How do fix a good solution against spam..

[Anton Alin-Adrian]

Giorgos Keramidas wrote:
> On 2004-05-16 01:46, Anton Alin-Adrian <aanton at reversedhell.net> wrote:
> 
>>M.Jessa> Not only it's way faster than perl based messagewall, amavisd and
>>M.Jessa> mailscanner etc but it also has neat stuff like making connections
>>M.Jessa> back to the sender's MX checking for validity of the sender's
>>M.Jessa> email.
>>
>>So far I can only release this code. It implements exactly what was
>>mentioned about exim. I use it with qmail because qmail I have, but
>>can be used with postfix/sendmail with ease. So now not only exim can
>>do that hack.
> 
> 
> Similar functionality to the one described above can be added to
> Sendmail with a milter.
> 
> Anyway, you surely can't be using the program you sent.  It doesn't even
> build here:
> 
>     giorgos at gothmog[11:31]/tmp/mxcheck$ cc -O2 -ggdb check.c
>     check.c: In function `filtervalidmail':
>     check.c:443: error: syntax error at end of input
>     giorgos at gothmog[11:31]/tmp/mxcheck$ _
> 
> 
>>PS - this is how i use it:
>>.qmail-file:
>>
>>| /usr/local/bin/check /usr/local/bin/safecat /path/to/Maildir/tmp
>>/path/to/Maildir/new
>>
>>#the above after | is on a single line.
> 
> 
> Putting aside the facts that the source is entirely undocumented, that
> it doesn't even build, that it has a million style bugs, that the
> comments aren't really helpful in understanding how it works, and that
> it's entirely undocumented...
> 
> - What does each one of these parameters exactly do?
> - What is safecat and why is it run with a full pathname?
> - Why are you using an internal Maildir/ path like 'tmp'?
> - Does this work in a .forward file too?
> 
> I know, I know that RTFS for such a small program documentation is most
> of the time unnecessary for the experienced hacker, but IMHO this is
> something that simple FreeBSD users might want to install too.  Without
> any sort of documentation or other hint about the way it works, you
> don't really expect anyone to run this or do you?  :-/
> 
> - Giorgos
> 
> 
> 

---dump---
%ls -l
total 10
-rw-------  1 bu  bu  10180 May 16 16:57 lacheck.tar.gz
%tar -zxvf *
gpl.txt
check.c
%ls -l
total 38
-rw-r--r--  1 bu  bu   9626 May 16 02:49 check.c
-rw-r--r--  1 bu  bu  18009 May 16 02:49 gpl.txt
-rw-------  1 bu  bu  10180 May 16 16:57 lacheck.tar.gz
%cc -o check check.c
%ls -l
total 50
-rwxr-xr-x  1 bu  bu  11518 May 16 16:58 check
-rw-r--r--  1 bu  bu   9626 May 16 02:49 check.c
-rw-r--r--  1 bu  bu  18009 May 16 02:49 gpl.txt
-rw-------  1 bu  bu  10180 May 16 16:57 lacheck.tar.gz
%
---dump---

It builds fine on both FreeBSD 4.x and FreeBSD 5.x.

You said it has millions of style bugs. Indeed. It is *not a program*, it 
is *snippet* who's functions are going to be used inside a large-scale 
anti-spam project, placed in ANSI-C header files and modularized. Obviously 
there's no doc for code snippet as this is not intended to be a 'real 
public release' for God's sake, it's *snipper*, I just thought it may be 
usefull for someone who knows how to use code.


As for docs, well yes, it's gonna have docs, but I doubt I'll write docs 
for a snippet till I add some more code and clean it up.As you can easely 
see the homepage of the project has no code released.

And yes, the snippet is for the hackers. Though I don't think one has to be 
a hacker to use it, if he wants. Can be adjusted to anything he/she likes.

The comments around the functions are written at different times, for 
private-circle purposes, and they served well.

-- 
Alin-Adrian Anton
Reversed Hell Networks
GPG keyID 0x1E2FFF2E (2963 0C11 1AF1 96F6 0030 6EE9 D323 639D 1E2F FF2E)
gpg --keyserver pgp.mit.edu --recv-keys 1E2FFF2E