how-to block ads
|
|
Uniqs:
4984 |
Share Topic
 |
 |
|
 tcp1Premium
join:2000-04-17
Herndon, VA
 ·Verizon FiOS
| reply to Jeremy341
Re: Windows OS is using wrong measurements said by Jeremy341:
said by OZO:
May be some little standard (as you mention about International Standard) is better then nothing. Don't you agree?
Nothing? A standard is in place, and everyone uses it exclusively, except the hard drive marketers. It's hardly nothing, except to zealot trolls like yourself.
Hey Jerw134.. We've bickered before but this is the first time I agree with you 110%! | |
| reply to tcp1
Re: Computers do not use 2^10 base said by tcp1:
Computers, in hardware, only use one "data type". One. That is the bit. Nothing else.
If you can point out a bit-slice or other form of CPU that allows operations on data down to bit-wise granularity of the size of operands, in commercial use today, then I would agree with that. Otherwise, you are simply being pendantic, and also wrong. Modern CPUs don't work on individual bits, they work on groupings of bits, called "operand sizes".
Modern x86 CPUs use three primary sizes, byte, word, and dword. There are also some specialized opcodes that operate on qwords. A high-level language can make it seem, to your program's source, that you are working on data types of arbitrary bit lengths, but it's all an illusion. The emitted machine code will only operate on the native hardware-supported data-types of the CPU.
said by tcp1:
What you are talking about is implementations and representations of human based systems in programming languages, as compilers interpret them to convert them to machine language - which again, is always, ALWAYS, only using bits.
As I explained above, quite wrong.
said by tcp1:
"Data Types" in programming languages are not how the "computer makes calculations". Go take an assembly language class. You'll learn a lot.
Perhaps you should as well.
(remaining irrelevant drivel snipped)
You seem enamoured at repeatedly pointing out that computers find it easiest to address memory using power-of-two sizes. Yes, we get the point. We already knew that. But that's wholly irrelevant.
The issue is why on earth did the early computer researchers usurp a previously-defined, well-understood, term of scientific measurement, and grossly redefine it to represent a number, that is "close to" its real value and meaning. It was simply sloppy. You may consider that "standard" now, because you grew up with it, but that makes its use no less wrong. The "metric system" is inherently decimal, and has been in use for far far longer than this brief recent "computer revolution" thing. | | |
|
| reply to tcp1
Re: Windows OS is using wrong measurements said by tcp1:
And the proper units are in powers of two. Using base 10 measurements in computers MAKES NO SENSE. Addressing in memory or storage in base 10 MAKES NO SENSE AT ALL.
Using YOUR system is dumbing things down to a more stupid level because people refuse to understand even the slightest, simplest details of how a computer works.
PS. CPU clock speeds are in decimal units as well, and always have been.
This is absolutely not true, first of all. The 8088 ran at 4.77MHz. Your 33Mhz 486 actually ran at 33.333Mhz, and your 266 MHz DDR ram runs at 266.666 MHz.
If you honestly don't understand the difference between non-integral, decimal numbers, and non-decimal numbers, then I truely pity you. If you have taken college CS courses (which I would assume include some mathematics courses), then I suggest you request a refund. If you paid by credit-card they may be able to help with that.
said by tcp1:
Oh, and look at those numbers! For decades, the progression of cpu speeds were as such: 16, 33, 66, 133, 166, 200, 266, 333.. Only around then did we start seeing "500" and "600" being more common; but this was simply due to the fact that the original basic 66Mhz Pentium bus was increased by 33 to 100Mhz, and CPUs worked off the multiplier of the front side bus.
See! Reasons! Not just to make it easy for you. CPU speeds, as you erroneously stated, have almost never been in even decimal units.
Learn a little bit and you wouldn't be making this argument.
You're right, of course. Those numbers like "500Mhz" aren't decimal numbers, no way could they be...
Never argue with the management, especially when the monkeys are running the zoo. | |
1 edit | reply to Hayward
Nevermind said\explained already. | |
tcp1Premium
join:2000-04-17
Herndon, VA Reviews:
·Verizon FiOS
3 edits | reply to VirtualLarry
1133, not 1130! |
Um, Larry? What numbers can't be represented in decimal? What exactly are you talking about? You said "CPU clock speeds are in decimal units as well.." Such an argument would also apply to the number 1024 as well, would it not?
If you read this whole thread, the concept is that the OP believed storage space and memory should be measured in metric-like even decimal units, such as kilo=1000. You said that "clock speeds are measured in decimal units." I think I pointed out, pretty clearly above, that clock speeds are NOT arbitrarily set at multiples of 10 like in the metric system, unless you're talking about the term "megahertz", which has nothing to do with computers or binary, and has been around long before computers have been.
My point is that clock speed is not arbitrarily set like the OP suggested; or else we'd always have multiples of 100 or 1000 to make it "easy"; it depends on the REASON for that number being what it is.
If your argument was simply 1 MHz = 1000 Hz, then I'm sorry I wasn't looking at it that simply. Perhaps it escaped me that you were talking about the derivation of megahertz, since CPU speed hasn't been measured in its base unit of hertz for ages. I concede that. It was 4am, you were right, I was wrong, but I was trying to say that when Intel or AMD choose a clock speed for a CPU, we don't see them putting out an "even" clock speed just to make it fit an exact multiple of say, two gigahertz or exactly three gigahertz.
