To add some info... I have develop a simple test case where I post a NNTP message using Jakarta-Commons-net and I have to force a \n\n at the end of the body to read it later without problems. If I don't send "\n\n" at the end.... I have again the same problem.

I have tested "\n" and I have problems reading the post. And when finally (timeout) I read it, there is a new "." at the end.
I have tested "\n.\n" and I have also problems reading the post. And when finally (timeout) I read it, there is a new "\n.." at the end.

So the solution is to send "\n\n". And my question here is: are really Mozilla, FireBird, Jakarta Common Net wrong? or is James wrong?

I will try to read the RFCs because there is something I don't understand. If I send "\n.\n" it should work and it doesn't.

Thanks in advance,

    Chemi.

Noel J. Bergman wrote:
Guys,

If some of you would like to help contribute to the NNTP handling, that
would be great.  Please note that we stick to the RFC specifications.  In
this case, RFCs 977 and 850 apply.

According to RFC 977, section 3.10.2 (and related), the article being posted
must end with <CR-LF>.<CR-LF>.

  
some clients arnt sending the first CRLF, so when the message gets sent
back, it doesnt include the CRLF on the end?
    

What are they sending?  The bug report,
http://nagoya.apache.org/jira/browse/JAMES-311, says that they are sending
<CRLF>.<CRLF>, but not <CRLF><CRLF>.<CRLF>.  The latter is not necessary.
If the code is not working properly for the former, then it is a bug in the
code.

	--- Noel

Reply via email to