Alexander Prinsier wrote:
Hugo Monteiro wrote:
Alexander Prinsier wrote:
Hello,
As suggested I tried the latest cvs snapshot. Here are some thoughts
on it:
-Can ldap be made optional through the configure script?
--enable-ldap was substituted by the --enable-external-lookup directive.
Besides the built-in LDAP lookup you can also choose a more generic user
lookup method by using external scripting. For configuration, look for
the ExtLookup* directives in dspam.conf.
Oh! that's a great new feature. Looking forward for a built-in mysql
lookup method ;)
But when I don't use --enable-external-lookup, then ldap shouldn't be
compiled in right? If that's the case it's not working :)
It's no real problem for me though, I just installed the ldap headers.
That's weird ... shouldn't need those if you're not using that
directive. I'll try to look at those #ifdef's. What's exactly the error
that you get when you're building?
-I previously was using the dspam package from debian stable. It has
some patches like the 'Include' directive. Obviously that wasn't working
anymore :) Any reason why these patches don't get integrated?
From what i recall, the actual code gets properly patched with those
patches. Problem is with the patch for the dspam.conf file. I have
rolled out my own debs with the cvs version. You might want to try them.
What's the problem with the patch for the dspam.conf file?
Nothing much. Apparently the original dspam.conf file changed a bit, and
patch can't find the right context to apply the diff. In the package
source, in the debian/patches directory (in my repository) you can find
those individual patches that i think apply cleanly to the original.
Drop me a line if you need help with those.
check them here:
http://hmonteiro.net/repository:debian
-I only have one user configured, but that user's 'False Positives' went
up quite alot suddenly. I'm sure it wasn't this high 2 days ago. Could
this upgrade have anything to do with it? (Ham strike rate went from
0.4% to 8.3%).
Actually i also found a slight increase of false positives right after i
upgraded, but still not as high as you report.
I've retrained those messages and the number of false positives has
droped to the small volume i had before.
Something mysterious is definitely happening here. Before this release
goes out as stable, this should be mentioned somewhere. You can imagine
that your users start to freak when they suddenly see their efficiency
drop alot :)
Alexander
Actually i'm not 100% sure it had something to do with it. I also made
some changes on the global group definitions i'm using. Hopefully
someone else trying cvs can verify that behaviour. I also think more
people should try out cvs. Maybe that way we could give a boost to the
new stable release.
Regards,
Hugo Monteiro.
--
ci.fct.unl.pt:~# cat .signature
Hugo Monteiro
Email : [EMAIL PROTECTED]
Telefone : +351 212948300 Ext.15307
Centro de Informática
Faculdade de Ciências e Tecnologia da
Universidade Nova de Lisboa
Quinta da Torre 2829-516 Caparica Portugal
Telefone: +351 212948596 Fax: +351 212948548
www.ci.fct.unl.pt [EMAIL PROTECTED]
ci.fct.unl.pt:~# _
!DSPAM:1011,48381345150926724514519!
