> The other issue I have is that runit-init (which I've renamed to init) 
> always wants to shut down.  If I do "init 6", I expect it to reboot, 
> but it halts the system.  I don't think runit should care that 
> runit-init is now called init, but I don't know.  I did write little 
> wrapper scripts to replace the SysVinit reboot/halt scripts (so that 
> you can set a time and it sends out a wall message), but the last thing 
> called is "init 0" for halt and "init 6" for reboot but in all cases it 
> shuts down.  I don't know why though.  Any ideas on this?

I'm wondering somewhat with your stage 3 script looks like.  I've only
ever had the shutdown troubles you describe on a FreeBSD box that was
using the port, which had some somewhat goofy stage scripts.  I took the
useful bits from my Debian GNU/Linux stage 3 scripts to fix it up, and
it now shutsdown and reboots like it is supposed to.

> Beyond those two issues, using runit rocks.  I can deal with the r/o 
> filesystem issue by adding some documentation noting the fact that this 
> is a problem and that there is no real damage to anything (because it's 
> all mounted r/o), it's just ugly and not very well handled (with 
> SysVinit, this has never been a problem).  And it won't actually affect 
> me again because now I mount those drives from rc.local *after* I 
> manually modprobe sata_promise.

This sounds like you either need to place the driver within the kernel,
or place sata_promise in /etc/modules...  I'm guessing you're able to
boot this drive at all due to an initrd or similar..

> The lack of a reboot issue is big though because it makes doing remote 
> administration (ie. a reboot due to a kernel upgrade) extremely 
> problematic.

Check out your stage 3 script.  Something is wrong there, I'm thinking.
Did you install via a distribution package, or from the slashpackage?