Talk:Megabyte

Discussion about centralization took place at Talk:Binary prefix.

Establish a Convention? (April 2002)

We need to establish a convention for Wikipedia about the word "megabyte" because of the conflicting definitions, or else we cannot ever use the word. Should we go with SI, i.e. 10⁶ bytes, and then use the new MiB for 2²⁰ bytes? AxelBoldt, Friday, April 26, 2002

If we accept this convention for the word "megabyte", we then should do the same for "kilobyte", i.e. 10³ bytes and kiB for 2¹⁰ bytes, too. I know children look rather confused when I explain to them that km is a 1000 m, while kB is 1024 bytes, because computers work better with binary numbers... Maria Renee Jenkins, Sunday, April 28, 2002

It is our task to follow standards, not ignore them. The plain reality is, majority usage for MB is the binary MB, and I ain't talking about a small majority either. The fact that "there are 1024 somethingorothers in a KB" is one of the very few things that Joe Average computer user really does understand. And where did this Wikipedia "policy" come from all of a sudden? It's a really bad idea. Tannin 13:18, 19 Sep 2003 (UTC)

I picked the "policy" since after my question above, nobody expressed an opinion either way for quite some time. The policy does not ignore standards, but follows them. I agree with your assessment of majority usage, but I don't think it is a large majority: the average computer user encounters megabytes in four contexts:

• hard drive capacity
• memory size
• file size
• bandwidth

Decimal megabytes are used in two of the four contexts. If you prefer we pick the opposite policy, then you need to invent a name for 10⁶ bytes, and you need to make the case why Wikipedia should ignore international standards in favor of this new name. AxelBoldt 17:14, 21 Sep 2003 (UTC)

These are all good reasons to not take sides by pretending that one of the two major meanings of the word doesn't exist or is wrong. Rather, when the difference matters, be explicit about the usage of the term. --Brion 17:49, 21 Sep 2003 (UTC)

I don't think that the sentence "This is the definition used in Wikipedia" pretends that the other meaning doesn't exist or is wrong; it just fixes a convention to avoid having to change many links from "megabyte" to "megabyte (i.e. 10⁶ bytes)" or "megabyte (i.e. 2²⁰ bytes)". But I can live with that as well. AxelBoldt 09:29, 22 Sep 2003 (UTC)

Proclaiming one to be the only definition used here would seem to require tracking down all uses of the term, deciding for sure which meaning was meant (possibly requiring further research), and either leaving it or changing it to "mebibyte". Letting normal usage stand allows further clarification where details are known and relevant. --Brion 09:38, 22 Sep 2003 (UTC)

A vote has been started on whether Wikipedia should use these prefixes all the time, only in highly technical contexts, or never. - Omegatron 14:50, July 12, 2005 (UTC)

Binary million -vs- decimal million (Dec 2004)

After some double-checking and extensive web searching, it appears that the overwhelming majority consider a MegaByte to be a binary million, i.e. 10^20. As far as I can see, only hard drive and floppy drive manufacturers use a decimal million (10^6), primarily for marketing/misleading reasons. example. I've worked in IT for 10 years, and I've never even heard of a Mebibyte. Regards, --Rebroad 19:58, 17 Dec 2004 (UTC)

Just realised that this article has been misleading since May 2003 (see) when the more commonly used usage was moved to 2nd in the list. --Rebroad 20:06, 17 Dec 2004 (UTC)

This was pretty clearly not NPOV, reproducing the "it's a conspiracy by hard drive manufacturers" line. I've added some unambiguous cases--for instance, the CD standard uses base 2, whereas the DVD standards use base 10. I've also tried to make it clear that hard drive manufacturers almost always use base 10, while OS software and humans mostly use base 2. As a final note, I've added a comment about MHz to go along with the discussion of Mb, making it clear that although people measure RAM capacity in base-2, they measure RAM speed in base 10. Now, can we all switch to SI and stop being needlessly confusing? Please? Metamatic 21:05, 23 Feb 2005 (UTC)

Nooo! We must adamantly resist all attempts to correct that which we have grown comfortable with!  :-)

“Megabytes have always been base 2, and always been written as MB”, they sneer. “Everyone knows that 1MB is 1024KB, unless you’re talking about DVDs, or reading manufacturer specs for a hard drive, and that’s just the hard drive manufacturers being stupid. Everyone knows that ‘K’ on a computer means 1024; except for speeds, where it means 1000, except for file download speeds where it means 1024, except when it’s the speed of your modem, when it’s 1000. Everyone knows that. What, are you stupid?” [1] - Omegatron 15:33, Jun 17, 2005 (UTC)

News flash -- majority rules.

