On Jun 1, 2:25 pm, Mxsmanic <mxsma....DeleteThis@gmail.com> wrote:
> The mistake concerns encoding of information in an address space in the
> mistaken belief that it reduces address capacity in a linear way, when in fact
> it reduces address space in an exponential way. It is the most persistent and
> classic of mistakes made by engineers and other people in IT. The fact that
> it doesn't make sense to you proves my point.
No, actually, it just proves you're unclear, are rambling and are out
of your depth. I understand very well how address spaces work. I've
taught C programming at undergraduate level. I know how pointers and
memory allocation work. I understand the difference between an address
and the contents of the address, and I understand dereferencing. I
passed a test which included questions on these things to get my
current job.
Firstly "Encoding" information in an address space doesn't make sense.
You don't talk about "encoding" information in an address space you
talk about storing it.
Secondly if you store one byte in an address space, you store one byte
If you store 100 bytes you still store 100 bytes. That's definitely a
linear relationship. If your addressing scheme enables you to address
2^32 bits, you use up those addresses in a linear fashion. The key is
that if you're addressing works with bytes or words you don't address
individual bits, so yes if you have a flag it takes a whole byte or
word instead of a single bit. However an integer or character requires
a whole byte/word. Yes you waste some but even the waste is quite
literally linear and not waste. Storing 2 integers uses 2 times as
much memory as storing 1 integer and requires 2 entries in an address
space. Storing 4 uses 4 times as much memory and 4 entries in the
address space. Unless you're using some mighty strange definitions of
the words linear and exponential, you're just wrong.
Thirdly it depends on your machine architecture (or VM architecture if
you're talking about a VM) as to how the address space is divided. For
example it's entirely possible to address more than the maximum number
of bytes in an address space, by creating multiple address spaces.
However data isn't shared between them and there's a performance
penalty to pay.
What you've said is so vague and badly worded that it basically
amounts to gibberish. If you're trying to say something coherent point
me to an article or at least cite a specific architecture. (Of course
I know you won't do this because you can't but will claim you don't
want to like a 3 year old).
You really should stop pretending to know more than you do. To anyone
with an education in the area you're prattling about, it makes you
sound completely absurd. This is just like when you claimed
anaphylaxis was all in the head. Just yesterday I was reminded of that
when I saw a show called RPA which follows the medical treatment of
patients at a hospital. A 2 year old came in with an anaphyalactic
reaction. His lips were swollen, and he was having trouble breathing.
I thought of you and quite frankly I thought "dumbass!" then turned to
my fiancee and said "it's all in his head" and chuckled at your
expense. I now double check every "fact" that you state since I find
what you say is occassionally true, sometimes there's a grain of
truth, but very often its complete bollox. Yet you claim to be an
expert on everything.
>> Stay informed about: Sammy, I am beginning to come around to your way of thinking 
FAQ
Profile
Private Messages
Log in
