wayne wrote:
Please create a patch of the exact stuff you want changed.
A decision about the one principle in these three points should
be good enough. For PermError ~ FAIL I've already proposed to
add some text in the style of "TempError" copied from -01pre2:
| If the message is rejected during the SMTP transaction for
| this reason, the software SHOULD use an SMTP reply code of
| 550 and, if supported, the 5.5.2 DSN code.
I've also proposed s/SHOULD/can/ for _all_ descriptions of DSN
codes, but actually that's an editorial detail from my POV
Do whatever you like as long as it's clear what might happen
in the "worst" case _if_ the message is rejected.
The list of issues that I know of for council review was
posted as part of the agenda items to the spf-council list.
It has been this way for the last several council meetings.
Yes, that's why I asked. I don't recall any Council agenda
covering my package of three points before 2005-05-22.
You asked here what to do about a redirect=any.invalid, and
IIRC it wasn't only me who said "like an include:any.invalid".
That's not yet reflected in -02pre1, maybe you're waiting for
the Council to decide about the old 2005-05-22 review request.
I admit that I have changed my mind on this subject
Shit happens. I stick to your original idea, because it's
simply what I expect from a robust and reliable protocol
design. "In the case of errors just ignore the policy" is
a waste of bandwidth (aka net abuse).
"If the MX is bogus try A, or try to add / remove a label
'mail' to the FQDN" would be no protocol but madness. And
like MX the SPF RR should be no nonsense.
the correct decision is to maintain compatiblity with
mengwong-spf-* on this issue.
Mengwong-01 had not only this wrong. It was fuzzy with
unknown mechanisms, ABNF errors, ignore syntax errors as
long as possible (left to right), and later why not just
use PRA based on CYA.
Obviously we disagree about this. But at least I defend
your ideas as they were back in November and in -00. And
in some lentczner drafts inspired by your ideas.
Write up the text you want to see.
Just add this to PermError:
| If the message is rejected during the SMTP transaction for
| this reason, the software can use an SMTP reply code of 550
| and, if supported, the 5.5.2 DSN code.
That's -01pre2 after s/SHOULD/can/ and maybe you like the same
s/SHOULD/can/ also for FAIL, SOFTFAIL, and TempError.
If the Council accepts my 2005-05-22 request and you need some
editorial help for handling a redirect=any.invalid in the same
way as an include:any.invalid feel free to ask. Bye, Frank