Number of users of Microsoft Windows (where MB= 1,024 KB and KB= 1,024 bytes) and similar computing environments: in the millions (or billions). Number of crusaders advocating "kibibytes", etc.: in the thousands (maybe). Number of people who actually consistently refer to binary byte sizes as a "Mebibyte": maybe a couple hundred?

I refer you to earlier comments; for example:

• "After some double-checking and extensive web searching, it appears that the overwhelming majority consider a MegaByte to be a binary million [1,024 KB]...";
• "not yet accepted and are simply ignored by ...This is the reality. Hardly anyone uses mebis. " ; etc.

- Liberty 23:34, 25 September 2005 (UTC)

Megabytes in Use" Useful? (June 2005)

I don't think the "Megabytes in Use" section is very useful, particularly the part about a megaybte storing roughly one book, 100 small images, or 1 minute of audio. Given that books come in all sizes, designs, and point sizes, this comparison is virtually useless. 100 small images? How small? JPEG? GIF? PNG? TIFF? 1 minute of audio as mp3 at 128kbit is indeed roughly 1 MB. As a WAV file, it's 10 MB however. I think this part should either be removed or replaced with a more clearly defined standard of measurement. pogo, June 2005

It should either be presented as "very roughly" or else made more specific, like saying 128kbit/s as you said. - Omegatron 02:45, Jun 20, 2005 (UTC)

Regarding "Adoption by the NIST" (Oct 2005)

A definition of the IEC proposal on the NIST does not, in itself, constitute advocacy of adoption, nor does it necessarily represent an official endorsement of the standard.

I refer you to the actual text appearing on the single page at the NIST that touches on the topic [2] (which, incidently, has an unclear authorship date). For example (emphasis mine):

"... the IEEE Standards Board decided that IEEE standards will use the conventional, internationally adopted, definitions of the SI prefixes. Mega will mean 1 000 000, except that the base-two definition may be used (if such usage is explicitly pointed out on a case-by-case basis) until such time that prefixes for binary multiples are adopted by an appropriate standards body."

I'd be interested in any references to actual, recent, implemented use of the newer terms in NIST activities... Just as a point of reference, note also that NIST is a non-regulatory federal agency... - Liberty 04:03, 13 October 2005 (UTC)

• IEEE
  • Standard: IEEE 1541-2002, IEEE Standard for Prefixes for Binary Multiples
    • "1541-2002 (SCC14) IEEE Trial-Use Standard for Prefixes for Binary Multiples [No negative comments received during trial-use period, which is now complete; Sponsor requests elevation of status to full-use.] Recommendation: Elevate status of standard from trial-use to full-use. Editorial staff will be notified to implement the necessary changes. The standard will be due for a maintenance action in 2007." IEEE-SA STANDARDS BOARD STANDARDS REVIEW COMMITTEE (RevCom) MEETING AGENDA 19 March 2005
    • "1541-2002 IEEE Standard for Prefixes for Binary Multiples (Upgraded to full use from trial use)" [3]
• NIST
  • "The IEC has adopted prefixes for binary multiples in International Standard IEC 60027-2, Second edition, 2000-11, Letter symbols to be used in electrical technology—Part 2: Telecommunications and electronics. ... Although these prefixes are not part of the SI, they should be used in the field of information technology to avoid the incorrect usage of the SI prefixes." NIST Special Publication 330 2001 Edition The International System of Units (SI)
  • "Because the SI prefixes strictly represent powers of 10, they should not be used to represent powers of 2. Thus, one kilobit, or 1 kbit, is 1000 bit and not 2¹⁰ bit = 1024 bit. To alleviate this ambiguity, prefixes for binary multiples have been adopted by the International Electrotechnical Commission (IEC) for use in information technology." nist.gov
  • "The new prefixes will eliminate the present confusion between powers of 1000 and powers of 1024 since in the field of information technology the SI prefix names and symbols for decimal multiples are now often used to represent binary multiples." News briefs Section 1.9
  • "With significant input from the National Institute of Standards and Technology, the IEC adopted kibi (Ki), mebi (Mi), gibi (Gi), tebi (Ti), pebi (Pi) and exbi (Ei) to represent exponentially increasing binary multiples. A kibibyte, therefore, equals 2 to the 10th power, or 1,024 bytes. Likewise a mebibyte equals 2 to the 20th power, or 1,048,576 bytes. The new prefixes for binary multiples, which parallel the metric prefixes, will increase precision in expressing electronic information." Representative's Report - April 1999

