On Oct 8, 2009, at 6:46 AM, Stephen J. Turnbull wrote:
Barry Warsaw writes:I've also heard convincing arguments from folks in the Python community in both camps: "using anything other than strings internally is insane; no, using anything other than bytes internally is insane."They're both right, of course. The problem is figuring out who is right when. ;-)
Indeed!
Just to ramble a little longer, it's been argued that we should give up on idempotency, but I'm not convinced.If we can't achieve ... ah, isn't "invertibility" what you mean here? ... "idempotency", then we're dropping information somewhere along the line. Also, there are part types (pgp-signed, I'm looking at you) where it's absolutely essential that we be able to roundtrip the body byte for byte. So I'm -1 on giving up.
Yeah, "idempotency" probably is not the right term, though I think historically that's what's been used. Math geeks, what's the right term here? :)
I completely agree with you (of course :). The way I look at it is that we lose this important principle only when the source data lacks complete information, i.e. is defective. Although we can still invert in the face of some defects (and we should), I think we officially make no such guarantees unless the model is defect-free.
-Barry
PGP.sig
Description: This is a digitally signed message part
_______________________________________________ Email-SIG mailing list Email-SIG@python.org Your options: http://mail.python.org/mailman/options/email-sig/archive%40mail-archive.com
