spf-discuss
[Top] [All Lists]

RE: [spf-discuss] proposed spf3 non-contiguous address notation

2007-01-14 17:15:34
Alex van den Bogaerdt <mailto:alex(_at_)ergens(_dot_)op(_dot_)het(_dot_)net> 
wrote on Sunday,
January 14, 2007 1:01 PM -0600:

I do like this, but where does it end ?

What about "v=spf3 192.168.10-15.0/24 -all" ?
What about "v=spf3 a:mailhost0{1-3}.example.org -all" ?

I think we also need to be careful not to make things too complex,
as this will introduce more room for mistakes.

Here's the motivation, you decide.  If you have many outbound hosts to
designate, listing all of them by hostname is not advisable for two
reasons:  each name requires another query to resolve and the record
gets very long.  In both cases, the problem is DNS bandwidth.  For very
long records, they will not fit in a single DNS response packet.  These
have been real objections that many people take seriously.  The %d macro
helps with the length only.

Listing a large number of hosts by IP avoids additional queries to
resolve the names, but the resulting record can still be very long.
While many domains can arrange their outbound relays to occupy one or
two contiguous ranges, they do not always align conveniently with bit
masks.  In those cases, a shortcut IP notation would greatly reduce
record length without confusion to human readers.  It does not make
sense to combine it with CIDR slash ranges and there is no need to
propose general text substitution macros.  If anyone wants those, make a
case for them separately.

--
Seth Goodman

-------
Sender Policy Framework: http://www.openspf.org/
Archives at http://archives.listbox.com/spf-discuss/current/
To unsubscribe, change your address, or temporarily deactivate your 
subscription, 
please go to http://v2.listbox.com/member/?list_id=735

<Prev in Thread] Current Thread [Next in Thread>