Sounds like an endorsement to me... — Omegatron 04:48, 13 October 2005 (UTC)

The facts on the ground are that the largest OSes (Microsoft's, POSIX (Unix, BSD, Linux, OS X...) use the binary system, and this it the computer arena where people most often come in contact with these metrics. At some (most?) CPUs run at binary-based speeds, e.g. the first 65c02 was 1.024 MHz, as I recall. Changed the page to reflect that.

That's a decimal speed. A binary speed would be something like "1 MHz" = 1.024 MHz. — Omegatron 23:37, 26 February 2006 (UTC)

The claim that "Linux uses binary" is false in its generality. The kernel boot messages refer to disk sizes in decimal units, fdisk uses decimal units, etc.

Why is there a link to a picture of flies having sex?

Usage objection

Someone inserted and someone reverted:

"Because of a traditional inconsistency, "megabytes" are often intended to mean mebibytes in common speech. This usage is not recommended as it creates confusion (see below) and has been facing increasing opposition by many technical standards and legal entities in the past few years."

Let's discuss this before putting it back in production. I'll state my objects to the addition first:

1. I think the entire addition is worded and structured poorly.

a. "Because of a traditional inconsistency" seems very vague and imprecise. Maybe it should be, "Because it has meant different things to different industries"

b. ""megabytes" are often intended to mean mebibytes in common speech. again seems vague and also seems to be a conclusion that should have a citation. I would suggest something like, ""megabyte" can mean either "megabyte" or "mebibyte" when used in common speech and even technical specifications."

c. "This usage is not recommended". What is "this usage"? Does this mean the term "megabyte" should not be used? Or does this mean "megabyte" should not be used when the speaker means, "mebibyte"?

d. "has been facing increasing opposition by many technical standards and legal entities". I think "increasing opposition" is a little extreme. The SI definition and the "mebi" prefix for the binary calculation are gaining support in various technical and international standards, but opposition implies something more.

2. Some parts of the addition seem speculative or make conclusions.

a. ""megabytes" are often intended to mean mebibytes in common speech.". Where is the evidence that this is the case? And what is "common speech"? Common to computer technicians, photo-journalists, and stamp collectors talking about system RAM, the size of the DVD rip they pulled of the web, or how much disk space they have left on their iPod?

b. "This usage is not recommended". Not recommended by whom? NIST, IBM, NASA, GAO, EU, UN? Certainly an encyclodedia isn't making such a recommendation.

The entire addition seems to be an objection to the use of megabyte with supporting argument. It's taking a position and promoting a particular POV. It might be a common POV, but "common" isn't necessarily "neutral". I vote to strike the edit. --JJLatWiki 16:57, 10 January 2007 (UTC)

I think this is the best version:

The usage of "megabytes" is ambiguous, as they can mean either 1000 (the technically accurate definition) or 1024 submultiples (a "mebibyte"). The confusion originated as compromise technical jargon for the byte multiples, that needed to be expressed by the powers of 2 but lacked convenient naming. As 1024 (2^10) is roughly equal to 1000 (10^3), rougly corresponding SI multiples began to be used as approximate binary multiples. However, in the past few years a number of technical standards and legal entities (IEC, IEEE, EU, etc.) have addressed this ambiguity and discouraged the use of "megabyte" as a synonym for a "mebibyte" by promoting the "mebibyte" (MiB) instead. This shift is reflected in an increasing number of software projects, but most file managers still show file sizes as "megabytes" ("MB").

If you don't like it, please don't just delete it, but improve it.

