World: r3wp

I am now able to compile RED program that throws "Illegal instruction" 
under OSX which is good thing as it means that the compiler now produces 
at least valid Mach-O header :)

You're implementing Mach-O?

I'm just trying to ;)

Cool!

Rebolek: great news! :-)

I'm glad someone picked up the task of implementing Mach-O output 
format.

Rebolek: maybe OS X requires position-independent code for executables? 
Checking it right now...

I should have basics required for functioning executable soon. If 
it still throws "illegal instruction" then, well...

Doc no it doesn't, I just still haven't got all the sections done.

but I'm close.

Good, it will make the porting easier.

A link for Andreas: http://developer.apple.com/library/mac/#documentation/DeveloperTools/Conceptual/MachOTopics/1-Articles/dynamic_code.html


I was thinking about using some similar trick for ELF to allow efficient 
PIC support (avoiding the indirection table). It works by having 
a fixed offset between CODE and DATA segments in memory, so DATA's 
start address can be calculated at runtime, and all globals can be 
accessed using a simple offset.


It should be enough to reserve enough space for CODE segment (let's 
say 10MB) and make DATA entry point fixed? I wonder if the relative 
offsets between segments are preserved when the segments are moved 
by OSes (I hope so)?

10 MB should be enough for everyone ;-)

Heck, 640K should be enough for anyone. ;-)

And...REALLY cool Bolek.

I first wrote 1MB, then I raised it thinking of possible libs that 
could get statically linked by users. But, it could be even easier, 
just hardcoding the DATA address offset somewhere in the CODE segment 
at linking time might be enough (no need to reserve fixed space in 
CODE segment).

Another milestone reached ;) Now I get "segmentation fault" instead 
of "illegal instruction" :)

Have you defined a new runtime in red-system/runtime/ for OS X?

Right now I just copied the Linux one.

Does OS X support the same syscalls?

Yes. It's not recommended, but I think that for basic test it's enough.

http://www.opensource.apple.com/source/xnu/xnu-1504.3.12/bsd/kern/syscalls.master

are there specific CPU limitations? could this be done for, say, 
a 6502?

It requires 32-bit integer support for source compatibility with 
other platforms, but if you don't care about compatibility, it should 
be possible to make a 6502 emitter ;-)

ok

6502 is a bit far reach with its 8/16 bits architecture, but 68k 
should be doable while preserving all Red/System's features.

that would also be quite interesting.

I'd been phantasising about that, and about porting my 6502 emulator 
:-)

porting 6502 emulator to 6502? :-)

If I would port the emulator and someone would do a 6502 backend, 
then yes :-)

doc, do you planing later external bindings features like SWIG / 
FFI  /ALIEN  etc ?
or would it be possible to access C/C++ libs directly with RED?

Red/System is a C level language, so you can just call C functions

Calling C++ requires a C wrapper for the C++ interface, as always 
with C++

Binding systems such as SWIG are way too heavy for Red, and C functions 
are neither foreign nor alien to Red/System, so FFI and ALIEN would 
be too heavy as well :-)

The only thing you have to do to call a C function is to translate 
the defintion of it in the C header to Red/System. It would be possible 
to write a C header parser to automate that process, one of the things 
that SWIG does. However, it would lead to ugly Red function definitions 
that don't make the most of the interface

Agreed with Kaj.
 

AFAIK, it might be possible to interface with C++ at very low-level, 
to be able to instanciate C++ objects and call their methods from 
Red. Same solution for obj-c interfacing, which will be required 
at some point to port Red to iOS.

thx, for info

That's in any case possible very naturally for Objective C. Io for 
example has a transparent Objective C bridge that allows accessing 
any Objective C object

For C++, you'd need RTTI to go in the direction of that level of 
transparency

Ruby has also a bridge to obj-c (I guess Python too).

The Io one is likely simplest, and BSD licensed, so you might be 
able to use some code from it

changing currently my serverbackends from rebol to luajit/FFI. what 
a relief and cool experience to just drop some c header inline and 

access the appropiate lib in seconds. the mapping is nice also.  
thought that this feature is quite important for Red.

Nice! We will see if we can come up with something even simpler with 
Red.

with the last beta you can detect also a real firework of new and 
planned very usefull bindings there.

a feature like this defines a real 'multipropose language' as I stated 
many times before.

I'll make sure that Red won't lag behind on that aspect. But, as 
Kaj says, binding to a library like SWIG might be overkill for Red, 
I'm pretty sure that we can come up with much shorter and better 
integrated solution for Red (remember that we will have PARSE support 
and ability to use dialects).

Objective-C messages can be passed with calls to C functions that 
are built into the runtime. All languages which have a different 
object model use these functions to interact with Objective-C in 
their compatibility layers. Internally, even Objective-C code compiles 
down to calls to these functions.

Good to know, seems easier at first look, to interface with than 
C++.

Interfacing with C++ functions and objects requires matching the 
name mangling and structure layout conventions, and those are different 
between compilers. There are some attempts to standardize these conventions, 
but their success is limited. This is why the developers of most 
Unix clones tend to hate C++. The best solution on most platforms 
tends to be to use C wrappers or one of the binary object interop 
standards like COM or CORBA.

Note that there is a typo in the manual you linked here: http://www.rebol.com/docs/core23/rebolcore-16.html#section-3.1
- the caret character is listed as #"^" when it should be #"^^".

Actually, almost all of those character literal examples are missing 
their leading ^.