Yeah, they are different for sure, The last attempt failed though the previous one was better.
The reson I have a hard time trying to code it is that In one end I wouldl like to build the spec on what's already there. And on the other hand make sure that it play well with dynamic states. The issue is that I don't want to spoil some natural optimisations due to a bad spec. The old approach meant that you alway's needed to use the expensive version of the construct and spoil much of it's simplicity in implemenetation. I will have new try, this evening. /Stefan On Wed, Mar 27, 2013 at 10:46 PM, Noah Lavine <noah.b.lav...@gmail.com> wrote: > On Wed, Mar 27, 2013 at 5:44 PM, Noah Lavine <noah.b.lav...@gmail.com> > wrote: >> >> >> Since it's definable with such a simple macro, I don't think it's >> appropriate for a SRFI. >> > > Sorry, I just realized that that's not a sensible objection. If it's a > better interface than parameters, then it certainly should be a SRFI. > However, I won't be convinced of that until I've played with both for a > while. So I would suggest waiting a bit until it's clear what the best way > to manage variables like this is. > > Noah
