Re: Changes in GC and in pure space

[Stefan Monnier]

> Four tag bits for "annotated" (e.g., immutable) objects: very far from
> ready, and problematic on 32-bit machines (perhaps this is no longer a
> concern for Emacs 28...)

I'm still using 32bit Debian on pretty much all my machines (except for
one).  I could change to 64bit OSes on most of them, but not on the
Thinkpad X60 I use for my in-class presentations nor on the two BananaPis
I use as servers.

So, AFAIC, 32bit is not something we can drop yet:
It's OK to support it less-well, but it's not OK not to support it.

> Turning pseudovectors into their own tag type, as miscellaneous
> objects: not convinced it's worth the change, yet.

I'm not sure what this is referring to.

But in this area, I'd welcome a change which makes symbols and strings
(and intervals) use the same representation as pseudovectors (just like
buffer and misc has been merged many years ago).

They could still use their own "enum Lisp_Type" tag (some benchmarking
might be needed to decide if it's worthwhile), but the idea would
be to drop their *_blocks from alloc.c.


        Stefan