Up until very recently, it was common to see 1133mhz CPUs or some other multiple (remember the 1.13ghz pentium III?) -- the speed is not based on the multiple of the unit; that is, it is based on a multiple of an operation parameter -- and that was my point.... Also noting that a "133 MHz" cpu was not actually running at 133,000,000 Hz, but 133,333,333 Hz -- so while 1 megahertz does technically equal 1,000,000 Hz, the pentium III 1.13 didn't run at 1130 MHz, but 1133. In this measurement system, the "error" is simply *ignored*:
And think of it this way. If some property of radio waves, for whatever reason, say on some other planet in some other dimension, resulted in radio waves always forming at a frequency multiple of say, 16 - wouldn't it make sense to measure those waves in that base, other than converting to base 10 simply so it would be easier for the layman to understand?
I know we're beating a dead horse, but you talk about early researchers.. It made absolutely no sense for early researchers to describe measurements in base 10. They found they were always deriving values such as 1024, 2048, 4096, 8192, 16384 -- and chose a unit that was a factor of those values, so they could describe values as 1, 2, 4, 8, and 16 "mega"bytes, vs 1.024, 2.048, 4.096, 8.192 and 16.384 decimal "mega"bytes. | |
tcp1Premium
join:2000-04-17
Herndon, VA Reviews:
·Verizon FiOS
1 edit | reply to VirtualLarry
Re: Computers do not use 2^10 base If you can point out a bit-slice or other form of CPU that allows operations on data down to bit-wise granularity of the size of operands, in commercial use today, then I would agree with that. Otherwise, you are simply being pendantic, and also wrong. Modern CPUs don't work on individual bits, they work on groupings of bits, called "operand sizes".
And a byte, word, dword, and qword are all made up of bits, are they not? And they're not 3 bits, or 7 bits, or 15 bits, are they? Nor are there operand sizes of 10, 100, or 1000 bits. If you measured in "metric" units, cpus would always be working on fractions of units, wouldn't they?
You seem enamoured at repeatedly pointing out that computers find it easiest to address memory using power-of-two sizes. Yes, we get the point. We already knew that. But that's wholly irrelevant.
No, not easiest. That's the only way they can do it. And that is precisely why storage space, as the original poster discovered, is measured as such - and it is not a Microsoft conspiracy.
The issue is why on earth did the early computer researchers usurp a previously-defined, well-understood, term of scientific measurement, and grossly redefine it to represent a number, that is "close to" its real value and meaning.
Is representing the number 15 as "F" in hexidecimal usurping the latin alphabet?
Why should something technical be measured in a way that makes it easier for a layman to understand, versus a way that is easier to work with for researchers and technicians? Just for history's sake? Just to appease the SI perfectionists?
Why must computers, which inherently work in base-2, be forced to be described in base-10 when describing a technical aspect of its operation; it's not like we're looking at Quicken here and it's saying your $1000 mortgage payment due is $1024. We're talking about an actual representation of an operational parameter, which makes much more sense to be described in a power of two. | |
| reply to OZO
Re: Windows OS is using wrong measurements OH MY god not this old discussion again....I thought this got buried years ago...get over it! | |
scomps
join:2001-06-05
Utica, NY | reply to koolman2
said by koolman2:
Why don't you try programing a computer using the decimal system- I think you'd be pleasantly surprised that you can't easily integrate operations in a computer using that system, hence why everything is done in binary, INCLUDING storage.
If he tried programming or even operating a computer in the decimal system, he'd be calculating the absolute pressure on the power switch to get it to come on, depth, pressure and speed of keystrokes for various characters, position, shade, lumens of video pixels etc and so on. (yes I know this is routinely done for games and graphics applications)
The point I'm making is that computers are a logical creature of True and False. On or Off. That's base 2. We "live" with it because it makes the most amount of sense for the technology, not necessarily the most amount of sense to the half clueless wanna be smarter than they really ares.
--
Scott Johnson -- developer of MWall. Contact me for more information. | |
scomps
join:2001-06-05
Utica, NY | reply to OZO
when did wikipedia become an authority on anything? and when in hell did a byte not equal 8 bits, a K not equal 1024 bytes, and so on? This is the way it is because it works, and the system can handle it. Just because some hairbrained simpleton can't handle that the k in the metric system is based on 10's and in the computer industry a k is based on the closest representation technically possible to those metric k's--well that's not the problem of the computer industry. That's the problem of those that can't handle the extra processing cycle in their own minds.
--
Scott Johnson -- developer of MWall. Contact me for more information. | |
 koolman2Premium
join:2002-10-01
Anchorage, AK | reply to scomps
That was my point- thank you for making it more understandable. | |
 MrFixitCTpay it forwardPremium,VIP,ExMod 2001-06
join:2000-12-01
Charleston, SC | reply to OZO
Now that this subject has been thoroughly beaten to death, time to say good night Gracie..
thank you all for playing  | |
|
