You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by Joseph Brennan <br...@columbia.edu> on 2018/01/09 15:56:19 UTC

FSL_MIME_NO_TEXT and MIME_NO_TEXT

Both FSL_MIME_NO_TEXT and MIME_NO_TEXT are very similar. Both look for a
multipart/mixed message with no "text/" part that has an attachment.
Combined score is just under 5. That's a lot.

The case in point is an application that sends a report to a few people as
a plain text document, and the only mime part is the attachment, which is
called application/octet-stream and has a .txt file extension. I feel like
this should count in __ANY_TEXT_ATTACH.

I think "multipart/mixed" allows a message with one part. Using
"application/" causes clients to treat a part as an attachment, and I think
the generic "octet-stream" is correct since there is no specific software
that must be used for a plain text file. (I'm actually surprised that there
is nothing like application/plaintext for this case, but I could not
identify such a type in a web search.)

Comments?


-- 
Joseph Brennan
Columbia University

Re: FSL_MIME_NO_TEXT and MIME_NO_TEXT

Posted by Bill Cole <sa...@billmail.scconsult.com>.
On 9 Jan 2018, at 13:47 (-0500), Matus UHLAR - fantomas wrote:

> this is a real duplicity...

Semantic note: "duplication" or "redundancy," NOT "duplicity," which is 
English for the flavor of dishonesty involving contradictory statements.


-- 
Bill Cole
bill@scconsult.com or billcole@apache.org
(AKA @grumpybozo and many *@billmail.scconsult.com addresses)
Currently Seeking Steady Work: https://linkedin.com/in/billcole

Re: FSL_MIME_NO_TEXT and MIME_NO_TEXT

Posted by Matus UHLAR - fantomas <uh...@fantomas.sk>.
>On 09.01.18 10:56, Joseph Brennan wrote:
>>Both FSL_MIME_NO_TEXT and MIME_NO_TEXT are very similar. Both look for a
>>multipart/mixed message with no "text/" part that has an attachment.
>>Combined score is just under 5. That's a lot.
>>
>>Comments?

On 09.01.18 19:43, Matus UHLAR - fantomas wrote:
>meta FSL_MIME_NO_TEXT (__CTYPE_MULTIPART_MIXED && !__ANY_TEXT_ATTACH_DOC)
>
>meta __MIME_NO_TEXT   (__CTYPE_MULTIPART_ANY && !__ANY_TEXT_ATTACH)
>
>meta MIME_NO_TEXT      __MIME_NO_TEXT && !__BOUNCE_CTYPE &&
>	!__CT_ENCRYPTED && !ALL_TRUSTED && !__MSGID_APPLEMAIL &&
>	!__USER_AGENT_APPLEMAIL && !__HAS_IN_REPLY_TO && !__HAS_X_REF &&
>	!__HS_SUBJ_RE_FW && !__LCL__ENV_AND_HDR_FROM_MATCH

btw:

mimeheader __ANY_TEXT_ATTACH     Content-Type =~ /text\/\w+/i
mimeheader __ANY_TEXT_ATTACH_DOC Content-Type =~ /text\/\w+/i

this is a real duplicity...

-- 
Matus UHLAR - fantomas, uhlar@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.
Emacs is a complicated operating system without good text editor.

Re: FSL_MIME_NO_TEXT and MIME_NO_TEXT

Posted by Matus UHLAR - fantomas <uh...@fantomas.sk>.
On 09.01.18 10:56, Joseph Brennan wrote:
>Both FSL_MIME_NO_TEXT and MIME_NO_TEXT are very similar. Both look for a
>multipart/mixed message with no "text/" part that has an attachment.
>Combined score is just under 5. That's a lot.
>
>Comments?

