Fog Creek Software
Discussion Board




kibibytes, methodologies and other stories

In an earlier thread, I was corrected about an error I made regarding kibibytes.
http://discuss.fogcreek.com/joelonsoftware/default.asp?cmd=show&ixPost=139568&ixReplies=6

To be honest, it was the first time I had heard of the word, and its kin, the gibibytes and mebibytes.

It seems the IEC introduced the word c1998, but it has not been widely adopted. People (including myself) still use the term kilobyte and its kin, gigabytes and megabytes.

"The new term “kibibyte” will more accurately describe the number of bytes in a kilobyte—rather than being 1,000, as could be inferred by the prefix “kilo,” a kilobyte actually has 1,024 (2 to the 10th power) bytes."

Contrast this with the wholesale adoption of the term methodologies. I still have issues with the way it is used, especially in techie circles. Say what you like, (and I am sure you will), a look into the etymology of this word would reveal that it is the study of methods. XP is a method, not a methodology!

"In recent years, however, methodology has been increasingly used as a pretentious substitute for method in scientific and technical contexts, as in The oil company has not yet decided on a <i>methodology</i> for restoring the beaches. .....  But the misuse of methodology obscures an important conceptual distinction between the tools of scientific investigation (properly methods) and the principles that determine how such tools are deployed and interpreted.

And how does one reconcile a words etymology with its colloquial use??

Any other new words out there that we should look out for? And what do they mean?

Tapiwa
Sunday, May 09, 2004

Forgot to reference properly
http://www.quinion.com/words/turnsofphrase/tp-kib1.htm

http://dictionary.reference.com/search?q=methodology&r=67

Tapiwa
Sunday, May 09, 2004

just thought up one..

boxen!! very popular on slashdot.

Tapiwa
Sunday, May 09, 2004

The only people who get any use out of the notion that a kilobyte is 1000 bytes instead of 1024 is drive manufacturers who are trying to justify selling you a "40GB" drive that only holds 37.25GB.  Other than as justification for this common bit of subterfuge, I've never heard the term "kibibytes" or any of its cousins used in any context.

Kyralessa
Sunday, May 09, 2004

Is there a reason you blew past the first definition of "methodology"?

"A body of practices, procedures, and rules used by those who work in a discipline or engage in an inquiry; a set of working methods: the methodology of genetic studies; a poll marred by faulty methodology."

I think "kibibyte" sounds silly and pretentious.

And I like "boxen" to indicate "multiple server boxes" because "servers" is ambiguous (and will only get worse with the growth of virtual servers)

Philo

Philo
Sunday, May 09, 2004

Actually there is a reason I ignored it. IMNSHO, it is incorrect. And hence lies the problem.

We can all throw seemingly authoritative sources who will disagree on what the work kilobyte or indeed methodology really means.

In such a crisis, do you turn to etymology (in which case methods win), or popular usage in some circles (in which case the methodology crowd wins hands down).

What about kilobytes? Technically the incorrect word was used because of the lack of a proper word. Years later, someone decides to rectify this, and most of the world does not even notice!

Do we call this person anally R, or do we knight him for his effort in correcting the misguided masses?

Tapiwa
Monday, May 10, 2004

I wouldn't mind if "kibi" wasn't such a terminally idiotic prefix. Who on earth came up with that?

Chris Nahr
Monday, May 10, 2004

I honestly promise to give up programming the day "kibibyte" becomes the established word.

Alex
Monday, May 10, 2004

A kibibyte sounds like some sort of dogfood.  I wonder if
it comes with a free chew-toy if you get the 50 lb bag?

x
Monday, May 10, 2004

It is just plain unfortunate that 2^10 and 10^3 are roughly the same and some twat in the past thought is was easier to use the same prefix for both. But in the case of Gb and Gibibytes the difference is already more than 7 %.

Jan Oosting
Monday, May 10, 2004

Dogfood. Now that is an argument.

I have dogfood for breakfast
Monday, May 10, 2004

> in the case of Gb and Gibibytes the difference is already more than 7 %.

Moving on to petabytes and whatnot, who really cares?

Technical people know what they mean; non-technical people just use 'gazillion' and are fine.

Do you know how much a parsec is? Do you care?

Alex
Monday, May 10, 2004

that rover is going all the way to mars. Who cares whether those are inches or centimetres we used in the design! In the grander scheme of things, the difference is so small anyway.....

Tapiwa
Monday, May 10, 2004

I've always maintained that where other languages have an academy English has an argument.  But what can you expect of a language that is at least 50% loan words and sports a simplified and contradictory grammar (For a start we've more irregular verbs than anyone else). 

English will proceed by custom & practice as ever.  What else from a language where the name of an Indian town can mean mad or "brassic" can mean "without cash"  (etymology available on request).

BTW A parsec is 3.26 light years and I only think it's relevant if you use parallax for measuring Estella distances.  Which I don't think they do any more.

a cynic writes...
Monday, May 10, 2004

"stella" not "Estella" - (over-enthusiastic Firefox spellchecker) - sorry.

a cynic writes...
Monday, May 10, 2004

"Actually there is a reason I ignored it. IMNSHO, it is incorrect. And hence lies the problem.

[snip]

In such a crisis, do you turn to etymology (in which case methods win), or popular usage in some circles (in which case the methodology crowd wins hands down)."

