On Mon, Apr 3, 2017 at 2:18 AM, Arthur A. Gleckler <xxxxxx@speechcode.com> wrote:

[Also, this doesn't sound like something that should be allowed as an errata.]

Generally, I agree, but please see my reply as part of the earlier discussion at http://srfi-email.schemers.org/srfi-133/msg/4219939.

I recall that, and it seems ok to make the exception in this case.

To emphasize Shiro's argument, though, the errata should be made

more clear. Instead of

2016/9/2 (Reversed order of arguments to vector-cumulate.)

let's explain what changed and why

3. 2016/9/2 Changed order of arguments to vector-cumulate

from (vector-cumulate f vec knil) to (vector-cumulate f knil vec)

to match the argument order of vector-fold.

to help anyone who downloaded or implemented a version

before said errata understand what happened, and to help

clarify when the SRFI or reference implementation itself

hasn't been updated, as in this case.

Alex