Re: Image spam

2006-11-02 Thread Ira Abramov
Quoting Shachar Shemesh, from the post of Thu, 02 Nov:
 
 They get deep and detailed information on all of your spam activity, and
 so long as you even remotely log in to their servers, they get to cross
 that with your searching, so they get to target-ad you. No, I highly
 doubt they will mind.
 

exactly what I meant in my previous post...

 The question is, knowing the real cost, do you?

I personally don't. I don't use it for illegal stuff so I have nothing
to hide. (If I was doing anything illegal, that needed real anonymity, I
would not use unencrypted text on a public server anyway :-)

-- 
Theoreticly sane
Ira Abramov
http://ira.abramov.org/email/

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-11-02 Thread Ira Abramov
Quoting Amos Shapira, from the post of Thu, 02 Nov:
 On 02/11/06, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
 ROTFL - What a great idea! Using gmail as a spam filter! Any idea if there 
 is
 any restriction on using it this way? (Obviously they didn't anticipate 
 this
 kind of usage, but I wonder if this breaks any of their published rules)

I'm quite sure they DID anticipate it, and don't mind doing a little
work on your mail. in return they get a lot of valuable statistics on
language structure, vocubulary and other things they need to make their
engines better.

Mind you, their spam filters are not that amazing though, I do get about
1-2 false positives a month and about 1-2 false negatives a week, though
always image spams.

 I have multiple gmail accounts, mostly to keep my anonymity. All of
 them just have a
forward all filter to my real gmail account.

and if you will look closely, when you send out from your mail, this may
have set your From: header correctly, but it sends your main account
in the envelope. some anonymity :)

-- 
E-Z Rider(tm)
Ira Abramov
http://ira.abramov.org/email/

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-11-02 Thread Amos Shapira

On 02/11/06, Ira Abramov [EMAIL PROTECTED] wrote:

and if you will look closely, when you send out from your mail, this may
have set your From: header correctly, but it sends your main account
in the envelope. some anonymity :)


I know, but it's good enough for my needs:

1. Avoid my address being added to google'able mailing list archives.
2. I can advertise my pseudo-address in public without it being
associated with me. Once someone contacts me either I don't care that
they know who I am or they aren't capable enough to find.

--Amos

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-11-01 Thread Linux Il

On 10/28/06, Oded Arbel [EMAIL PROTECTED] wrote:
On Wed, 2006-10-25 at 01:33 +0200, Michael Vasiliev wrote: On Monday October 23 2006 20:28, Gil Freund wrote:
  Hi,   I am using a Amavis+SA+CLAM for mail filtering (debian sarge packages).  Recently I am being hit by a lot of image spam. Bayesian filtering and  RBL's are not enough.
 These mails are terrible, while my personal accounts get a few dozen each, Linux-IL posting address gets a hundred per day minimum. If we only relied on SA's checking, this list would look much worse than LKML. Most of these
 emails are a crazy mix of sentences on random topics, probably web-crawler generated, plus a gif image with a nasty ad of some kind (I never care enough to mimedecode them).The problem is that the image has some random noise added to it, so you
can't even get your bayesian filter to recognize the mime data. Thiswill also, IMO, in the long run defeat OCRs in the same way thatCAPTCHAs defeat OCRs. SA's Bayesian filtering is terrible, I don't know why. (All right, I admit I
 am too lazy and illiterate at adult-level mathematics to figure out why). I'm running SA and bogofilter in parallel, clearly bogofilter is at least twenty times more accurate.I'm using bogofilter in conjunction with a large set of RBLs, and while
I do see a surge in the number of small-caps and viagra image spam, onlyrarely there is a false negative - about 4-5 times a week.--Oded::..If you lost your left arm, your right arm would be left.



This could be off topic... 

but I have unorthodox way to filter spam...

I auto forward my emails to agmail account that i opened forthis purpose only.
then from the gmail account i Auto forward the email back to other secret email for viewing the emails.

Gmail have a very strong spam filters and the spam will not be auto forwarded to you.

My old personal email was heavily spamedin a way that I gave up on that email, since i got hundreds of spams everyday butnowI get only the real emails.

