http://bugzilla.spamassassin.org/show_bug.cgi?id=3997
------- Additional Comments From [EMAIL PROTECTED] 2005-04-13 10:01 ------- 'I just looked at the source code in for Net::Dns::Header. It not only calls nextid() each time to increment the ID, but it initializes it to a random value when it loads the class. We don't need any of that $DNS_ID_COUNTER code in the patch!' that's only true if we can *get* that packet's ID after it's been created, so we know what ID it had. creating the packet is only half the story. however I notice that the API Matt found does allow the caller to get that... hmm. 'does anything like this have to be done in the DCC plugin?' I don't think it uses Net::DNS, and afaik it just calls into DCC code which can do this stuff for itself. regarding Matt's single-socket code -- this bug is definitely not the right place to put that, it should be in a separate bug that this depends on. reasons: 1. the code we have here may fix the problem 2. the current Dns.pm code would need some pretty extensive rewriting to work in the event-driven style that code needs, and 3. aren't we supposed to be finishing up for a 3.1.0 release? ;) having said that, it does look like a cleaner and more efficient fix for this problem, AND the problem of Net::DNS creating hundreds of fds. I may have a try at implementing that (and beating Dns.pm into something nearer real OO ;). ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee.
