Mailing List Archive

wayne schrieb:
> In <45178540.7000303@byteaction.de> Norman Maurer <nm@byteaction.de> writes:
>
>
>> After work more on jspf and fix a bug in our tests we notice that the
>> longlabel test is wrong. We think that there are 10 chars more needed in
>> the label.
>> The label limit of 63 chars is not matched cause there are only 53 chars.
>>
>
> I agree that you need more than 63 characters.
>
> I haven't checked the test suite, but I guess you should also test
> the email addresses "foo@.example.com" and "bar@a..example.org". Both
> of those contain zero length labels that are not on the far right.
>
> Of course, none of these are valid domain names and therefore can't
> have SPF records published in the DNS so some implementations may get
> away with just doing a DNS lookup and not directly checking the syntax
> of the domain.
>
> Similar problems can produced by having the longlabel, or zerolabel
> being in the localpart of the email address and using the %{l} macro
> variable. E.g. "v=spf1 exists:%{l}._foo.%{d} -all" with
> "spam...a@example.net".
>
>
>
> -wayne
>
I think the main problem is that the label is incorrect in the
testsuite. It has only 53 chars.. So im almost sure its a bug in the
testsuite..

i also not understand where you got the mail from of ? The testdata is:
longlabel:
spec: 4.3/1
helo: mail.example.net
host: 1.2.3.5
mailfrom:
lyme.eater@A12345678901234567890123456789012345678901234567890123.example.com

result: none

so i don't see whats wrong with the mailfrom . Maybe im wrong ;-)

bye
Norman


I

-------
To unsubscribe, change your address, or temporarily deactivate your subscription,
please go to http://v2.listbox.com/member/?listname=spf-devel@v2.listbox.com

Norman Maurer wrote:

> It has only 53 chars.. So im almost sure its a bug in the
> testsuite..

Yes, an "off by one" Murphy, five instead of six '1234567890',
plus 'A' and '123': 5*10+4=54, but 6*10+4=64 is what we want.

Frank


-------
To unsubscribe, change your address, or temporarily deactivate your subscription,
please go to http://v2.listbox.com/member/?listname=spf-devel@v2.listbox.com

On Tue, 26 Sep 2006, Frank Ellermann wrote:

> > It has only 53 chars.. So im almost sure its a bug in the
> > testsuite..
>
> Yes, an "off by one" Murphy, five instead of six '1234567890',
> plus 'A' and '123': 5*10+4=54, but 6*10+4=64 is what we want.

The test-suite is fixed. It is not an important test, since most
DNS implementations will simply return nothing when presented with such
an invalid label. However, it is important that implementations not
crash with bad data.

--
Stuart D. Gathman <stuart@bmsi.com>
Business Management Systems Inc. Phone: 703 591-0911 Fax: 703 591-6154
"Confutatis maledictis, flammis acribus addictis" - background song for
a Microsoft sponsored "Where do you want to go from here?" commercial.

-------
To unsubscribe, change your address, or temporarily deactivate your subscription,
please go to http://v2.listbox.com/member/?listname=spf-devel@v2.listbox.com