- There's no example or explanation on how to use installboot on old 
world machines and/or that it's not needed for new world machines in 
the boot or installboot man pages for 2.0 and current.

- Certain key combos on some models of {Power/i}book cause the system 
to power off, not sure if this is still true or not.

- Running 2.99.10 patched kernel I decided I'd try to see what happens 
under low memory, high memory pressure on my PDQ. I have 256MB of real 
ram and a ~300MB swap partition. So I made a 480MB memory disk and 
started copying usr into it. I saw two messages from copy : cannot 
allocate memory. And then nothing. No thrashing, no hard disk activity. 
So I hopped into ddb, I'm not sure if what I did was just dumb and to 
be expected behavior or indication of a lurking bug.

Stopped in pid 11.1 (pagedaemon) at netbsd:cpu_Debugger+0x10: lwz r0, 
r1, 0x14
db> bt
0x0067cd40: at internal_command+0x1a8
0x0067cd60: at wskbd_translate+0x8c
0x0067cdb0: at wskbd_input+0x18c
0x0067cdf0: at kbd_passup+0x58
0x0067ce00: at kbd_intr+0x40
0x0067ce20: at kbd_processevent+0x44
0x0067ce70: at kbd_adbcomplete+0x58
0x0067cec0: at adb_soft_intr+0xe8
0x0067cf00: at softclock+0x248
0x0067cf40: at hardclock+0x314
0x0067cf70: at decr_intr+0xfc
0x0067cfa0: at trapstart+0xbd0
0xd520be10: at uvmpd_scan_inactive+0x360
0xd520bed0: at uvmpd_scan+0x7c
0xd520bf10: at uvm_pageout+0x114
0xd520bf40: at cpu_switchto+0x44
saved LR(0x9f2f0f6b) is invalid.
db> sync
syncing disks... 24 (hang)