Re: apple's objc runtime on linux?

[Benhur Stein]

Citando Dietmar Planitzer <dave.pl@ping.at>:

> Apple's runtime achieves lock free multi-threading support by tacking 
> advantage of some Mach APIs and by implementing a simple GC strategy. 
That's good to hear. As I said, last time I had heard of it was a long
time ago. On another side, if it depends on some Mach API, maybe it is
not very portable (that is a problem in the case of unifying runtime).

> Well, I really, if not extremely, doubt that GNU's runtime could be 
> faster than Apple's on the _same_ hardware and OS.
That's exactly what I was wanting to do. I was hoping to be able to 
compile the runtime on the machine I have access to, an x86 PC, and
it seemed to me that apple's runtime have support for it.
But from what you say in the rest of the message, I conclude that I will
need to install darwin to be able to test it, and even then, maybe 
it will not be a fair comparision, because the code has only been 
further optimized for ppc.

Anyway, it was just out of curiosity. Maybe one day I'll extract the
messenger from the library so that I can compile it without the 
dependencies, just to satisfy my curiosity.

> I think its _very_ important to keep a few things in mind when 
> comparing the GNU and Apple runtimes, in order to get meaningful 
> results:
> 
> I) First and foremost, remember that Apple's messenger is de facto only 
> optimized for PPC. This is the platform which is currently relevant to 
> Apple. The x86 messenger has not seen as many optimizations over the 
> past years.
But it was optimized for the x86 by next, wasn't it? The GNU runtime
has not been optimized for many years also.

> II) Compare both runtimes on the same hardware platform and OS, only.
Sure, that was my intention. 

I only posted the numbers I got for the
GNU runtime in case someone was interested on them, to know how much
is paid on a message delivery. I for one think that too much, too early,
optimization has been put on GNUstep's core caching IMP's, obscuring 
code, and most of them unnecessarily. Maybe because of hearing that
you pay a high price on each message (and that scares a lot of beginners).

> Anyway, regarding a merge of the Apple and GNU runtimes: I'm personally 
> not a very big fan of this idea.
Nor am I.
But seeing that as one of the objectives of the new maintainers of the
runtime, I just wanted to start some discussion on it (and satisfy this
old curiosity of mine).

Best regards,
Benhur Stein