While I appreciate the effort put in to this work, I don't think the
above statement is justification for putting this code into nmh when it
really does belong in the external SASL provider libraries.
Eric, have you approached the Cyrus people to see if they will accept
the code? I can't think of any reason why they wouldn't.
I looked on the SASL mailing lists; it's not clear to me who's
responsible for developing Cyrus SASL anymore, but there hasn't been a
release in 3 years (okay, my bad; there was a stealth 2.1.26 that was
released in 2012. So, two years?). There was someone working on OAUTH
for it here:
http://lists.andrew.cmu.edu/pipermail/cyrus-sasl/2014-May/002726.html
But ... the overall impression I get is that Cyrus-SASL is kind of
languishing. Also, from what I've seen at the first glance of Eric's code,
putting it into Cyrus-SASL is going to require some significant work
since there's extra stuff you need to do to before you login that doesn't
quite jibe with the SASL library implementation.
The potential for code conflict here is pretty high. What happens when
the Cyrus SASL library grows its own OAUTH capabilities? Most likely we
would just end up ripping this code out again. It would be much better
if the OAUTH code was incorporated in the correct place to begin with
(i.e. Cyrus).
Sigh. I understand where you're coming from ... but I'll be honest. My
dealings with the Cyrus SASL people haven't always been positive (okay,
that was a while ago, but still ... they're not so open). My glance at
Eric's code is that it is reasonable (but I still need to look more closely);
if the choice is wait for someone to put this in Cyrus-SASL, or cram Eric's
code into nmh, well, I'm 100% voting for taking Eric's code into nmh.
--Ken
_______________________________________________
Nmh-workers mailing list
Nmh-workers(_at_)nongnu(_dot_)org
https://lists.nongnu.org/mailman/listinfo/nmh-workers