John M McIntosh <johnmci at smalltalkconsulting.com> is widely believed to have written:

> >...Even the simple logic I used requires at least some
> >understanding about typical timing behavior when writing a primitive. If
> >it's not really gaining anything in general (and 1% doesn't really count)
> >then I'd probably leave it out.
> >
> >Cheers,
> >   - Andreas
> 
> Ok, 5% is quite different from 1%, it's hard to decide if it's noise then.
> So leave it out.
I disagree; 5% or 1% is still an improvement and although Andreas _demo_
changes are a little complicated I think that implementing the original
suggestion (indexed prims are assumed short by default, named prims are
defined as long via the calling protocol) is very simple to live with.
Generally speaking noone much should be adding indexed prims and there
are plenty of us here that understand enough to be able to point out
when a suggested new prim needs any tweaking.

It would also be a possible help for the hoped-for day when we can deal
with translated prims in some future jitter. In general the right
approach to making changes is to decide if they are Pareto-efficient. If
they benefit anyone and disadvantage noone, make the change. 5%
improvement on windows is good; 1% on Mac is no disadvantage. I don't
have any time to try the result on Acorn right now but I suspect it
would be at the high end of the 1-5% range since the timer is quite
expensive. I think the cleanup in the logic of primitiveResponse is
almost worth it on its own!

tim
-- 
Tim Rowledge, tim at sumeru.stanford.edu, http://sumeru.stanford.edu/tim
Strange OpCodes: SHUTDOWN: (See EFB)