Well, looking at the modern english language, either you have to start talking *very* strangely, or you have to acknowledge that the latter is the way this language works. If you insist on only using "original" english, how come your posts don't look like this:

Hwæt! We Gardena in geardagum,
þeodcyninga, þrym gefrunon,
hu ða æþelingas ellen fremedon.
Oft Scyld Scefing sceaþena þreatum,
monegum mægþum, meodosetla ofteah,
egsode eorlas. Syððan ærest wearð
feasceaft funden, he þæs frofre gebad,
weox under wolcnum, weorðmyndum þah,
oðþæt him æghwylc þara ymbsittendra
ofer hronrade hyran scolde,
gomban gyldan.

?

;-)
Philo

Philo
Monday, May 10, 2004

The question is Philo - do you know Old English or do you just have a copy of Beowulf handy?

a cynic writes...
Monday, May 10, 2004

and then we will have kibibit ... try to repeat it 50 times ... kibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibitkibibit


Monday, May 10, 2004

On a slight aside, there was a series on UK television recently by Melvyn Bragg about the development of English. If it makes PBS (I doubt it would be shown by CBS or NBC), I'd recommend watching it.

Gaius
Monday, May 10, 2004

It's only confusing now because machines with 8 bit byte and 16/32 bit words won out over other word-size CPUs. They were'nt always built on powers of 2. There were once processors with 6 bit bytes and 12 or 36 bit words. 1K was never consistently 1000 or 1024. It was an approximation to the nearest thousand of whatever size words your machine was constructed of.

old_timer
Monday, May 10, 2004

Not only that, there have been computers that were not even binary. From that point of view, equalling kilo with 1024 is an inappropriate implementation detail. It is a leaky abstraction.

Karel Thönissen (www.hello.nl)
Monday, May 10, 2004

"The question is Philo - do you know Old English or do you just have a copy of Beowulf handy?"

Sadly, I'm not as erudite as either option would suggest. I googled it.

Oops, sorry Tapiwa - "I searched for the information at http://www.google.com/ " Didn't mean to use a word that didn't exist 500 years ago.

[g,d,r]
Philo

Philo
Monday, May 10, 2004

Tapiwa dude! You too visit Michael Quinion's website, I am glad.

Sathyaish Chakravarthy
Monday, May 10, 2004

----"parallax for measuring Estella distances.  Which I don't think they do any more.

a cynic writes...
Monday, May 10, 2004

"stella" not "Estella" - (over-enthusiastic Firefox spellchecker) - sorry. "-----

Actually the adjective is stellar. Stella is a girl's name, though a parsec would be a reasonable diistance to keep from the one I know.

Stephen Jones
Monday, May 10, 2004

---"Not only that, there have been computers that were not even binary. From that point of view, equalling kilo with 1024 is an inappropriate implementation detail. It is a leaky abstraction.

Karel Thönissen (www.hello.nl)
Monday, May 10, 2004 "------

So you debug code for Babbage's Mechanical Engine do you. I feel for you. Debugging in Ada is enough to make anybody lose their grip on reality.

Stephen Jones
Monday, May 10, 2004

Stephen - it's a fair cop.

Philo - I don't read OE either -  it was Scyld Scefing who gave it away.

a cynic writes...
Monday, May 10, 2004

When I was at university, some called the plural of Vax, Vaxen. I think it was derived from the plural of ox.

MilesArcher
Monday, May 10, 2004

As I stated in the other thread dealing with the subject: we actually find bugs by using a very precise vocabulary, including the distinction between kilo and kibi. It really works, and that is the only convincing argument.

Wrt the analystical machine, that was by far not the last machine that was not (pure) binary. I programmed non-pure binary machine in the eighties.

I do not understand the statement about Ada and debugging. We do not work in Ada, and debugging is almost unknown in the Ada-community for good reasons. Care to explain?

Karel Thönissen (www.hello.nl)
Monday, May 10, 2004

Find out who Ada is named after.

Stephen Jones
Monday, May 10, 2004

Karel is hammering away so hard at this issue, I begin to suspect that he works for a hard-drive manufacturer.

Kyralessa
Monday, May 10, 2004

Stephen: Are you suggesting that I am programming in Ada Lovelace? Is this an obscene suggestion, or what?

Kyralessa: one little check would have sufficed to find the answer.

Karel Thönissen (www.hello.nl)
Monday, May 10, 2004

Aha! But are you programming in Ada Lovelace digitally?

MilesArcher
Monday, May 10, 2004

(-8

Karel Thönissen (www.hello.nl)
Monday, May 10, 2004

How long will it take me to transfer 10 Gigabytes of data over a gigabit serial link?

If I can read one byte of memory for every bus cycle, and my bus runs at 100 Mhz, how long will it take to read a gigabyte of memory?

Ham Fisted
Tuesday, May 11, 2004

----"Aha! But are you programming in Ada Lovelace digitally? "----

You mean has he got his digits in her?

Stephen Jones
Tuesday, May 11, 2004

>> how long will it take to read a gigabyte of memory?

Ah, that's what *we* are here for.

Please make an appointment.

Alex
Wednesday, May 12, 2004

*  Recent Topics

*  Fog Creek Home