[fetchmail]Rejecting large messages when using "sendmail" for delivery

Ed W lists@wildgooses.com
Tue, 28 Jun 2005 23:51:22 +0100


This is a multi-part message in MIME format.
--------------090300040505000203000302
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit


>No, not return codes, but MTA error codes.  You need to be passing the
>email via SMTP.
>  
>

?? Ehh? My first post made it clear that I wasn't passing email via SMTP 
because I couldn't get the behaviour I was after (easily).  You then 
said I could check the spam return codes?  I think my confusion is 
obvious at this point?


>> The desired behaviour here is to bounce the large message...
>>    
>>
>
>Which you can't do.  You can either keep repolling it, or delete it. 
>  
>

How do I delete it (safely) bearing in mind some users may be marking 
"option keep" for some servers?  I sense that this isn't the correct 
solution for me?

>Or, you can either increase the local limit to match the server you're
>pulling from (or talk to them about lowering theirs).
>  
>

Hmm, difficult to see how I could ring up huge ISPs and persuade them 
that 20Mb was the correct limit for their servers.  I doubt this is a 
workable suggestion in practice.


>Well, you could always provide the project with a patch :-)  All I can
>do is tell you that fetchmail doesn't support the behaviour you're
>looking for.
>  
>

Thanks, writing the code would of course probably be trivial. 

We seem to be talking cross purposes though - at this stage I'm trying 
to configure a complicated tool and have the usual pressures of time.  I 
was *hoping* just to get a few simple answers as to current capabilities 
before I started cracking open code.  Perhaps I phrased my original 
question badly but I don't feel that I obtained this simple answer very 
easily.

Thanks for taking the time (in the end) to clarify the current fetchmail 
behaviour.  Much appreciated

Ed W

--------------090300040505000203000302
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
  <title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
<br>
<blockquote cite="mid43ea8d070506281516376fb02a@mail.gmail.com"
 type="cite">
  <pre wrap="">No, not return codes, but MTA error codes.  You need to be passing the
email via SMTP.
  </pre>
</blockquote>
<br>
?? Ehh? My first post made it clear that I wasn't passing email via
SMTP because I couldn't get the behaviour I was after (easily).&nbsp; You
then said I could check the spam return codes?&nbsp; I think my confusion is
obvious at this point?<br>
<br>
<br>
<blockquote cite="mid43ea8d070506281516376fb02a@mail.gmail.com"
 type="cite">
  <blockquote type="cite">
    <pre wrap="">
 The desired behaviour here is to bounce the large message...
    </pre>
  </blockquote>
  <pre wrap=""><!---->
Which you can't do.  You can either keep repolling it, or delete it. 
  </pre>
</blockquote>
<br>
How do I delete it (safely) bearing in mind some users may be marking
"option keep" for some servers?&nbsp; I sense that this isn't the correct
solution for me?<br>
<br>
<blockquote cite="mid43ea8d070506281516376fb02a@mail.gmail.com"
 type="cite">
  <pre wrap="">Or, you can either increase the local limit to match the server you're
pulling from (or talk to them about lowering theirs).
  </pre>
</blockquote>
<br>
Hmm, difficult to see how I could ring up huge ISPs and persuade them
that 20Mb was the correct limit for their servers.&nbsp; I doubt this is a
workable suggestion in practice.<br>
<br>
<br>
<blockquote cite="mid43ea8d070506281516376fb02a@mail.gmail.com"
 type="cite">
  <pre wrap="">Well, you could always provide the project with a patch :-)  All I can
do is tell you that fetchmail doesn't support the behaviour you're
looking for.
  </pre>
</blockquote>
<br>
Thanks, writing the code would of course probably be trivial.&nbsp; <br>
<br>
We seem to be talking cross purposes though - at this stage I'm trying
to configure a complicated tool and have the usual pressures of time.&nbsp;
I was *hoping* just to get a few simple answers as to current
capabilities before I started cracking open code.&nbsp; Perhaps I phrased my
original question badly but I don't feel that I obtained this simple
answer very easily.<br>
<br>
Thanks for taking the time (in the end) to clarify the current
fetchmail behaviour.&nbsp; Much appreciated<br>
<br>
Ed W<br>
</body>
</html>

--------------090300040505000203000302--