Bill Arbaugh <[EMAIL PROTECTED]> writes:

> Guys,
> 
> If you look at the list of items, it starts to resemble the functionality of a
> real OS.
> This was Ron's insight with LinuxBIOS. Unfortunately, Linux is now too big to
> fit
> 
> into most flash chips.
> 
> We tried here at UMCP using Red Hat's eCos, but it was too limited. The *BSD
> OS's are too large as well. Etherboot has been a good alternative, but as we
> look
> 
> at adding more and more. It will become OS like...something it wasn't really
> designed
> to become.
> 
> Ron played around with Plan9, and that seems to work.
> It also has much of the support asked for on the wish list already.
> 
> What are people's thoughts about Plan9?

- Plan9 isn't a widely used OS so this doesn't trivially solve the
  hardware support disparity.
- Plan9 might not go small enough.

Large scale usage depends on a strategy that is generally useful.  And for
that size is an enourmous consideration.

I will check it out, as there looks to be some potential with plan9.

A lot really depends on what it takes to deploy this stuff.

I guess we should also include in the wishlist encrypted signature checking.

My working hypothesis is that I will have to build a mini-os, to use
for bootloader purposes.  And I will measure every solution against
what it will take to build a mini-os.  The mini-os advantage is that
it can steadily improve and not be tied to other goals.

Eric

Reply via email to