On Thu Feb 16 15:34:57 2012, Cyrus Daboo wrote:
Hi Dave,
--On February 16, 2012 10:25:22 AM +0000 Dave Cridland
<[email protected]> wrote:
Which is a vote from me for "yes, SRV option makes sense".
Oh, totally agree. Discovery is awesome, user-configuration is
insane. I
don't see there's any valid argument the other way.
Practical experience with SRV has shown that, whilst it is fine for
large service providers to do that sort of thing (c.f., Google,
iCloud etc), the practicalities of actually being able to setup SRV
records and have proper SSL cert validation is heard, particularly
for "hosted domain" type applications.
I agree with what you're saying, but I think it's getting
substantially easier as time goes on.
It used to be a rarity for XMPP services to have SRV records, and
very rare to have them relied upon, but it's now commonplace, and the
numbers of systems that require SRV resolution for them to be
reachable is increasing.
Having discussed with several engineers who have implemented SRV it
is clear that having each app do it separately is problematic
because there are a bunch of awkward issues. Not in the least is
the simple issue of even getting SRV from standard DNS libraries
(e.g. one Android developer mentioned that the size of his app
increased significantly when he had to bring in libraries to do
SRV). Then there is all the certificate verification stuff that
needs to be done to properly implement discovery. Overall this is
much harder than it needs to be, and other non-standard approaches
have proved that.
Again, I think SRV resolution in software is increasingly common, and
becoming generally easier.
It used to be next to impossible in many environments, whereas it's
now merely awkward.
Dave.
--
Dave Cridland - mailto:[email protected] - xmpp:[email protected]
- acap://acap.dave.cridland.net/byowner/user/dwd/bookmarks/
- http://dave.cridland.net/
Infotrope Polymer - ACAP, IMAP, ESMTP, and Lemonade
_______________________________________________
imap5 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/imap5
