Re: [ANN] New package: live-free-or-die
Alex Knauth <[email protected]>
| Newsgroups | gmane.comp.lang.racket.user,gmane.lisp.scheme.plt |
|---|---|
| Message-ID | <[email protected]> |
> On Sep 22, 2016, at 9:39 PM, Jay McCarthy <[email protected]> wrote: > > If I knew how, I'd do unsafe-in, but I just know how to fake a typed context. The whole point is to use the optimizer so libraries like math are usable. Libraries like math/array would still be usable without the optimizer. What you really want is a version of require (or a require transformer) that would allow untyped code to *selectively* unsafely require typed code. I would want to import as few things unsafely as possible. It's impossible to choose what to be unsafe about if it's an all-or-nothing imperative flag like that. > On Friday, September 23, 2016, Leif Andersen <leif-hCbPq2Sbls7o/[email protected] <mailto:leif-hCbPq2Sbls7o/[email protected]>> wrote: > Umm...if you're going to do this, why is it imperative, and not a require transformer? > > Also, I kind of agree with Matthias here...sigh. :'( Like, as far as I can tell, this is even worse then typed/racket/unsafe, as the optimizer is still run. O_o Also, with typed/racket/unsafe the optimizer *is* still run, so that's not any better *unless* you selectively choose what you want to be unsafe. This is worse, but not because of the optimizer; because of the all-or-nothing part. Alex Knauth -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/[email protected] For more options, visit https://groups.google.com/d/optout.