This might be a good and fast solution for personal emails and private machines, I am not sure if it will work for a mail server that serve lot of people, though it might be possible with some tweaks to theDNS and mail server.


forgive me for suggesting this (no offense)but you can alsoput youremail domainunder google apps
https://www.google.com/a/

Michel


Re: Image spam

2006-11-01 Thread michael




On Wed, 1 Nov 2006, Linux Il wrote:


On 10/28/06, Oded Arbel [EMAIL PROTECTED] wrote:


 On Wed, 2006-10-25 at 01:33 +0200, Michael Vasiliev wrote:
  On Monday October 23 2006 20:28, Gil Freund wrote:
   Hi,
  
   I am using a Amavis+SA+CLAM for mail filtering (debian sarge

 packages).
   Recently I am being hit by a lot of image spam. Bayesian filtering and
   RBL's are not enough.
 
  These mails are terrible, while my personal accounts get a few dozen

 each,
  Linux-IL posting address gets a hundred per day minimum. If we only
 relied on
  SA's checking, this list would look much worse than LKML. Most of these
  emails are a crazy mix of sentences on random topics, probably
 web-crawler
  generated, plus a gif image with a nasty ad of some kind (I never care
 enough
  to mimedecode them).

 The problem is that the image has some random noise added to it, so you
 can't even get your bayesian filter to recognize the mime data. This
 will also, IMO, in the long run defeat OCRs in the same way that
 CAPTCHAs defeat OCRs.

  SA's Bayesian filtering is terrible, I don't know why. (All right, I
 admit I
  am too lazy and illiterate at adult-level mathematics to figure out
 why). I'm
  running SA and bogofilter in parallel, clearly bogofilter is at least
 twenty
  times more accurate.

 I'm using bogofilter in conjunction with a large set of RBLs, and while
 I do see a surge in the number of small-caps and viagra image spam, only
 rarely there is a false negative - about 4-5 times a week.

 --
 Oded
: : ..
 If you lost your left arm, your right arm would be left.




This could be off topic...
but I have unorthodox way to filter spam...

I auto forward my emails to a gmail account that i opened for this purpose
only.
then from the gmail account i Auto forward the email back to other secret
email for viewing the emails.

Gmail have a very strong spam filters and the spam will not be auto
forwarded to you.

My old personal email was heavily spamed in a way that I gave up on that
email, since i got hundreds of spams everyday but now I get only the real
emails.

This might be a good and fast solution for personal emails and private
machines, I am not sure if it will work for a mail server that serve lot of
people, though it might be possible with some tweaks to the DNS and mail
server.

forgive me for suggesting this (no offense) but you can also put your email
domain under google apps
https://www.google.com/a/

Michel




ROTFL - What a great idea! Using gmail as a spam filter! Any idea if there is
any restriction on using it this way? (Obviously they didn't anticipate this
kind of usage, but I wonder if this breaks any of their published rules)

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-11-01 Thread Michael Vasiliev
On Wednesday November 1 2006 21:36, [EMAIL PROTECTED] wrote:
 On Wed, 1 Nov 2006, Linux Il wrote:
  I auto forward my emails to a gmail account that i opened for this
  purpose only.
  then from the gmail account i Auto forward the email back to other secret
  email for viewing the emails.
 
  Gmail have a very strong spam filters and the spam will not be auto
  forwarded to you.
 
  My old personal email was heavily spamed in a way that I gave up on that
  email, since i got hundreds of spams everyday but now I get only the real
  emails.
 
  This might be a good and fast solution for personal emails and private
  machines, I am not sure if it will work for a mail server that serve lot
  of people, though it might be possible with some tweaks to the DNS and
  mail server.
 
  forgive me for suggesting this (no offense) but you can also put your
  email domain under google apps
  https://www.google.com/a/

 ROTFL - What a great idea! Using gmail as a spam filter! Any idea if there
 is any restriction on using it this way? (Obviously they didn't anticipate
 this kind of usage, but I wonder if this breaks any of their published
 rules)

Either they don't add any headers to indicate their spam status, or they are 
shadowed by my own spam filters when I fetch them with POP3. If they don't, 
how do you filter your mail and check for false positives?

