Why doesn't the discrete computer operates in basis 10?

[luquinhas0021]

Comparing with the binary processor, would the basis 10 processor slow? or far more complex?

[StainlessS]

would the basis 10 processor slow?
or far more complex?
Probably Yes, and Yes.
We only use base 10 because we have 10 fingers, its not really ideal for use in a digital computer. [False/True, No/Yes, 0/1]
Additionally, we as humans find it easier to work in and remember decimal/denary numbers rather than vast strings of zero's and one's.
[So conversions to and from base 10 often occur]
There are/were often instructions in microprocessors for BCD, Binary Coded Decimal.

EDIT:
BCD was used in many early decimal computers, and is implemented in the instruction set of machines such as the IBM System/360 series and its descendants, Digital Equipment Corporation's VAX, the Burroughs B1700, and the Motorola 68000-series processors. BCD per se is not as widely used as in the past and it is no longer implemented in newer computers' instruction sets (e.g. ARM); x86 does not support its BCD instructions in long mode any more. However, decimal fixed-point and floating-point formats are still important and continue to be used in financial, commercial, and industrial computing, where the subtle conversion and fractional rounding errors that are inherent in floating point binary representations cannot be tolerated.[11]
BCD:- https://en.wikipedia.org/wiki/Binary-coded_decimal
Not sure, but think Visual Basic Currency Type is implemented in above mentioned decimal fixed-point and floating-point formats.

Binary arithmetic can be performed using digital logic gates, AND, OR, XOR, NOT, and variations, [at its most basic chip level] whereas
base 10 would be somewhat more awkward.

Analogue computers however can easily use decimal number base or eg base 12, as its just a matter of calibration.
[just a matter of how far apart the markings are on a meter]

[feisty2]

what is it like to compute Hamming code in a base 10 system (without any conversion to/back from base 2)?
I suppose it should be possible, but with a lot of extra complexity.

first we need to define some new mathematical structures for our beloved integers, namely stuff like base10 xor operator and other base10 logical operators. I suppose most of these structures should constitute at least an Abelian group (but don't bother me to prove it), so we start from some nice properties we get to work with.

apparently base10 xor and base2 xor must be homomorphic at the very least, but do we need to make them isomorphic, or even identical (the homomorphism being an identity function from base10 to base2)?

since base2 xor is embedded into base10 xor (given by injective homomorphism), the parity property of base2 xor (which is essential to Hamming code) must have been preserved in base10 xor, but preserved in what manner? Does it still make apparent sense with base10, like something we can reason with? a lot of questions you need to answer

anyways, I guess you'd get the exact same algorithm if base 10 xor is defined to be identical to base 2 xor, but there're other possibilities if base 10 xor is defined to have a richer structure than base 2 xor, maybe you'd get an algorithm that searches for error in a 10-branch manner rather than bisection? who knows

EDIT: I was gonna say "don't ask stupid questions", but changed my mind after seeing grandpa stainlessssssss actually took some time to write a meaningful reply. His reply leans towards more of a low level perspective, digital logic gates and stuff, my reply should give you some insights from a high level perspective.

[StainlessS]

Yeah, what young Sheldon said.

[filler56789]

Comparing with the binary processor, would the basis 10 processor slow? or far more complex?
Yes and yes, of course, indeed, surely.
What puzzles me is, we are in the so-called 21st century and there still are people ignoring what should be very obvious :-/

On the other hand, or maybe in the same hand, many people keep treating certain stupid limitations as "facts of Nature" and/or "science"...

"Must Follow Standards freezes the Cybernetic Sector into existing functionality. In some cases, e.g., ASCII that's an undeniable good ... for a while. ASCII has a range of coding to support teleprinters. Who the heck uses teleprinters these days? But there they sit, hogging space that these days could be used for other, more important, purposes. It's possible to state ASCII is obsolete; it was designed for 8 bit systems in a 64 bit world."