72.36.245.34 07:20, 11 January 2007 (UTC)

POV

I've again removed the sentence "So, technically, it is the operating system that creates the confusion, rather than the hard disk vendor." This sentence implies that the hard disk vendor is more correct than the operating system vendor in their choice of units, which cannot be established without the implied POV that the SI units are better than binary units. JulesH 14:45, 5 February 2007 (UTC)

I weakly disagree. "better" would imply POV, but "technically" is neutral. Based on the current standard of "Mebi", "Gibi", etc, and the fact that kilo and mega meant 1000 and 1000000 long before computer programmers inappropriately used them to describe 1024 and 1024^2, I would have to argue that the sentence is technically correct. I would also argue that the hard disk vendors ARE more correct. Mega meant 1000000 long before it ever meant 1024^2, and rather than create an appropriate abbreviation for 1024^2, programmers decided to use mega for almost (but not quite) everything. So drive vendors have at least a couple hundred years of history to support there usage and even some scattered support within modern non-disk related, computing history. OS vendors have a few decades of spotty support for their use of mega, which is entirely (that I know of) isolated to their niche. I don't like the way the sentence sounds, but technically it is correct. --JJLatWiki 21:08, 5 February 2007 (UTC)

I think the replacement sentence you've put in works a lot better. Thanks. JulesH 23:25, 7 February 2007 (UTC)

Time ambiguity

"In the past few years, standards and government authorities including IEC, IEEE, EU, and NIST, have addressed this ambiguity by promoting the use of megabyte to describe strictly 1000² bytes and "mebibyte" to describe 1024² bytes."

This is unclear as to when "the past few years" is. 20 years down the line, it certainly won't be true. Does anyone know when IEC, IEEE, EU, and NIST began to promote the use of these terms? If so, it should be added in the article. Meviin 01:31, 28 May 2007 (UTC)

from the 1999 for the IEC (I believe that the 'full use' recommendation of the IEC was in actually in 2005) to somewhere at the end of this year for the EU... (I think that a European Directive was voted on recently (few weeks), but it's official publication is planed for Q2 2007. -- Shmget 02:28, 28 May 2007 (UTC)

Mega = 1024

I'm very disappointed to find something so stupid on Wikipedia; every child in computer science learns that 1 Megabyte = 1024 Kilobytes, it's widely accepted in the computer science community to count in 1024 (2^10), Microsoft Windows & Linux use it to show storage capacities. So where does this 1000 come from ? stupid, stupid, stupid. Maybe it's better to throw wikipedia to the garbage can and buy a real encyclopedia.

~~Marc —Preceding unsigned comment added by 62.197.65.107 (talk) 18:12, 8 December 2007 (UTC)

Check the real history. MB in floppy disk capacities really were 1 024 000 bytes. — Arthur Rubin | (talk) 19:31, 8 December 2007 (UTC)

... and, whether you like it or not, in hard disk drives, 1 GB is 1 000 000 000 B. The computer industry has created the confusion. Wikipedia is simply explaining it as best it can. Thunderbird2 (talk) 09:44, 9 December 2007 (UTC)

The way I see it there isn't much confusion, either you know why there are two definitions and apply them according to the circumstances... Or you have no idea so it either a. shouldn't matter, or b. you learn what the situation is. People who bitch and moan that hard drive manufacturers are ripping them off are using phony naivety to get all worked up. — Preceding unsigned comment added by 124.168.198.102 (talk)

Three definitions, actually: 1000×1000 (SI compatible) 1024×1024 (RAM/CPU compatible) and 1024×1000 (hybrid, used by some FDD/HDD manufacturers). SHEFFIELDSTEEL^TALK 16:32, 10 November 2008 (UTC)

1024×1024 is "RAM/CPU compatible"? It is more than that: it is both historically and currently the predominant usage. Edam (talk) 15:34, 13 January 2009 (UTC)

We have the SI, which is very well established. As a standards committee we have ISO, which is THE authority on units and their naming. I think it's a BAD idea to settle these things on the web, by voting, where every TD&H can yell something. ISO cannot define kilo to be anything else than 1000 anymore, so in support of their inevitable decision:
1: Kilo = 1000 was first. Us computer guys /gals misused it FIRST, to mean 1024. So let's not keep investing in a lost cause, admit the error and correct it as soon as we can.
2: Users, in the age of terabyte disks and gigabyte memory, don't really care if their memory is a few percent larger or smaller than stated.
3: We can use mega=1000000 and leave mebibi etc for specialists and spec sheets.
4: What is the problem of printing "1.024 kB" ?
Erik6000 (talk) 08:34, 17 November 2008 (UTC)

Well, for starters, if someone asks me how much memory is available for dynamic instances of object type X, and I answer two megabytes, then they will go away thinking that they can create 2048 dynamic objects each using 1024 bytes of memory... and they'll be correct. If you were responsible for setting the size of that memory pool and you set it to 2,000,000 bytes and answered the same question with the same answer, your application would fail at some point in the future, and it would be your fault.

