Message245654
| Author |
barry |
| Recipients |
barry, catalin.iacob, jesstess, python-dev, r.david.murray, sandro.tosi, zvyn |
| Date |
2015年06月22日.22:05:55 |
| SpamBayes Score |
-1.0 |
| Marked as misclassified |
Yes |
| Message-id |
<20150622180549.343b50d5@anarchist.wooz.org> |
| In-reply-to |
<1435010428.05.0.000926948967956.issue15014@psf.upfronthosting.co.za> |
| Content |
On Jun 22, 2015, at 10:00 PM, R. David Murray wrote:
>We change the authobj signature to challenge=None, then the first thing we do
>in auth is 'initial_response = authobj()'. The return value can be the empty
>string or a real initial value, and we send the auth command with '
>'.join(mechanism, initial_response).strip(). Then we do the challenge part
>only if we get the 334.
Sounds good to me. I'll see if I can work up a patch. I have a hack in my
testing code to work around the lack of auth in smtpd. It's not pretty but it
works. |
|