/

[amigaksi]

by Lorraine on 2008/11/10 16:34:09
...
>Yes I also am stuck in the functional PC ways of today, >with Windows XP being my primary OS. Shame on me
...
>@ amigaksi
>
>I use Windows every day but I don't really like the way >it's dominant - maybe why I'm so stuck in the past with my >hobby.

I use Windows 3.11 networked to another Windows XP machine.  Windows 3.11 is not 'dominant' and lets you do whatever you want with the machine and also allows me to run real-mode DOS.  I don't consider this nor using Amiga as "stuck in the past".  Windows 3.11 is also not prone to viruses/spyware as most of those are 32-bit apps so you don't need to constantly have some antivirus software hogging up 50% of your CPU time and there's no I/O protection on the ports so everything is as fast as possible.  I use hardware that is (was) standard so I don't need any drivers (sound blaster, IDE HDs, VESA-compatible VGA, etc.).  In fact, I backed up my entire Windows 3.11 OS on one 1.44MB floppy disk (and can boot from it).  I still need to have XP running because of "others".  Believe it or not you can directly write to the Sound Blaster I/O ports and VGA ports and create real-time sound and visual effects like on Amiga/Atari although not as exact and not with as much control.  I believe creativity in general took a hit by making standards at API level rather than at hardware level.  
Some things in the past are better than what's in the present so "stuck in the past" has no meaning in regards to technology.  I prefer computers that are simple so complicated programs can be written without having to worry about the complications of the processor.  I suggest that Intel drop the L1 and L2 caches, power management, branch prediction caches, and few other things from their processors.  They should slow down their processors so that each instruction executes in fixed amount of time or can easily be predicted as to the number of cycles.  I am pretty sure they can do 1 Ghz machine and run all of RAM at the same rate as the processor.  They can use a single branch prediction so we can predict whether a branch will occur or not and that would take care of most of tight inner loops.  Get rid of virtual/physical/linear memory bullcrap and let you align opcodes to physical memory locations (that factually exist) for optimizing code. A few more things but don't want to hog up the thread...

[amigaksi]

>...They can use a single branch prediction so we can predict whether a branch will occur or not and that would take care of most of tight inner loops....

That should be "whether a branch will occur cycles or not".  For example, if we were upgrading a 486-type processor (to some theoretical 886) and take the 3 cycles for branch taken and 1 cycle if it does not and put in a single branch prediction in parallel with previous instruction then following code for SQRT(EAX) for EAX=4 would take 18 cycles.

;EBX=FLOOR(SQRT(EAX))
Sqrt Proc Near
Xor EBX,EBX        ;1 cycle
SqrtEAX: Sub EAX,EBX   ;1 cycle
Inc EBX   ;1 cycle
Sub EAX,EBX   ;1 cycle
Jns SqrtEAX   ;3 first time, 0 afterwards, 1 last iteration
Dec EBX   ;1 cycle
RET   ;say 3 cycles

[KimmoK]

It starts from each single person.
MAKE PEACE.


(waiting my HW for the RED and for the BLUE, most likely will use use every possible color there exist in Amigaworld, the original amiga checkmark had all colors anyway... ;-)  The FUTURE IS NOW ! ) :-o  :crazy:  :crazy:  :crazy:  :crazy:
 8-)

[Atheist]

Hi  amigaksi,

I agree 100%!!!

So much of those UNKNOWN ACTIONS interfere with making a predictable outcome.

That is what made Amiga so AWESOME. It was an ultimately flexible/hackable system, and yet it was just like a console because what was done on one, would happen on every other model.

So, you write an assembler groovy hack, and it worked on all the other machines (A1000, A2000, A500, A600, A3000 and the CDTV), while the A1200, A4000, and CD32 weren't 100% backwards compatible. A disaster! (Even that contributed to the breakdown of Amiga market share.)


What will bring everything all back to where it should be is the NatAmi!!!!

Full OCS/ECS/AGA compatibility. AWESOME!

For example, it'll be like having what is essentially the power of a Playstation 2 (with much more RAM!!!!), but it is ONE HUNDRED PERCENT hackable!!!!!!

NOTHING beats that!

[Atheist]

Hi Lorraine,

Why are there so many camps?

AOS1.0 to 3.9
AmigaForever-UAE
AOS4.x
AROS
Amithlon
AmigaXL
MorphOS
Anubis
(did I miss any?)

Simple, actually. Everyone thought they were doing the right thing.


AROS's problem is, every program has to be recompiled.... who is going to do that, and WHERE are the source code for most of the Amiga SW available? Too difficult.

MorphOS. Not based from the original source code.

Amithlon. Does'nt run software that uses the custom chipsets, ist that like more than 75% of the SW that's available for AOS? Probably doesn't run purely assembler written programs.

AmigaXL. Don't know anything about it.

Amiga Forever, to this day, it's MOSTLY compatible. Not enough at all in the AGA realm.

Anubis. Well, it's not Amiga, but something similar. We lose all old SW. Who is going to recreate AMOSProfessional? And a billion other programs?

Only AmigaOS4.x is the continuation of AOS1-3.9, as the original source code was looked at and everything that Amiga was attempting to do oh so many years ago is reimplemented and/or improved.

[jahc]

:roll:

Just for the record, this man (Atheist) does not represent me.

[benJamin]

Because of the number of people running *UAE having difficulty with their host machine/OS?

Honestly, I've asked some non-Amiga questions here simply because I have a good idea who to trust, if not to give immediate answers then to point me in the right direction (should Google already have failed).

I also remember the early days when all the latest tech. developments were reported here, I had even no need or want to go to Slashdot (for example).

Amiga.org has always rocked!

:-D

[takemehomegrandma]

Why?

Well, Atheist's post above pictures the reason in essence quite well:

Only AmigaOS4.x is the continuation of AOS1-3.9

I.e. an elitist and uncompromising "we are so much holier than you" attitude, "there can be only one, and it is OS4", "only OS4 uses the original source code", "only OS4 has the Amiga name", etc. The point is that this attitude has proved to be so totally mis-fitted, time after time, and outright untrue, and as a result many of the "hardcore BAF's" has softened during the years and turned more into *the pluralists* we encounter here at amiga.org (it's very positive to see that amiga.org with its allowing attitude and pluralistic direction has 6467 members, while for instance amigans.net only has 865 ;-)).