Re: Geniuses at expedia.com

2009-08-07 Thread Henrik K
On Thu, Aug 06, 2009 at 01:50:45PM -0700, Mike Cappella wrote:
 On 8/6/09 6:31 AM, Mark Martinec wrote:

 No it doesn't. Header fields names are case-insensitive.

 A space after : is shown in every example in 2822, but I don't see a
 requirement that it be there.  It is extremely unusual not to see it.

 There is no requirement for a space after a colon.

 (but yes, the rest of the header field body is wrong)

 Perhaps useful, this has proven to be 100% no FP over the past 3 years

 # From w/tab
 header L_TAB_IN_FROMALL =~ /\nFrom:\t/s
 score  L_TAB_IN_FROM6

 ymmv

Pretty good here..

OVERALLSPAM% HAM% S/ORANK   SCORE  NAME
  024942799550.238   0.000.00  (all messages)
  0.676   2.7504   0.02880.990   0.000.01  T_TAB_IN_FROM

For some reason all the FPs appeared to come through MailScanner. Seems it
liked to delimit all headers with tabs? Wonder what was that.. maybe some
old version bug.

Mind if I put that in SA mass checks?



Re: Geniuses at expedia.com

2009-08-07 Thread Mark Martinec
  header L_TAB_IN_FROM  ALL =~ /\nFrom:\t/s
- header L_TAB_IN_FROM  From:raw =~ /^\t/m

  Mark


Re: Geniuses at expedia.com

2009-08-07 Thread Mike Cappella

On 8/6/09 11:44 PM, Henrik K wrote:

Pretty good here..

OVERALLSPAM% HAM% S/ORANK   SCORE  NAME
   024942799550.238   0.000.00  (all messages)
   0.676   2.7504   0.02880.990   0.000.01  T_TAB_IN_FROM

For some reason all the FPs appeared to come through MailScanner. Seems it
liked to delimit all headers with tabs? Wonder what was that.. maybe some
old version bug.

Mind if I put that in SA mass checks?



Good to see other's stats.  Go ahead, its all yours.

This was a real hitter a while back; it has slowed somewhat of late.

On 8/7/09 4:45 AM, Mark Martinec wrote:
 header L_TAB_IN_FROM  ALL =~ /\nFrom:\t/s
 -  header L_TAB_IN_FROM  From:raw =~ /^\t/m


Thanks Mark.  Much nicer.

--
   Mike


Re: Geniuses at expedia.com

2009-08-06 Thread Jari Fredriksson
 tests=[BAYES_00=0.1, DCC_CHECK=1.5, DCC_REPUT_60_69=0.1,
 HTML_MESSAGE=0.001, INVALID_DATE=1.245,
 MIME_HTML_ONLY=0.957, NO_REAL_NAME=1,
 RELAY_COUNTRY_US=0.001, SARE_OEM_S_PRICE=1,
 SPF_SOFTFAIL=1] autolearn=no received:from

Why positive score for BAYES_00? It's supposed to mean that the mail is HAM.




Re: Geniuses at expedia.com

2009-08-06 Thread Matus UHLAR - fantomas
  tests=[BAYES_00=0.1, DCC_CHECK=1.5, DCC_REPUT_60_69=0.1,
  HTML_MESSAGE=0.001, INVALID_DATE=1.245,
  MIME_HTML_ONLY=0.957, NO_REAL_NAME=1,
  RELAY_COUNTRY_US=0.001, SARE_OEM_S_PRICE=1,
  SPF_SOFTFAIL=1] autolearn=no received:from

On 06.08.09 09:36, Jari Fredriksson wrote:
 Why positive score for BAYES_00? It's supposed to mean that the mail is HAM.

because he apparently uses own scoreset with different scores:

score INVALID_DATE 2.303 1.651 1.329 1.245
score SPF_SOFTFAIL 2.301 0.654 0.698 0.596
score DCC_CHECK 0 1.37 0 2.17
score MIME_HTML_ONLY 2.299 1.672 1.925 1.457

DCC_CHECK and MIME_HTML_ONLY have higher default scores than he uses...

-- 
Matus UHLAR - fantomas, uh...@fantomas.sk ; http://www.fantomas.sk/
Warning: I wish NOT to receive e-mail advertising to this address.
Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu.
It's now safe to throw off your computer.


Re: Geniuses at expedia.com

2009-08-06 Thread Joseph Brennan


Michael Scheidell scheid...@secnap.net wrote:


 and did you ever hear of Y2K?  can't you afford to send out two more
digits in the year?

date:31 Jul 09 10:13 -0800



Do they really write date: instead of Date:?  That violates RFC 2822.

A space after : is shown in every example in 2822, but I don't see a
requirement that it be there.  It is extremely unusual not to see it.

The two-digit year is obs-year, and MUST NOT be used to generate
messages, but MUST be honored when interpreting messages.   Hm.

Is the Expedia server really in Alaska?  I think that's the only place
in timezone -0800 this time of year.

Joseph Brennan
Lead Email Systems Engineer
Columbia University Information Technology




Re: Geniuses at expedia.com

2009-08-06 Thread Mark Martinec
Joseph Brennan writes:
  date:31 Jul 09 10:13 -0800
 Do they really write date: instead of Date:?  That violates RFC 2822.

No it doesn't. Header fields names are case-insensitive.

 A space after : is shown in every example in 2822, but I don't see a
 requirement that it be there.  It is extremely unusual not to see it.

There is no requirement for a space after a colon.

(but yes, the rest of the header field body is wrong)

  Mark


Re: Geniuses at expedia.com

2009-08-06 Thread Neil Schwartzman
The geniuses send their regards; they are a customer so I pinged them:

Hi Neil,
 
Thanks for heads-up. I've forwarded the information to our corporate
domain/smtp management folks.
 
Sincerely,

MUNGED 
Lead Operations Manager, EWW Database Marketing


On 06/08/09 9:23 AM, Joseph Brennan bren...@columbia.edu wrote:

 
 
 Michael Scheidell scheid...@secnap.net wrote:
 
  and did you ever hear of Y2K?  can't you afford to send out two more
 digits in the year?
 
 date:31 Jul 09 10:13 -0800
 
 
 Do they really write date: instead of Date:?  That violates RFC 2822.
 
 A space after : is shown in every example in 2822, but I don't see a
 requirement that it be there.  It is extremely unusual not to see it.
 
 The two-digit year is obs-year, and MUST NOT be used to generate
 messages, but MUST be honored when interpreting messages.   Hm.
 
 Is the Expedia server really in Alaska?  I think that's the only place
 in timezone -0800 this time of year.
 
 Joseph Brennan
 Lead Email Systems Engineer
 Columbia University Information Technology
 
 
 

-- 
Neil Schwartzman
Director, Certification Security  Standards
Return Path Inc.
0142002038




Re: Geniuses at expedia.com

2009-08-06 Thread Mike Cappella

On 8/6/09 6:31 AM, Mark Martinec wrote:


No it doesn't. Header fields names are case-insensitive.


A space after : is shown in every example in 2822, but I don't see a
requirement that it be there.  It is extremely unusual not to see it.


There is no requirement for a space after a colon.

(but yes, the rest of the header field body is wrong)


Perhaps useful, this has proven to be 100% no FP over the past 3 years

# From w/tab
header L_TAB_IN_FROMALL =~ /\nFrom:\t/s
score  L_TAB_IN_FROM6

ymmv

--
Mike