On Wed, Jul 9, 2014 at 6:51 PM, Matt Oliveri <[email protected]> wrote:
> On Wed, Jul 9, 2014 at 9:44 PM, Jonathan S. Shapiro <[email protected]> > wrote: > > If the separate checker terminated, we'd be using it as the primary > checker. > > The idea is that the separate checker wouldn't need to be standardized > or especially well behaved, since only code for the primary checker > would be standards-compliant. The inference system would be some > optional tool on top for developers to use at their discretion. I understood you. But the standard for inference is sound and complete inference. If you have that, no "auxiliary" inference engine is necessary, required, or useful.
_______________________________________________ bitc-dev mailing list [email protected] http://www.coyotos.org/mailman/listinfo/bitc-dev
