code review 2751042: crypto: switch block ciphers to detination first.

crypto: switch block ciphers to detination first.

Previously all the functions took two arguments: src, dst. This is the
reverse of the usual Go style and worth changing sooner rather than
later.

Unfortunately, this is a change that the type system doesn't help
with. However, it's not a subtle change: any unittest worth the name
should catch this.

[r, 2010-10-27 17:11:00]

thanks for this. it's been on our list for a long time.

are there no consequences outside the crypto packages? if so, LGTM.

[rsc, 2010-10-27 17:12:44]

this looks good to me but please do not submit yet.
we're tagging a release in the next 24 hours and
there's already a lot in that one.  i'd like to delay
landing of this until after the release tag.

russ

[agl1, 2010-10-27 17:44:39]

On Wed, Oct 27, 2010 at 1:12 PM, Russ Cox <rsc@golang.org> wrote:
> this looks good to me but please do not submit yet.

Understood. There are other cleanups which I believe need to be made
around crypto/block (including probably renaming it to crypto/cipher).
I'll try to get them all in one release, which probably means waiting
a bit till I have time to knock them all out in a week.

> are there no consequences outside the crypto packages? if so, LGTM.

All tests pass. In fact, nothing outside of crypto uses any of the
block ciphers. crypto/rand is the only real user and all its arguments
are (x, x), so no change needed.


AGL

[rsc1, 2010-11-01 20:31:00]

LGTM

fine to submit now

[agl1, 2010-11-02 17:54:10]

Hello rsc, r (cc: golang-dev@googlegroups.com),

I'd like you to review this change.

[r, 2010-11-02 17:57:41]

LGTM but please fix typo in CL description

[rsc, 2010-12-07 17:01:32]

Appears to have been submitted; removing reviewers.