Phoenix Labs - View Single Post - repackaging pg for debian

Quote:

Originally Posted by jre

First: Nice package, one step nearer to the goal to get the best from PG 1.5. As I see you used parts from an old script from me. Have you also checked these newer versions, they're more flexible and easier in handling the different blocklists to download. But at least you gave me a wonderful inspiration for my script to comply better to formal rules. And the most important thing: The blocking seems to works.

I'll have a look at the newer scripts. Thanks!

Quote:

Originally Posted by jre

etc/cron.daily/peerguardnf
Unfortunately the daily update doesn't work because your test if there is a connection to the internet always gives a negative result here. But pinging www.sourceforge.net manually works here, so it's a problem in your script.

Hmm. It works for me. At the moment I have no better idea to check for a working connection without a timeout blocking the installation... Maybe I should ping pheonixlabs...

Quote:

Originally Posted by jre

etc/PG.conf

Code:

# Please edit below if you wish not to block certain ports with
PORTS_NOT_TO_BLOCK=80,5222,5223,5190,6667,1863, 993, 110, 443;

I wouldn't unblock that many ports in a package for the public.

I'll fix this.

Quote:

Originally Posted by jre

etc/init.d/peerguardnf

Code:

DAEMON_ARGS="-d -l /var/log/peerguardnf.log -c /etc/PG.conf -m"

You start the webserver. That's ok, but it's also a small security risk.

I'll fix this, too.

Quote:

Originally Posted by jre

/var/log/peerguardnf.log
I agree the naming with PG/peerguardnf/peerguardian is messed up. But I'd recommend to use /var/log/PG.log since otherwise JFM's Gambas GUI doesn't work.

Ok. Missed this, because I did not use the gui so far...

Quote:

Originally Posted by jre

usr/sbin
Unnecessary that you create this folder.

ok. did not notice...

Quote:

Originally Posted by jre

usr/share/doc/peerguardnf/NEWS
Still refers to methlabs.org

usr/share/doc/peerguardnf/ipcopinstall.txt:
Are you sure your advise is right? See the other threads in this forum about the iptables-problems!

I haven't even read this - it was in the cvs source and it was automatically included while constructing the package. I'll remove it from the package for the time being.

Quote:

Originally Posted by jre

Some minutes later: I think I take that back, at least before I think about it again. Did you change the code from PG? Because there's also a QUEUE in the FORWARD rule.
What else did you change in the code?

jre

I changed pg to react to SIGHUP by reopening the logfile (this was necessary for log file rotation)