counting in all cases.  As for you comment about 1 byte being a single character, this is only true in certain character coding schemes.  East Asian languages use 2 bytes, as does Unicode in certain flavors.  And UTF-8 will use up to 4 bytes, depending upon what characters you are encoding.

    "And, finally, SI is not a unit as you mentioned. It’s a standard that puts in order mess created by recent misusing of the common decimal prefixes (x1000) with non decimal meanings (x1024). ‘Kilo’ as prefix is always meant x1000, nothing more and nothing less. SI standard makes it possible to use both decimal and binary prefixes without any confusion and ambiguity."

    I believe your English reading comprehension is very low.  Not once did *I* use the word "unit" in my post; that word only occurs in the quote from Ian.  You do understand what quotation marks designate, I assume?

    You are pedantically correct in that "SI is not a unit." SI is short for Le Système International d’Unités (International System of Units in English), and it defines a set of units and gives a set of handy base 10 prefixes to make life easier (because it’s faster to write 5 km than to write 5 000 m).  Thus saying "an SI unit" is perfectly correct, because the whole purpose of SI is to define units.  Even so, my only mention of SI at all was "SI style prefixes," which I assume doesn’t actually run afoul of your definition of what SI is because I’m not even talking about SI; only talking about something that mimics its naming convention.

    I’m sorry if I’m coming off as rude or condescending, but I find it rather offensive to be called out for saying something I did not even say.

  72. nocturnal says:

    While we’re on the subject, how come this stupid speedometer in my car shows m/h instead of km/h?

    Because kilomiles are too big to be useful.

  73. Timo says:

    So sayeth Worf: "does Europe really get better gas mileage? Or does the use of the larger imperial gallon skew things?"

    Dunno about Brits (who should be using SI, though EU commission keeps granting exceptions to that), but for most Europeans "gas mileage" means nothing – first of all, we use SI (well, mostly – still km/h instead of m/s) which means that gallons don’t enter into the equation anywhere, and second of all, it’s usual to give the reciprocal of gas mileage, i.e. the consumption of petrol (in units of volume) per unit of length – usually litres/100 km.

    And as to the topic – it’s deplorable that a prefix is named like an SI prefix and is almost of the same size (same magnitude, at least). KiB could have a better name, but a "silly" name is no reason not to use it. At least, please stop using k, M, G, etc as something else than a SI prefix.

Comments are closed.