Wikipedia isn't the venue for us computer gals/guys to correct things. It's the venue for us editors to document things as they are in the real world. Things like, for example, that this unit is ambiguous and has had three different meanings. SHEFFIELDSTEEL^TALK 14:27, 17 November 2008 (UTC)

WP:BOLD edits

Kbrose has been making bold revisions to the article without discussion, and I've reverted twice. The essay was supposed to be WP:BRD, not WP:BRRRRRD. — Arthur Rubin (talk) 19:10, 16 November 2008 (UTC)

I don't see any major differences between the version as of 2008-11-13 and the current version -- it seems to me that the "bold" (and inaccurate) changes were actually introduced by Likebox without discussion. Shreevatsa (talk) 19:16, 16 November 2008 (UTC)

It looked to me that Likebox just reordered the discussion. However, I'll concede that, provide that we add a note that the stadards bodies' efforts are apparently unsuccessful. — Arthur Rubin (talk) 19:36, 16 November 2008 (UTC)

This also is factually wrong. They are not unsuccessful. Naturally these standardization efforts take time, people don't like change. In other technical fields, the prefix Mega- means exactly what it should. In telecom, a T1 line always has transmitted 1.544 Megabits/s, that is 1,544,000 bits/s. It is about time, that a technical discipline like IT, adheres to existing standards and ends the confusion it has created. Kbrose (talk) 19:44, 16 November 2008 (UTC)

Perhaps you're right that it should be happening. However, it isn't happening. — Arthur Rubin (talk) 20:15, 16 November 2008 (UTC)

If it weren't happening, we wouldn't have these exchanges. Facts remain, that these definitions are in place, are being adopted increasingly in documentation, software, and only people who don't like them argue about it. WP should reflect on technical issues what the standards are, and also mention the history and the struggles of transition. I am sure there are better ways to present that, but the most recent changes were steps by refuseniks in a backward direction. Kbrose (talk) 20:26, 16 November 2008 (UTC)

This is false, and you should know it, per the Wikipedia debate as to whether these should be included in Wikipedia articles. Wikipedia was the most common use outside of the standards publications themselves, before the decision was made not to use them for the very reason that no one else was. — Arthur Rubin (talk) 22:19, 16 November 2008 (UTC)

Furthermore, it doesn't seem to be the dispute should be in the lead. But that's a matter of opinion. What's not a matter of opinion is that Wikipedia, itself, was the most common use of the term. — Arthur Rubin (talk) 22:24, 16 November 2008 (UTC)

As for previous discussions, the most recent seems to be at Wikipedia talk:Manual of Style (dates and numbers)/Archive/Complete rewrite of Units of Measurements (June 2008)#IEC Prefixes (Purplebox), it which it was stated that it was used in scientific publications, but not in "the real world". However, it doesn't seem to be being used, in reality, in scientific publications. — Arthur Rubin (talk) 23:00, 16 November 2008 (UTC)

I reverted back to the pre-edit-war / pre-bold-edit version. Per WP:BRD, we've done the bold, revert bit, the next step is to discuss. Per WP:EW, the edit warrring needs to stop. Whichever way you look at it, proposed changes should be discussed here, and of course they should be verifiably supported by reliable sources. SHEFFIELDSTEEL^TALK 14:45, 17 November 2008 (UTC)

Merge in Mebibyte ?