The technological change over the past 40 years continues today. Much of it is not reaching the consumer market because of existing "standards," e.g., WinTel. And the fact 95% of the people on the software side know bugger-all about hardware, its design, architectural trade-offs between hardware and software, and hardware/software integration.
Putting it simply, COMPUTER SYSTEMS AVAILABLE IN 2013 ARE SQUARELY BASED ON THE LIMITATIONS OF 1975 HARDWARE USING PARADIGMS AND HEURISTICS DEVELOPED IN 1956.

source: https://eurotrib1.eurotrib.com/comments/2013/6/7/05318/71420/2

[StainlessS]

Tis the Operating System that is king.
Once Linux really does take over, then can finally rid ourselves of the antiquity that is Intel based. [although there is ARM version of Win].
With an OS that can move relatively easily to another CPU, so we will no longer be locked into using the 8086/8080 based chip and all
of the horror that it involves. It is basically the businessman that has locked us into Wintel, refusing to use anything other than that
which all other businessmen use, ie Windows and Intel based chip.
Due to linux, the infinite loop will finally get broken, the competition between chip manufacturers will finally recommence.

[feisty2]

modern x86 has absorbed many features from other architectures for its internal implementation (e.g. a modern x86 chip has many physical registers, one logical register like rax actually corresponds to multiple physical registers, x86 instructions are also internally translated to RISC-like micro-operations), x86 is really just a binary level interface, the internal architecture could be something totally different. I doubt that switching to another architecture would magically bring any significant performance boost.

[StainlessS]

Aha, so you're an Intel man.
Well when it finally does happen, you will be able to stay Intel.
CISC has and will always be the final goal, RISC is a stop-gap measure to allow for faster processing of simple instructions
with less silicone producing heat [less heat means faster clock speeds possible], but complex functions done slower
using sequence of RISC instructions. When heat/clockspeed problem solved, then CISC will be the norm again.
I just dont like the LDA style of intel processors, tastes way too much like Z80/8080, MC68000 style was much more flavoursome.
[although had some microcode layer above real registers, much nicer to use assembler for 68000].
I understand that current intel is much improved, but I just wanna heave when I see assembler source for it.
Also, I prefer Big Endian, although I do see the advantages of Little Endian.

EDIT: To each his own.

[cogman]

Perhaps surprisingly, early computers (vacuum tube no less) were base 10. They had 10 different voltages that represented the 10 different digits.

So why did we abandon that? Simple, it's just too hard to tell the difference between 10 different voltage levels. It ended up causing a bunch of difficulty in managing the system. Further, the tubes to do base 10 arithmetic were very fragile and difficult to manufacture.

Those early computers were so bad and unreliable that you'd often still need a person to double check the result and they'd have the computer compute the result multiple times and take the average as the "real" answer :D

In fact, mechanical computers of the time were far more reliable and in some cases faster than the tube based computers (certainly used less power). Their only downside was they weren't as easy to program.

It seriously amazes me that we kept pouring funds into computer technology. They were so wildly impractical in the early era.

Now, if you want something truly interesting, russians used base 3 computers for quite a while. In a lot of ways, base3 computers would be superior to today's base2 computer if for no other reason than the increase in memory density. Most of the base2 circuits we have today would work just fine if we did base 3 instead (particularly, DRAM would be practically unchanged).

[wonkey_monkey]

they'd have the computer compute the result multiple times and take the average as the "real" answer

Funny that we're sort of going back to that when it comes to quantum computing.

[StainlessS]

Base 3 indeed, Yes, No, and Maybe.

EDIT: I got a book somewhere on Quantum Mechanics, I never understood the first page so did not read any further.

EDIT: Quantum mechanics on wikipedia:- https://en.wikipedia.org/wiki/Quantum_mechanics

EDIT: Schrodingers Cat:- https://www.youtube.com/watch?v=HCOE__N6v4o

Big Bang Theory Quote 2056: Penny:
No, no, no, no, I didn't forget. Um, there's this cat in a box and until you open it, it's either dead or alive or both. Although, back in Nebraska, our cat got stuck in my brother's camp trunk, and we did not need to open it to know there was all kinds of dead cat in there.

[wonkey_monkey]

Base 5 adds "I don't know" and "Can you repeat the question"