On Thu, 2004-01-29 at 02:43, Gustav Foseid wrote:
I have read some of the discussion og whether or not to use _spf as a
namsepace in DNS.
It seems that using a namespace would make it easier to get SPF
approved as an RFC, and in most ways it is a cleaner implementation
seen from a DNS point of view.
Can someone point me to where what I'm about to describe has already
been addressed, because I think I missed it. I am also operating under
a limited knowledge of DNS transactions, so I may be off-base. Please
forgive me and correct me if I'm wrong.
As far as versioning goes, wouldn't it be easier, and faster, and fewer
packets if SPF clients that supported version X look up these records in
this order:
v2a._spf.domain
version 2 with experimental extensions
(most likely used for local testing only)
v2._spf.domain
plain version 2
v1x._spf.domain
version 1 with experimental exts
(rare if v2 is available, and again only used for local
new features testing)
v1._spf.domain
plain version 1
(or whatever the syntax is for these kinds of records, my memory fails
as to what has been discussed). You're welcome to dig these records for
domain = leave-it-to-grace.com, v2a is the only one that doesn't exist.
And if they don't find anything, fall back on querying for previous
supported versions (for known values of X)? This would seem to save
time and bandwidth compared to requesting all TXT records, and then
searching for the highest version number you support, and filtering out
records that don't match the prefix? If there are a number of TXT
records, how are they returned? Are they returned in a small number of
DNS transactions? Is the packet size limit (512) an issue in this
case? Is there any issue with DNS queries being synchronous for
retrieving multiple records?
I would suggest that the draft is changed to incorporate a namespace
as soon as possible.
I agree. It is much cleaner, and while does use the TXT RR, it lessens
the argument for actually needing another RR just for SPF. Because the
difficulty of getting new RRs assigned, it might be wise to consider
_-style extensions to fit within the same realm that the assigned RRs
do. That is, they are new "RRs", but not "RRs" in the same sense that
true RRs are.
But I also agree with everyone else that it might be too late to change
it. And the majority is on this side. Bah.
--
Andy Bakun <spf(_at_)leave-it-to-grace(_dot_)com>
-------
Sender Permitted From: http://spf.pobox.com/
Archives at http://archives.listbox.com/spf-discuss/current/
Latest draft at http://spf.pobox.com/draft-mengwong-spf-02.9.4.txt
Wiki:
http://spfwiki.infinitepenguins.net/pmwiki.php/SenderPermittedFrom/HomePage
To unsubscribe, change your address, or temporarily deactivate your
subscription,
please go to
http://v2.listbox.com/member/?listname(_at_)©#«Mo\¯HÝÜîU;±¤Ö¤Íµø?¡