meta FSL_MIME_NO_TEXT (__CTYPE_MULTIPART_MIXED && !__ANY_TEXT_ATTACH_DOC)

meta __MIME_NO_TEXT   (__CTYPE_MULTIPART_ANY && !__ANY_TEXT_ATTACH)

meta MIME_NO_TEXT      __MIME_NO_TEXT && !__BOUNCE_CTYPE &&
	!__CT_ENCRYPTED && !ALL_TRUSTED && !__MSGID_APPLEMAIL &&
	!__USER_AGENT_APPLEMAIL && !__HAS_IN_REPLY_TO && !__HAS_X_REF &&
	!__HS_SUBJ_RE_FW && !__LCL__ENV_AND_HDR_FROM_MATCH

score FSL_MIME_NO_TEXT 3.274 1.728 3.274 1.728
score MIME_NO_TEXT     1.726 1.999 1.726 1.999

why the rest of comments is correct, this is imho a good example where
similar rules that should not have that big combined score.

looking at those scores, one of those rules should be removed, or at least a
meta should be created, lowering combined scores (aparently through mass-check)


-- 
Matus UHLAR - fantomas, uhlar@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.
Spam = (S)tupid (P)eople's (A)dvertising (M)ethod

Re: FSL_MIME_NO_TEXT and MIME_NO_TEXT

Posted by John Hardin <jh...@impsec.org>.
On Tue, 9 Jan 2018, Kris Deugau wrote:

> Joseph Brennan wrote:
>>  The case in point is an application that sends a report to a few people as
>>  a plain text document, and the only mime part is the attachment, which is
>>  called application/octet-stream and has a .txt file extension. I feel like
>>  this should count in __ANY_TEXT_ATTACH.
>
>>  (I'm actually surprised that there is nothing like application/plaintext
>>  for this case, but I could not identify such a type in a web search.)
>
> Personally, I think anything using "application/octet-stream" for any ASCII 
> text anything is Doing It Wrong.
>
> If it's really just pure ASCII text, "text/plain" would be a lot better.

+1 - that application is behaving in a very spammy manner.

Also (making an assumption here), you should probably save processing 
overhead by bypassing internal automatically-generated emails in your 
glue.


-- 
  John Hardin KA7OHZ                    http://www.impsec.org/~jhardin/
  jhardin@impsec.org    FALaholic #11174     pgpk -a jhardin@impsec.org
  key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C  AF76 D822 E6E6 B873 2E79
-----------------------------------------------------------------------
   Maxim XXXV: That which does not kill you has made a tactical error.
-----------------------------------------------------------------------
  8 days until Benjamin Franklin's 312th Birthday

Re: FSL_MIME_NO_TEXT and MIME_NO_TEXT

Posted by Kris Deugau <kd...@vianet.ca>.
Joseph Brennan wrote:
> The case in point is an application that sends a report to a few people 
> as a plain text document, and the only mime part is the attachment, 
> which is called application/octet-stream and has a .txt file extension. 
> I feel like this should count in __ANY_TEXT_ATTACH.

> (I'm actually 
> surprised that there is nothing like application/plaintext for this 
> case, but I could not identify such a type in a web search.)

Personally, I think anything using "application/octet-stream" for any 
ASCII text anything is Doing It Wrong.

If it's really just pure ASCII text, "text/plain" would be a lot better.

-kgd

Re: FSL_MIME_NO_TEXT and MIME_NO_TEXT

Posted by RW <rw...@googlemail.com>.
On Tue, 9 Jan 2018 10:56:19 -0500
Joseph Brennan wrote:

> Both FSL_MIME_NO_TEXT and MIME_NO_TEXT are very similar. Both look
> for a multipart/mixed message with no "text/" part that has an
> attachment. Combined score is just under 5. That's a lot.
>
> The case in point is an application that sends a report to a few
> people as a plain text document, and the only mime part is the
> attachment, which is called application/octet-stream and has a .txt
> file extension. I feel like this should count in __ANY_TEXT_ATTACH.

more like __ANY_TEXT_ATTACH_DOC, but deliberately disguised text
shouldn't count as text.
 
> Using "application/" causes clients to treat a part as an attachment,

What's wrong with:

Content-Disposition: attachment; filename=...;