If you want a civil discussion than be civil in your own post in the first place. Not very polite to use a sentence like "Please feel free to look up "Hard Disk Drive" on wikipedia for a decent explanation on the technology and how it actually stores data.",
If that came off the wrong way, please pardon me. My intent was that I was too lazy to type out the details of how/why HDD's store data, and figured if someone really cared about that level of detail, they could reference that link. I had actually considered explaining that magnetic medium stored by directional magnetic charges, etc,etc... but figured that most casual readers would get annoyed by the extreme details needed for that to make much sense.
Yep, but not all of them and not all of the time. I wanted to make the point that it's not just a marketing decision of the disk companies out there.
You know, since I started reading this thread, I've been going nuts trying to remember how the old reel-to-reel tape systems used to advertise storage capacities, and can't seem to remember.
I can't remember the actual size of the first "hard drive" I used either, so I decided to see what wikipedia had on it, and found this:
http://en.wikipedia.org/wiki/Early_IBM_disk_storage
I'm not using that article as any type of reference, btw... I just found it really interesting reading. From what I've been able to tell, the 350 was the first "hard drive" system... and only stored 7bit characters. Actually, several of the first drives listed only stored "characters" or "digits", but not "bytes." The first one I see there in that list that stores "bytes" is the 2311. I have no idea how it was marketed, though, so don't know if they claimed 7,250,000 bytes, 7.25 MB, or 7MB (7MiB.)
Anyway, back to the actual thread...
I _do_ remember that floppy disk sizes were abbreviated as base-2 (1 KB == 1024 bytes.) I can't speak directly to the original marketing of HDD's. I seem to remember the original IBM XT having a 10MB 5 1/4" full size HDD that had a formatted capacity a bit over 10,000,000 bytes. (Bill Gates: Who would ever need more than 10 million bytes?) However, without one in hand (and an old RLL controller) I wouldn't care to say if it was closer to 10,000,000 or 10,485,760.
Never stated it otherwise. The point with that example was to make people see that sometimes base-10 is being used instead of base-2.
Used or presented as? Even the old BCD-based math processors used binary (base 2) number systems under the covers. There was a certain predictability of rounding errors as a result. Then again, there are many examples of computer parts that aren't binary based themselves, but were made to be binary based in order to be compatible with the "rest of the system."
I don't mind the sarcasm as it implies that you agree that "All this... so the HDD companies can put bigger drive size numbers on their marketing materials... " is indeed untrue and just plain idiotic to say
Oh, I think it IS true (even if the statement is idiotic.)
Things haven't been base-2 all of the time, that is something from the last couple of years. Also, not every part of a computer uses base-2. Things like clock speed and bus speed use base-10. Base-10 is also common in networking.
I can't speak for the "tube era" of computing devices, but ever since the concept of transistor based processors took hold, they've been based on 2 state (on/off) logic. There are many technical limitations/efficiencies that result in items being based on 2^x and many measurements have been the victim of that. Decimal 100 is inefficient as a max value... it required the use of 7 on/off states (7 binary digits.) Those 7 bits could actually hold up to 128 unique combinations.
All of this, however, relates to storage (both persistent and volatile) limitations/efficiencies, and is due to the logical/binary nature of transistors.
Other systems/measurements have their own oddities and quirks. Processor speeds... actually oscillators are based on resonant frequencies, data TRANSFER speeds are usually represented in BITS per second instead of bytes per second (exception being parallel transfers which are limited to the width of the transfer mechanism.)
I'll be the first to agree that it's confusing - which is (this will sound backwards) why I don't want it changed... half-way. If we change storage to be described with proper metric abbreviations, then ALL storage should be described with proper metric abbreviations throughout the entire system. Anything less results in MORE confusion. 2 "megabytes" of memory that's volatile should be the same as 2 "megabytes" of memory thats persistent on a memory card. As Apple has done things with SL, 2 MB volatile is 2097152 bytes while 2 MB persistent is 2000000 bytes. Both numbers are describing the amount of space something takes...
I won't say they are "fixing" the system. They are changing the meaning of a prefix from one technology accepted "standard" (which also happens to be inaccurate) to a different and more "people friendly" accepted standard (that is more technically accurate given the meanings of the prefixes.) However, if the meaning is only changed for part of a single concept, they've done nothing but make the mess even messier.
Oh, and thanks for the conversation.
I doubt either one of us will change the other's mind, but it's still fun talking about it.
