ALTqd vs PF/ALTQ

[Swift Griggs <swift.griggs%coloradovnet.com@localhost>]

	I noticed that in 4.0 PF+ALTQ is a non-starter. Pf works great 
(provided you use the lkm or compile it in), but altq rules in 
/etc/pf.conf just elicit the same old error about lack of kernel support. 
I did, of course, recompile the kernel to support all the ALTQ_* options. 
However, I did so with the suspicion that this was only for the 
gratification of the the userspace altqd. My suspicion was confirmed as 
soon as I booted the new kernel. Altqd worked, pf+altq didn't.
	 I actually have no problem with altqd whatsoever. I've managed to 
make it do everything I need. I've also got no problem using ipfilter over 
pf, either. I could simply care less about the whole Darren Reed license 
hoopla.
	If I had a fairy-wish, I guess like to see someone brighter than 
myself the do enough to the altq.conf(5) page to be able to remove the BUG 
section that reads: "This man page is incomplete.  For more information 
read the source.". I was able to glean what I needed from the source, but 
others might not fare as well. Not seeing a mention of altq in the NetBSD 
Guide makes me think there is still some question about what exactly the 
future of altq is.

	Moving right along, what is the plan for 5.0 ? Will all 
permutations be supported (pf+altq, pf+altqd, even ipf + pf-altq) ? I just 
want to know where my energy is best spent if I want to fully master altq 
syntax for the next-gen NetBSD (/etc/pf.conf or /etc/altq.conf) ?

Thanks,
  Swift
--
Swift Griggs -- <swift.griggs%coloradovnet.com@localhost>
Unix Systems Administrator