Justin Murdock writes:
Sam Varshavchik wrote:Justin writes:
Sam Varshavchik wrote:
Justin Murdock writes:
Am I missing something obvious?
"MYRIGHTS" is an ACL1 command. The draft actually refers to the "LIST(MYRIGHTS)" command.
I think this confusion is down to the moving target, and while we were refering to "the draft" we were in fact referring to different drafts - you the ACL2, I the ACL1.
I'm looking at draft-ietf-imapext-2086upd-01.txt (and now -02.txt) which is a proposed "updated list of rights" for ACL1 which has as standard "lrswipkxtea" with "cd" as virtual rights - which MUST be returned /in addition/ to the "kxet" flags in a MYRIGHTS response.
Courier-IMAP does ACL1, per RFC 2086, and the ACL2 draft. It doesn't do this mutation, which is better described as ACL1.5.
I note they've "fixed" the "c" flag to be "k" and "x"
They didn't really "fix" anything. They just broke more things, and called it a "fix".
Hence the inverted commas. Abandoning the "d" flag (or augmenting it) makes sense as its function has been split. Changing the "c" flag allows an updated ACL client know if it is dealing with an old or updated ALC1 server.
I'm afraid that if I did that, I suspect that a lot of ACL1 code would break.
I though the point of having backwards compatability was that you didn't?
ACL2 is not backwards compatible with ACL1.
this becoming mute with the
ACL1.5 is also not backwards compatible with ACL1 also. As you've discovered. And everyone else who uses ACL1 will also discover this, in due time.
It's bad enough that IMAP already has major software interoperability issues. But it's still not enough for those pinheads. They are on a mission to break even more stuff.
I believe that the prudent thing for me to do is to step aside, and sit on the sidelines watching the major train wreck unfolding in front of me, from a safe distance.
And taking pictures.
pgpKQbeuuSegj.pgp
Description: PGP signature
