mjb (02/23/83)
The article "4.2 wish list" brings up something I've always wondered about, namely, does anyone have any evidence that humans actually read bug reports mailed to !ucbvax!arpavax:4bsd-bugs? I've sent dozens of bug reports (with fixes) to them and all I get back is an obviously machine-generated acknowledgment. I have never received anything which communicates the disposition of the fix. This is extremely annoying when I get a new release and have to re-install all the bug fixes. I'd also say it's pretty anti-social on the part of the Computer Science Research Group at Berkeley. Anyhow of the bug reports I submitted one which I think people might want to know about in particular is that 4.1BSD (and to my annoyance, 4.1aBSD) does not compute the swap space map correctly at boot time. If you've ever gotten an "IO err in push" panic and couldn't trace it to a hardware error, it's because you hit a "bad spot" in the swap map. The fix I have to the problem also allows you to specify different size swapping filesystems (unlike 4BSD where they all have to be the same size). Most sites will not notice this bug because of the large (16+ MB) swap filesystems specified in the 4BSD release (the "bad spots" occur near the end of each filesystem). If anyone is interested in this fix, please let me know by mail. Mike Braca ..!decvax!brunix!mjb mjb.brown@udel-relay