Moving up...

[GabrielCA]

Why on earth would anyone want to program in assembly?

Because it once was a standard, because it allows to control things in a more direct, immediate manner, because it represents a challenge (to quote Sir Edmond Hillary, first man to climb the Everest, "I climbed it because it was there"), because many great games were presumably made with it, etc.

[Silverwind]

It's the programmer's equivalent to Irish politicians: it's completely unnecessary.

because many great games were presumably made with it, etc.

Not even Rogue was written in assembly.

[GabrielCA]

To the best of my knowledge, it is, indeed, totally unnecessary by today, but certain people such as me (and this is strictly our point of view, our opinion) believe it is interesting to program assembly, because it tends to be more challenging.

[WarHampster]

1. You learn a lot about how computers work.

2. You directly manipulate registers, so you have unparalleled efficiency.

[Gan]

My interest is piqued. Do you have any example code you could post to satisfy our nerdy desires?


-Gan

[GabrielCA]

My interest is piqued. Do you have any example code you could post to satisfy our nerdy desires?

Why, yes, although I must warn you I designed this system when I was about 13...
This is probably the first thing I ever programmed for this virtual system

Code: [Select]

;Print out the first 16 Fibonacci numbers
Do 1 10 1 1 ;x=1
Do 4 10 1 1 ;y=1
Echo 1 ;print x
Do 5 10 1 15 ;timeout=15
PtrTo 6 ;store location to myrepeat
Do 5 30 1 1 ;timeout-1
Do 1 20 2 4 ;x=x+y
Do 2 10 2 1 ;ox=x
Do 3 10 2 4 ;oy=y
Do 4 10 2 2 ;y=ox
Do 1 10 2 3 ;x=oy
Echo 1 ;print x
BoolDie 5 ;exit if timeout<0
PtrFrom 6 ;goto myrepeat

Note that "Do" is an operation that is used for all types of arithmetic.

[EDIT] As I have previously said, the most complex thing I managed to program was a BREAKOUT clone; it took approximately 550 lines of macro-assembler code plus the development of a primitive tile graphics "library"

[Gan]

 
It's like trying to read Japanese... so foreign.


-Gan

[GabrielCA]

As with Japanese (or 6502 assembly, or English, for that matter), it is a matter of practice (or of having designed that silly system)

[Connors]

Nice... even Gandolf looks impressed by this. (lols)

[Silverwind]

My dad had to use Assembly all the time back in the days of Kobalt, Fortran and... well... Assembly. It's way too time consuming to be practical. It's a nightmare to do anything you could do simply in anything else.

1. You learn a lot about how computers work.

2. You directly manipulate registers, so you have unparalleled efficiency.

I wouldn't build apps in assembly for efficiency anymore than I'd build a TV for efficiency.

ertain people such as me (and this is strictly our point of view, our opinion) believe it is interesting to program assembly, because it tends to be more challenging.

Well if that's what drives you, let nobody stand in your way.

Just don't expect to finish your first app before you're 70! Hehehe...

[WarHampster]

Why are you even arguing about this... it can be fun to play with outdated stuff that is useless by today's standards.

[j]

Different people are interested in different aspects of development. Some coders continue to script within engines, others go on to write those engines. I'd imagine there's some appeal to the latter interest group in learning and working with something in Assembly, for all the reasons they've previously argued.

[GMG Mike]

I have no idea... I've only recently 'migrated' from Mac OS 9 to Mac OS X (old version)

Macintosh forward compatibility...
I've managed to run a test application both in System 6 and Mac OS 9. Everything was identical except for the fonts, colors and GUI.

Is the forward compatibility as good in Mac OS X ?

Kind of an old thread but stuff like this interests me.

The easiest and perhaps strictest interpretation of backward/forward compatibility would be to write a strictly Carbon C or C++ program, adhering only to the "CarbonLib 1.6" specification and not using any exclusive OS 9 or OS X features. This program, compiled as CFM PPC would technically run on anything from a Power Macintosh 6100 (1994) to a brand new Mac Pro with Snow Leopard and Rosetta (2010), using only a single binary. You would need to use CodeWarrior.

It may then be possible to compile the same project in Xcode. You could then produce an Intel binary so that Intel Macs will be able to run the program more efficiently.

Supporting very early OS X versions (10.0, 10.1) usually isn't worth the time. Mac OS 9 probably has more daily users than those versions of OS X.

Tiger is the most universal version of the Mac OS, running on anything from a Power Macintosh 7500 (with at least 604 processor, using XPostFacto patches) (1995) to various Intel Macs released in 2007, with only a few rare exceptions. Leopard, in theory, could run on a Power Macintosh 7500 with a G4 processor, but no one has attained this feat. Leopard can run on anything from a Power Macintosh G3 (with G4 upgrade) (1997) to various Intel Macs released in 2009, but there are numerous exceptions (all G3 laptops and G3 iMacs and any other Macs not upgraded to a G4 processor)

System 7.5.3 is the most universal version of the Classic Mac OS, running on anything from a Mac Plus (1986)  to various Macs released in 1996, possibly some Macs released in 1997 exclusive of G3s.

[GabrielCA]

Wow, that's a lot of information !
Very interesting, thank you.

What about the "FATCarbon" module, or compiling a program as a combination of fat binary and Carbon ?

[GabrielCA]

Wait, you made a language? Elaborate  

BTW if this still interests anybody, I am now making a limited pseudo-C compiler for my VM system.
Here is the new way to write the Fibonnaci numbers test (it compiles and works !) :

Code: [Select]

#define QUANTITY 20
int x = 1;
int y = 1;
int ox;
int oy;
int remainingNumbers=QUANTITY;

Echo(x);
remainingNumbers--;

sub loopPoint:
x+=y;
ox=x;
oy=y;
y=ox;
x=oy;

Echo(x);

remainingNumbers--;

if(remainingNumbers>0){
goto loopPoint;
}