|
|
|
| Authored by: Anonymous on Sunday, October 14 2012 @ 11:36 AM EDT |
one could argue then that all rational thought in affect is a
type of math because you must
have a beggining and add the stuff of life to get the result
of YOUR life.
all things equate down to math BUT faith....[ Reply to This | Parent | # ]
|
- Ehh... - Authored by: Anonymous on Sunday, October 14 2012 @ 01:12 PM EDT
| |
| Authored by: Ian Al on Sunday, October 14 2012 @ 11:44 AM EDT |
Does it change the argument? The symbology in today's practical computers is
based on 8, 16, 32, or 64 groups of binary bit symbols.
I made the comment that "the only things that the processor can get from
and put to memory is the binary symbols '0' and '1'". That is still true no
matter what the word length employed.
What is even more true is that the processor engineer only determines the
symbology used at the processor instruction level. As you work up through the
layers of software, the symbology of the software changes according to the
mathematical language used.
Take letters of the alphabet, for instance, when manipulated according to the
math algorithms higher up in the language heirarchy. The processor cannot
manipulate those symbols directly even if, say, a 7/8 bit ASCII symbol
expressed as a binary number is the same word length as the word length in
memory. With Unicode, that relationship is lost.
In fact, the processor does not process 32 bit binary numbers at all. Some
endian fixated engineer decided that he could make a processor manipulation of a
word made up of a number of individual binary symbols look like an arithmetic
manipulation of a binary number as long as you look at it from the right end!
All the processor 'knows' is to apply AND, OR, NOT to the individual bits as
'instructed' by the processor design engineer.
---
Regards
Ian Al
Software Patents: It's the disclosed functions in the patent, stupid![ Reply to This | Parent | # ]
|
| |
| Authored by: jesse on Sunday, October 14 2012 @ 04:21 PM EDT |
Take the simple case of a winmodem...
All it does is convert a byte to a bit stream...
the computer can only see one bit at a time.
You are strictly talking about an IMPLEMENTATION, not fundamentals. The original
computers were bit serial devices, as were the early calculators.
This was done simply to keep the physical expense down. It was also one of the
reasons they were slow.[ Reply to This | Parent | # ]
|
| |
| Authored by: Ian Al on Tuesday, October 16 2012 @ 05:14 AM EDT |
Not as easy as one, two three.
I see I missed a bit! That gives me a chance to patronise you with stuff you
know very well in order to complete my thoughts.
My written counting is,
zero
1
The colour, red
III
0100
The colour, green
six
.......
1000
Which demonstrates a few of the ways of writing symbols to represent numeric
values. The elemental symbols are the letters of whatever alphabet I use. Some
elemental symbols have meaning. '1' means the numeric value of one. 'k' means
'thousands of'. The full value of the alphabet comes from combining symbols into
words. The combination symbol '1000' means the numeric value 'eight'. Or 'one
thousand'. I have to make it clear in my writings, which symbolic meaning I
intend by the word I use.
The elemental symbols in computing are the pairs 1 and 0, true and false and Off
and On. The intended meaning is the meaning when applied to the electrical
circuits enacting boolean logic.
I have seen the truth table for an AND gate shown using all three symbol pairs.
A B Output
0 0 0
0 1 0
1 0 0
1 1 1
A B Output
Off Off Off
Off On Off
On Off Off
On On On
A B Output
False False False
False True False
True False False
True True True
Most use the '0' and '1' symbolism. The designer of the electrical circuit is
saying that if you use the standard electrical voltage conventions for the
symbols '1' and '0' then the circuit will correctly perform the boolean algebra
in the truth table. It is well known that if the reverse of the conventions are
used, the same electrical circuit will perform another boolean function.
A B Output
1 1 1
1 0 1
0 1 1
0 0 0
The output becomes A OR B. If the convention is used at the input and the logic
is reversed at the output then the output becomes A NAND B. Again, this is all
with the same electrical circuit. In summary, the processor logic only does what
it says on the tin if one uses the symbol/voltage relationship intended by the
circuit designer.
