Hrvoje Niksic <hniksic(a)srce.hr> writes:
Uh, I really don't understand what the code above is supposed to
do.
The FSF Emacs Lisp manual seems a bit unclear:
Each char-table has a "subtype" which is a symbol. In order to be a
valid subtype, a symbol must have a `char-table-extra-slots' property
which is an integer between 0 and 10. This integer specifies the
number of "extra slots" in the char-table.
What does this mean? What are the "extra slots" good for, and why is
the code above setting them to zero? I'm baffled!
From what I gather, a char-table is a vector that has more than one
slot. (Kinda like an obarray, but not.) And parse-time used creates
a char-table with 0 extra slots, which means that it's just an
ordinary vector. Or something.
So I've now modified parse-time to just use a vector instead. Perhaps
it was an effeciency issue? Are char tables meant to be more
efficient than vectors?
Anyway, I can now use parse-time under XEmacs as well, which means
that the Gnus date handling functions are several magnitudes faster
than they used to be back when they used timezone.el to fiddle with
dates.
--
(domestic pets only, the antidote for overdose, milk.)
larsi(a)ifi.uio.no * Lars Magne Ingebrigtsen