flat assembler
Message board for the users of flat assembler.
Index
> Projects and Ideas > fasm modification for expanded data types Goto page 1, 2 Next 
Author 

revolution
tbyte is a reserved symbol used for 10 byte extended floats.


17 Feb 2011, 16:07 

John Spera
revolution wrote: tbyte is a reserved symbol used for 10 byte extended floats. That must be my problem. I will need to come up with another label. Thanks, John 

17 Feb 2011, 16:19 

edfed
use the symbol dword, and ignore MSB. result, work on 24 bits, 3 chars.


17 Feb 2011, 17:15 

John Spera
I was thinking more about creating compact file structures. I was able to do this with C programs, now I want to do it with fasm.
Just a thought. We will eventually have computers with 12, 24, and 48 bit registers. It is my understanding that a UTF12 character would contain the language symbols for our entire world population. I am just trying to get started in that direction. Thanks, John 

17 Feb 2011, 17:51 

revolution
Even UTF16 had to be extended beyond 65536 to contain all the world's language symbols. Some languages have more than 4096.
And I highly doubt that any mainstream computer will have 12, 24, or 48 bit registers. Where did you read/hear about that? 

17 Feb 2011, 17:59 

John Spera
Hi,
The information I was reading about languages must be outdated. For some reason I tend to think that base 12 bit structures would be a good idea. I have not read that any where. It is a reflection of my thinking?? Everything is a matter of perspective and speculation. What will the next generation of computer be like? Thanks for the input. John 

17 Feb 2011, 18:14 

f0dder
John Spera wrote: For some reason I tend to think that base 12 bit structures would be a good idea. John Spera wrote: Everything is a matter of perspective and speculation. What will the next generation of computer be like? _________________  carpe noctem 

17 Feb 2011, 19:45 

edfed
with RAX expanded to 128 bits. SAX.
but still no 4GB onchip RAM nor onchip Vidéo DAC. nowadays, the goal of IA32 is not to be good for personnal development, but is made to generate the more money as possible (for them), with the help of microsoft and others... 

17 Feb 2011, 21:05 

revolution
edfed wrote: with RAX expanded to 128 bits. SAX. 

18 Feb 2011, 01:08 

Tyler
Base 6(and, by inclusion, 12) have the special property of being multiples of 3#( = 2*3 = 6). Numbers not coprime to 2 and 3 are easily distinguished in base 6 by there least significant digit. If a number in base 6 doesn't have a least significant digit of 1 or 5, it is divisible by either 2 or 3. The general case of this is that, a number, n, is not coprime to a prime, p, and all primes less than p, if it is not coprime to 1 or an prime between p and p#, where # denotes the primorial function.
I discovered this a while ago, and haven't spent too much time thinking of why it is. But it is a cool relationship. 

18 Feb 2011, 03:33 

JohnFound
BTW, there was 12 bit machines in the past. I managed to code for one of them. It was a minimachine (it occupied only half a room instead of a hall).


18 Feb 2011, 07:14 

Tomasz Grysztar
revolution wrote: Pity we don't have an EX register to expand in this way. 

18 Feb 2011, 07:22 

John Spera
Tyler wrote: Base 6(and, by inclusion, 12) have the special property of being multiples of 3#( = 2*3 = 6). Numbers not coprime to 2 and 3 are easily distinguished in base 6 by there least significant digit. If a number in base 6 doesn't have a least significant digit of 1 or 5, it is divisible by either 2 or 3. The general case of this is that, a number, n, is not coprime to a prime, p, and all primes less than p, if it is not coprime to 1 or an prime between p and p#, where # denotes the primorial function. I am not that deep into math, however I understand that base 12 math will eventually become the standard for physics. I read your reply several times. I would really like to appreciate your insights better. That relationship has much to offer based on what I have read. Sometimes I think we all need more humility. I suspect that it would help us to envision options that are beyond conventional wisdom. This applies to all subjects. May I ask if you played chess at the unexplained mysteries forum? A member there was Tyler? BTW I was John343. Thanks, John 

18 Feb 2011, 16:08 

John Spera
I switched the reserved word tbyte and used bytet instead. Changed that word in the memory allocations.
Rebuilding fasm generated the following error when allocating memory. The process I used involved 2 steps. The first part used the table modification. Once that worked, I used that executable for those memory region modifications. "error: invalid operand" for "stos bytet [edi]. No doubt I would get a similar error when reserving memory. Any insights as to why this happened will be appreciated. Thanks, John 

18 Feb 2011, 17:15 

Tyler
John Spera wrote:
What you are talking in just word size. It would be less ambiguous to say "12bit word." Where "word" in this sense is the machine word, and you're specifying it is 12 bits wide. No, I've never played chess online. I suck at chess. 

18 Feb 2011, 18:51 

John Spera
Tyler wrote:
OK, I asked because the name Tyler was the same. That base 6 insight impressed me. At the moment I am focusing on that 24 bit word. After that I would be interested in those 12 and 6 bit storage words. Compact informational labels do have value, as I see it naturally. Thanks, John 

18 Feb 2011, 19:28 

Tomasz Grysztar
This concept of base being divisible by as many of the initial numbers as possible was already utilized by ancients when the used base 60.


18 Feb 2011, 19:43 

edfed
john spera:
fact is that X86 is based on bytes (8 bits) and even multiples of bytes (16, 32, 64). fasm cannot do anything to the CPU in order to support other bases than 8, 16, 32, 64, 80, 128 an 256 bits. it is like that, and if a HLL can give the support, it is just because they are many layer over assembler, where the machine code is an abstraction, and where developper don't understand binary. 

18 Feb 2011, 20:35 

John Spera
Tomasz:
I see your work as the foundation for next generation information tools. These seemingly unnecessary features do provide some building blocks for more advanced structural changes, as I visualize them. edfed: If I could be sure that we would have only 8/16/32/64 based computers for another 40 years, I might agree with you. Since I can build data structures in C that are 24 bit values, I am convinced it can be done with fasm. I am starting to reinvolve myself with assembler after many years of using higher level languages. My assembler skills go back to those old times when computers used JCL as an operating system. As a result I am on a steep learning curve. Community efforts appeal to me. Very few of us get anywhere in life, or with a project, without the support of others. So more than anything else, I am looking for that. Thanks, John 

19 Feb 2011, 14:08 

Goto page 1, 2 Next < Last Thread  Next Thread > 
Forum Rules:

Copyright © 19992020, Tomasz Grysztar. Also on GitHub, YouTube, Twitter.
Website powered by rwasa.