|
Boost : |
From: Terje Slettebø (tslettebo_at_[hidden])
Date: 2002-08-13 08:27:38
>From: "Paul Mensonides" <pmenso57_at_[hidden]>
> From: "Douglas Gregor" <gregod_at_[hidden]>
>
> > That's one form of type-vector, but not the type that Burton et al.
used. They
> > used type vectors that look like this:
>
> I see what you mean, but why couldn't this be done with a regular
type-vector?
> Just efficiency at huge sizes? 10,000 elements isn't that many compared
to a
> database, but having a database encoded in the source is just plain ugly!
:) I'm
> still wondering if this example even applies. As in, can this sequence
(up to
> several thousands) even be run through an MPL algorithm?
I see no reason why not. If you specialise begin<> and end<> to return
iterators to it.
Besides, you don't need to traverse it all. I go more into this in another
posting.
Regards,
Terje
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk