[OLPC-devel] Re: Boot tasks -- the iPAQ way

[Zephaniah E. Hull]

On Thu, Jul 20, 2006 at 02:10:27PM -0400, Joshua Wise wrote:
> On Thu, 20 Jul 2006, Jordan Crouse wrote:
> >On 20/07/06 11:16 -0400, Jim Gettys wrote:
> >>Jordan,
> >>
> >>Some history may be useful: Joshua Wise was a significant contributor to
> >>Linux on iPAQ at age 12, before he started high school. (Joshua, your
> >>ears should start burning...).  He taught me to not under-estimate what
> >>kids can do....
> 
> At every opportunity, Jim... :)
> 
> >Point taken. :)
> >
> >I guess my reasoning wasn't so much that kids *can't* handle a shell, more 
> >that
> >they shouldn't need to if they don't want to.  But your argument is
> >valid - enough to convince me that we shouldn't get rid of /bin/ash quite
> >yet.
> 
> For what it's worth, LAB does have a shell environment (in fact, that was
> its original environment, and arguably the only useful one currently).
> There's all the power still there, just tailored to the bootloader's
> environment (labcopy is a good example of this: copy ymodem: nand:0,raw)

How does LAB, with a shell environment and all, compare boot speed and
size wise with some of the current initramfs work?

I have not looked at the size usage for a tailored shared library and
binaries compiled against it for klibc, but given the size of the
library and binaries compiled staticly against it, and the flexibility
it would give us for it to all be userspace and yet still very tiny...

(By tailored, I mean what would probably be a two stage run of building
a full featured klibc, compiling the tools we want/need against it, then
seeing what parts of it are actually used and building a klibc with all
but those parts ripped out.  Given how the compiler reduces things when
compiling static binaries, I suspect that much of this could be
automated if someone were to spend the time on it.  But it may not be
worth the effort for OLPC to do it.)

Zephaniah E. Hull.

-- 
	  1024D/E65A7801 Zephaniah E. Hull <warp at aehallh.com>
	   92ED 94E4 B1E6 3624 226D  5727 4453 008B E65A 7801
	    CCs of replies from mailing lists are requested.

A little knowledge is a dangerous thing.
So is a lot.
  -- Albert Einstein.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.laptop.org/pipermail/devel/attachments/20060720/2430f170/attachment.sig>