[Openmcl-devel] OpenMCL Versions

Justin Fagnani-Bell justin2 at fagnani.com
Wed Aug 16 19:32:06 PDT 2006


I just joined the list so I could chime in on this issue.

I've used MCL for algorithmic music programs for a little while now,  
and I know there's a lot of other's out there who do as well (as well  
as OpenMusic, based on MCL). I was looking forward to using OpenMCL  
for it's Cocoa support so I could see if I could tie into CoreAudio/ 
MIDI, but I own a MacBook Pro now, and I'm pretty much left out in  
the cold as far as Lisp goes.



> In a perfect world, you wouldn't have to pick one or the other.  It's
> not a perfect world.  I would certainly rather not have to say that
> OpenMCL doesn't support most current Apple hardware; unless and until
> time and/or money appear for a 32-bit Intel port appear, I have to
> say that.
>
>

Can you put some sort of donation link up on the website? I had a  
friend do this for a stupid little iTunes utility he wrote. Only a  
small percentage donated, but the amounts were pretty decent. He  
generated thousands of dollars in donations. Considering that  
commercial Lisp packages cost hundreds of dollars, you might see some  
very good donation amounts. You can set it up so that if you don't  
get a port out you refund the donations, which seems to increase  
donation rates quite a bit. I'd certainly donate some if you did  
this. It's cheaper than buying a new 64-bit MacBook Pro when it comes  
out.



> The good news is that (if rumor sites can be trusted and if the port
> to x86-64/Darwin goes reasonably smoothly) that won't be true
> indefinitely and may not be true for very long.  If a 32-bit Intel
> port was started now and was done at end of the year, how much (then)
> current Apple hardware would it support that the 64-bit port doesn't ?
>
>

Maybe none, but maybe just the mini and the MacBook, which could  
represent 100,000's of units. But there will certainly be millions of  
very new 32-bit Intel Macs that aren't going to be replaced for a while.



> One of the things to bear in mind is that (unlike the PPC and Sparc
> and (AFAIK) MIPS architectures), x86 progams generally benefit from
> being compiled in 64-bit mode: 64-bit programs generally burn memory
> and cache lines faster, but (unlike their counterparts on saner
> architectures) 64-bit x86 programs have twice as many registers
> available to them.  C compilers like that at least as much as Lisp
> runtime systems do, and the Intel world has additional motivation
> for switching to 64 bits beyond the (still fairly esoteric) need
> to overcome address space limitations.
>
>

The advantage of the extra registers is often overstated. On average  
it looks like much less than a 10% performance increase. Some  
specific apps really use the extra registers, and OpenMCL might be  
one of them, but even then it's not some insane performance boost.



> I'm not sure how long it'll take for 32-bit Intel systems to turn into
> G3s (and however long it takes there will certainly still be many
> millions of them out there).
>
>

I hate to think that I'll need to upgrade my brand new notebook soon  
just to run Lisp. Apple really has sold a ton of Intel iMacs,  
MacBooks, MacBooks, and Mac minis recently. A lot of people are left  
out. The most recent G3 is only 3 years old. There's going to be a  
lot of 32-bit Macs around 3 years from now. There will be whole CS  
and music labs full of Intel Macs this fall.

I hope there's a way that it's worth it to you to do this.

Thanks,
   Justin





More information about the Openmcl-devel mailing list