-- 
Sincerely Yours,
Michael Vasiliev

If you don't have time to do it right, where are you going to find the time to 
do it over?

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-11-01 Thread Amos Shapira

On 02/11/06, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:

ROTFL - What a great idea! Using gmail as a spam filter! Any idea if there is
any restriction on using it this way? (Obviously they didn't anticipate this
kind of usage, but I wonder if this breaks any of their published rules)


I have multiple gmail accounts, mostly to keep my anonymity. All of
them just have a
forward all filter to my real gmail account.

Michel should still check his spam folder from time to time if he
expects important messages,
gmail's spam filter is excellent but not 100% perfect.

--Amos

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-11-01 Thread Shachar Shemesh
[EMAIL PROTECTED] wrote:
 ROTFL - What a great idea! Using gmail as a spam filter! Any idea if
 there is
 any restriction on using it this way? (Obviously they didn't
 anticipate this
 kind of usage, but I wonder if this breaks any of their published rules)
Why should they mind?

They get deep and detailed information on all of your spam activity, and
so long as you even remotely log in to their servers, they get to cross
that with your searching, so they get to target-ad you. No, I highly
doubt they will mind.

The question is, knowing the real cost, do you?

Shachar

-- 
Shachar Shemesh
Lingnu Open Source Consulting ltd.
Have you backed up today's work? http://www.lingnu.com/backup.html


=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-10-28 Thread Oded Arbel
On Wed, 2006-10-25 at 01:33 +0200, Michael Vasiliev wrote:
 On Monday October 23 2006 20:28, Gil Freund wrote:
  Hi,
 
  I am using a Amavis+SA+CLAM for mail filtering (debian sarge packages).
  Recently I am being hit by a lot of image spam. Bayesian filtering and 
  RBL's are not enough.
 
 These mails are terrible, while my personal accounts get a few dozen each, 
 Linux-IL posting address gets a hundred per day minimum. If we only relied on 
 SA's checking, this list would look much worse than LKML. Most of these 
 emails are a crazy mix of sentences on random topics, probably web-crawler 
 generated, plus a gif image with a nasty ad of some kind (I never care enough 
 to mimedecode them).

The problem is that the image has some random noise added to it, so you
can't even get your bayesian filter to recognize the mime data. This
will also, IMO, in the long run defeat OCRs in the same way that
CAPTCHAs defeat OCRs.

 SA's Bayesian filtering is terrible, I don't know why. (All right, I admit I 
 am too lazy and illiterate at adult-level mathematics to figure out why). I'm 
 running SA and bogofilter in parallel, clearly bogofilter is at least twenty 
 times more accurate. 

I'm using bogofilter in conjunction with a large set of RBLs, and while
I do see a surge in the number of small-caps and viagra image spam, only
rarely there is a false negative - about 4-5 times a week.

--
Oded
::..
If you lost your left arm, your right arm would be left.



=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-10-25 Thread Amos Shapira
On 24/10/06, Amos Shapira [EMAIL PROTECTED] wrote:
It's funny you mention this - right after I read a whole thread about this issue in another Linux mailing list.
I haven't tried this, but people recommended the following: http://wiki.apache.org/spamassassin/FuzzyOcrPlugin
In continuation to this - I've just saw a positive report about installation of Maia, a web-based interface to this whole SpamAssassin+Amavis setup.
http://www.renaissoft.com/maia/Cheers,--Amos


Re: Image spam

2006-10-24 Thread Michael Vasiliev
On Monday October 23 2006 20:28, Gil Freund wrote:
 Hi,

 I am using a Amavis+SA+CLAM for mail filtering (debian sarge packages).
 Recently I am being hit by a lot of image spam. Bayesian filtering and 
 RBL's are not enough.

These mails are terrible, while my personal accounts get a few dozen each, 
Linux-IL posting address gets a hundred per day minimum. If we only relied on 
SA's checking, this list would look much worse than LKML. Most of these 
emails are a crazy mix of sentences on random topics, probably web-crawler 
generated, plus a gif image with a nasty ad of some kind (I never care enough 
to mimedecode them). The problem is that many of these are full of IT terms, 
some even Linux-related and making sense to a certain degree, at least to SA. 
These require me to actually read spam, and not skim through as I usually do. 