The two input AND gate demonstrates the use of the elemental binary symbols 1
and 0 together to form words. Word length depends on the design objectives of
the computer circuit designer.
In logic circuits there are rarely word size conventions. It is only when the
concepts of 'data' and 'address' are introduced that word size becomes
significant. A standard LCD display module has eight logic input pins and a
Read/Write logic input pin. Alphanumeric characters are represented by symbolic
binary data words from the ASCII logic table. When the R/W input is set to
logical 1 then the ASCII data word symbol on the data logic buss is (boolean)
gated via the eight d (for 'data') input ports into the display memory and
translated into a displayed symbol from the alphabet.
A full computer system adds the 'address' word. Read/Write becomes a more
meaningful logical symbol. To 'write' a symbolic data word 'into' a word memory
location the processor arranges for the symbolic word in a processor registor to
be presented via a boolean gate array at the data logical word bus. The data bus
can be both a set of processor logical inputs or a set of outputs depending on
the processor function.
In addition, the processor arranges for the address word symbol for the required
word memory location to be presented at the processor memory word buss output.
This is passed to a boolean gate array that puts a 1 on the memory location
enable port. Since only one memory location has the enable port set to logical
1, all the other locations' logic prevents any change as a result of the word on
the data buss.
Finally, the memory Read/Write input port is set to 1. Setting it to 0 would be
the logical symbol to pass its contents to the data buss for reading. Or perhaps
not. Perhaps the electrical convention for DRAM is the opposite. As long as all
the computer design engineers use the same symbolic convention then the computer
should work.
That reminds me that the bit symbol pair of 0 and 1 does not tell the whole
symbolic story of computers. The AND gate always has an output logical value of
0 or 1. If the R/W port of memory is set to R then, by AND gate standards, it
would always force its contents on to the data buss. Every memory word would try
to be on the data buss at the same time. The designers of boolean logic have
added a third logic state of 'undefined', 'no symbol' or NC using tri-state
logic. The memory word location data outputs are set to this third, no symbol,
state when not having the enable port set at logical ON.
In fact, everything connected to the address and data bus must have tristate
logic ports so that individual devices like the processor, memory and PCI hubs
can share the busses for their own logical words.
One question that relates to my previous comments: does the operating system use
low addresses for its own use or high addresses? Logically speaking the question
is nonsense. The concept of high and low addresses comes from interpreting
address words as binary numeric values. It helps us when programming at the
machine level, but the processor treats every bit in the address word as having
equal value. There is no most significant bit.
The same applies to the data bus. The processor treats every bit in a data word
as having equal value. The processor is designed so that the word pattern is
interpreted by the processor using the declared convention in the processor
device specification.
However, in the case of arithmetic instructions, the designer arranges the
boolean logic into arrays that do arithmetic and informs us that, for this part
of processor logic only, one end of the word has greater numeric significance
than the other end. Where two words are combined to produce a long word
expressing a larger range of numeric values, the designer has to declare whether
the first word read is more significant than the second word or vice verse. I
can't remember which end is up in the Intel endian system.
It is the experience of decades that makes us treat all words in computer
systems as binary representations of numeric values. Actually, we don't do that.
We write the imagined values as hexadecimal (hex) numbers. Its all part of the
abstract thought and ideas that get us writing software.
Come to think of it, we do the same thing when designing and using ICs. Data
inputs are numbered from 0 upwards. Except when its '1'. Even the device pins
have standardised numbering. d6 port is rarely on pin 6 of the IC. In fact, a
data 'bus' sometimes seems almost randomly allocated to numbered IC pins. We
have to read the device data.
So, next time that someone tells you that the ASCII hex value for a space is 20,
pedant him to death!
---
Regards
Ian Al
Software Patents: It's the disclosed functions in the patent, stupid![ Reply to This | Parent | # ]
|
|
|
|
|