Any objections ? Megapixie (talk) 09:49, 15 December 2008 (UTC)

Opposed. As long as all the multiples of the bit and byte units have their own article, this should obviously not be merged. However, I am for the DELETION (conversion to redirects) of all articles that discuss the multiples separately and merge them ALL into one article discussing the unit, as they all just discuss the same topic, a multiple of their respective unit. The customary or historic usages of these multiples in various fields can still be discussed in paragraphs. There is a lot of duplication in these articles regarding the binary/decimal controversy. Kbrose (talk) 16:49, 15 December 2008 (UTC)

Seems like a reasonable approach to me. Step one might be the creation of a draft article in userspace... any other thoughts ? Megapixie (talk) 08:31, 24 December 2008 (UTC)

Opposed. Megabyte is the contentious term. When I search for Mebibyte, I want to know about the Mebibyte, not the Megabyte (even if it is sometimes a synonym, it's not always and it would simply lead to more of the confusion the term is meant to eliminate). BossAnders (talk) 02:42, 27 February 2009 (UTC)

Strong Oppose 1.) Megabyte is the more common term, and 2.) they cover two completely different aspects. Since these merge templates have become stale, I'll be removing them shortly unless there is further discussion. — Ched ~ ^(yes?)/_© 08:02, 31 March 2009 (UTC)

Oppose merge, support removal of template, per Ched. SHEFFIELDSTEEL^TALK 12:46, 31 March 2009 (UTC)

Abstract Knowledge vs Useful, Accessible Knowledge

- I browsed to this page hoping to quickly clarify a simple and straightforward problem and instead stumbled across this wearisome, abstract from someone's Master thesis. I became mired therein. This page is uselessly tortuous. Please remember that such a commonplace subject as this is likely to draw many people for perfectly mundane reasons. Please modify the article so that it responds to simple, superficial inquiries (suh as my own) as well as to more indepth inquiries. Thank you. (Have not logged on in a few years)—Preceding unsigned comment added by 71.129.108.112 (talk) 20:46, 31 December 2008 (UTC)

Are you sure you are on the talk page of the right article? This has one paragraph on what a megabyte is, and a section on its three (unfortunately) different definitions, followed by a list of examples. It's a pretty short article. What would you like to see in it? Shreevatsa (talk) 21:17, 31 December 2008 (UTC)

"Mebibyte" is rediculous

Anyone who has worked in the computer and networking industries long enough should not have much of a problem understanding which means which. Bandwidth is in metric bits, raw HDD capacity is in metric bytes, and memory is in binary bytes. If I tell people (who are confused) how to differentiate I advise them to use the following:

   "m" = 1000  (smaller sum), metric
   "M" = 1024  (larger sum), binary
   "b" = bit   (usually 1/8th of a byte)
   "B" = byte

If this makes sense to anyone, bandwidth would be in "mb", raw HDD capacity "mB", and memory "MB". It makes more sense than making up what sounds like computer "baby talk".

—Preceding unsigned comment added by Wdl71 (talk • contribs) 16:34, 12 February 2009 (UTC)

The SI (metric) symbol for 1000 is a capital M, not m. Nicer is Knuth's proposal to use "MMB", "GGB" etc. for the larger (binary) units. Anyway, this article (and Wikipedia in general) is merely to report what is (and the fact is that "MiB" and so on are standards now), not to make our own suggestions. Shreevatsa (talk) 18:32, 12 February 2009 (UTC)

I doubt that this will ever change/be sorted out. Everyone in the IT, except for hard disk manufacturers, uses Megabyte/Gigabyte in the classic sense - and will keep doing so because it simply doesn't make any sense to change it. Hard disk manufacturers will keep selling their stuff with the artificial definition because it makes their drives look bigger. I don't see any way how this could be changed/solved. The so called "standard" is no standard because the majority blatantly ignores it and goes for the definition that has been de-facto in place for tenths of years longer. I actually find the Wikipedia article pretty good because it clearly points out the difference between artificial definition and reality. Lodan (talk) 05:24, 1 June 2009 (UTC)

Megabyte equals two values?

It can contain a value of 1,000,000 or 1,048,572. How can something equal two values? A Megabyte equals 1048572. Do not use the back of a HDD box which contains *advertising* as a way to prove it equals something less. Otherwise I would believe 'coke is it'. —Preceding unsigned comment added by 203.134.124.36 (talk) 22:54, 24 February 2009 (UTC)

To repeat what Shreevatsa wrote above, "[T]his article (and Wikipedia in general) is merely to report what is". The term "Megabyte" is used to mean two different values. You don't like this; I don't like this. Tough, because that is how it is used, and so that is what the article must contain. HairyWombat (talk) 18:38, 23 May 2009 (UTC)