On Sat, 2 Feb 2002, Martin McClure wrote:

> At 7:46 PM -0400 2/2/02, Lex Spoon wrote:
> >
> >For a persistence framework, etc., I have no idea.
>
> This second class of usage is the primary motivation for adding the
> header bit. All the alternatives to this that I've seen either
> severely limit the functionality or are extremely ugly.
>

Do you really need a header bit? What about just reserving a seperate
range of memory for such objects, then, you just see if its in that range
before deciding whether or not to allow the mutation.

There are also issues of what about:

   foo array mutate
   foo mutate


> So I still think the proposal to be worth seriously considering.
>

Can you explain how, where, and why you need these bits? There doesn't
seem room to get them..

Also, I would be careful about considering removing hashBits. Take away
just two hashbits, and Morph becomes at least twice as slow, or even
1000x.

May I suggest not doing any serious revamps of the image format yet. Just
enough for the new VM. This is because I'm beginning to suspect that we
may wish to consider a really big revamp of object memory and the GC
system. Such a revamp should have a more careful consideration and would
not be anywhere near to usable for at least a year.

Scott