On the positive side of this annoyance is that I've developed an unusual and 
useless skill of reading through HTML pages without even noticing the tags. 
I wish they were sending spam in assembly language, preferrably of various 
arch-s, I'm becoming a bit rusty lately. :)

SA's Bayesian filtering is terrible, I don't know why. (All right, I admit I 
am too lazy and illiterate at adult-level mathematics to figure out why). I'm 
running SA and bogofilter in parallel, clearly bogofilter is at least twenty 
times more accurate. SA's learning is not really effective, you can feed in a 
dozen of spam emails following the same pattern, and seconds later you get 
another false negative, which is something completely the same. Are the 
algorithms really identical?

 I have read of an OCR option for SA 3.01, but I before going to source
 packages or moving to etch, I would like to see of there are other
 options available for the current installation.

Isn't that very, very slow and very, very inaccurate? Spamassasin is notorious 
for being a memory and CPU hog, adding another extension to fight a lost 
battle will make it even worse.

--
Sincerely Yours,
Michael Vasiliev
Linux-IL Moderator

Let me have men about me that are fat
Sleek-headed men and such as sleep o' nights
Yond Cassius has a lean and hungry look
He thinks too much: such men are dangerous.
-- William Shakespeare: Julius Caesar

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-10-23 Thread Meir Kriheli
Gil Freund wrote:
 Hi,
 
 I am using a Amavis+SA+CLAM for mail filtering (debian sarge packages).
 Recently I am being hit by a lot of image spam. Bayesian filtering and
 RBL's are not enough.
 
 I have read of an OCR option for SA 3.01, but I before going to source
 packages or moving to etch, I would like to see of there are other
 options available for the current installation.
 
 Thanks
 
 Gil
 

This article might help:

http://lwn.net/Articles/196704/

Cheers
--
Meir Kriheli

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-10-23 Thread Baruch Even
* Gil Freund [EMAIL PROTECTED] [061023 20:42]:
 Hi,
 
 I am using a Amavis+SA+CLAM for mail filtering (debian sarge packages).
 Recently I am being hit by a lot of image spam. Bayesian filtering and
 RBL's are not enough.
 
 I have read of an OCR option for SA 3.01, but I before going to source
 packages or moving to etch, I would like to see of there are other
 options available for the current installation.

For spamassassin I'm using the volatile repository which is maintained
by Debian folks though I'm not sure if it's considered part of Debian
officialy.

Cheers,
Baruch

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-10-23 Thread Gil Freund

On 10/23/06, Baruch Even [EMAIL PROTECTED] wrote:

* Gil Freund [EMAIL PROTECTED] [061023 20:42]:
 Hi,

 I am using a Amavis+SA+CLAM for mail filtering (debian sarge packages).
 Recently I am being hit by a lot of image spam. Bayesian filtering and
 RBL's are not enough.

 I have read of an OCR option for SA 3.01, but I before going to source
 packages or moving to etch, I would like to see of there are other
 options available for the current installation.

For spamassassin I'm using the volatile repository which is maintained
by Debian folks though I'm not sure if it's considered part of Debian
officialy.


Yes, it somewhat vague. I use it too. SA is in version 3.0.3. I guess
I'll give to OCR thing a try...



Cheers,
Baruch

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]





--
Gil Freund, Systems Analyst
---
Sysnet consulting
[EMAIL PROTECTED],  http://www.sysnet.co.il
voice: +972-54-2035888, Fax: +972-8-9356026

=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]



Re: Image spam

2006-10-23 Thread Amos Shapira
On 24/10/06, Gil Freund [EMAIL PROTECTED] wrote:
I have read of an OCR option for SA 3.01, but I before going to sourcepackages or moving to etch, I would like to see of there are otheroptions available for the current installation.It's funny you mention this - right after I read a whole thread about this issue in another Linux mailing list.
I haven't tried this, but people recommended the following: http://wiki.apache.org/spamassassin/FuzzyOcrPluginHTH,--Amos