[Home] [Groups] - Message: [Prev in Group] [Next in Group]
11821: [MUD-Dev] Re: about MOO
[Full Header] [Plain Text]
From: Gevan <shanos@es.co.nz>
Newsgroups: nu.kanga.list.mud-dev
Date: Sat, 20 Nov 1999 19:08:51 +1300
References: [1] [2] <-newest
Organization: Kanga.Nu
Quoting Dan Root (dar@thekeep.org):
> >There was once a disk based version of LambdaMOO, IIRC, though it may have
> >never really gotten off the ground.
>
> Lots of talk, but I never saw any code, nor anyone who claimed to have it,
> other than, as you mention, LPMOO, which is amusing, but not really the
> right way to run a MOO. :)
I hacked one up in a couple of afternoons a couple of years ago, then
promptly threw the code away again because the MOO I was working on at
the time didn't really need it -- it was just a bit of fun. I never
bothered coming up with a good on-disk representation, so it was quite
slow. (As I recall I just used the normal db object write/read code,
which is text-based.)
So I claim to have had *something*. Of course, I don't now.
The internal db abstraction in 1.8.0ish LambdaMOO is just good enough
to make it a lot easier than you might expect, particularly as most
aquisition of an (internal) object pointer goes through a single
function. Scatter a few unfind()s to match the find()s and you're,
well, at least some of the way there (although not heading for
stunning efficiency.)
--k
_______________________________________________
MUD-Dev maillist - MUD-Dev@kanga.nu
http://www.kanga.nu/lists/listinfo/mud-dev