Re: Proposal: immediate strings

[Paul Eggert]

On 05/24/2012 05:51 AM, Stefan Monnier wrote:
>>> we could just grab a second bit from `size'
>>> >> for the `immbit'.
>> > Only if MOST_POSITIVE_FIXNUM <= min (SIZE_MAX, PTRDIFF_MAX) / 2.
> Why?
> This would be true if `size' were changed to ptrdiff_t (which would make
> sense, indeed), but it's currently EMACS_INT,

Ah, sorry, I was assuming the integer-overflow patch
(for bug#9874).  That patch indeed changes 'size'
to ptrdiff_t.  Any changes for immediate strings
should assume that this patch has been applied.
I haven't applied that patch only because the trunk
has been frozen against big changes since God
knows when.