Bootloader question

[Guylhem Aznar]

Hello,

On Thu, Jun 4, 2009 at 07:18, John Gilmore <gnu at toad.com> wrote:
> Um, you're not using OLPC hardware to teach kids, but because someone
> gave it to you for free?  It's not appropriate -- an ordinary netbook

I think you deserve an explain.

My first contact with OLPC was during my PhD. I studied the various
problems arising from a computer support of written sign language. It
was actually used in southern Brasil to teach to deaf kids, and I
wanted them to have the same computer-learning opportunities than
hearing kids would have.

Unfortunately, for various reasons, I could not pursue the project
after my PhD. It did needed further research and a lot of development,
including  major software rewrites, custom font drawing, changes in
the unicode engine, etc. This means a lot of funding, and I couldn't
find any then.

For my MD thesis, I had a completely different project, basically
using mobile computing to help provide independant drug information to
healthcare professionals - for this like making better decisions,
avoiding side effects, finding alternatives to the drug based on cost,
allergies to excipients, etc. This means less funding, and most things
I could do myself.

Since I already had an OLPC and I've found some strong points with it
(cheap, resistant to shocks, outdoor capable screen etc), I reused it.

Yet the project was also on the slow side and was likely to fold,
until I moved from mainland France to Martinique, FWI. Here, my
university hospital is very supportive, and believes such projects
fits well within their agenda of exisiting cooperation efforts in the
carribean. They'd also like to deploy the computer for bedridden kids,
who do have some teacher time, but could do much more with an
educational laptop.

Unfortunately, they also do not have funding at the moment, even if
they are working hard to find it. So basically I now pay everything
for my pet project until they have such funding (EU, WHO, we are
trying in various ways...), or they decide to fund it themselves.

For the last years I've been renting a place to put servers with an
optic fiber connection, to provide the information to people with
internet connections, paid for said servers and optic connection, paid
for hardware used in field tests, paid for my trips to present the
project and try to find funding, etc. You really don't want to know
the current grand total. I tried free hosting at ibiblio.org, which is
good, but too slow. Compare www.therapeutique.info (slow) and
www2.therapeutique.info (fast).

In the end, I couldn't purchase OLPCs during the G1G1 for the pilot I
wanted to run.

So yes, shame on me, I asked for free OLPCs during the repair stock
clearance because it was a great opportunity to help a dwidnling
project and maybe resume another one, by making a case to my hospital
and thus proving the project was worth helping.

So in the next months, I will deploy these refitted OLPCs in Haiti,
loaded with drug informations, free medical books, etc.

I could have picked up Dells or whatever, but I believe the OLPCs are
more tough, best suited for the field with they display and membrane
keyboards, and more importantly can be made cheaper. X0 1.5 is also
very promising.

In the worst case, these OLPCS could be retrofitted with sign language
software and go to brasilian schools.

> would be much better -- but you're bashing it to fit.  By the time
> you're done, there will be nobody who understands how it works except
> you.

What exactly will nobody understand? Removing jffs2? A separate
partition? A debian instead of a fedora?

Anyway, that's the beauty of free software. If you are unsatisfied
with the current way, you can try a different way :-)

But I believe this new way will benefits to more people than only myself.

> (0) How does your work help OLPC reach its goals?  Or do you just want
>    us to help you, while you provide no help to us?

If you believe a faster NAND boot, to say 30 seconds, is worthless for
XO and completely outside of its goals, then yes I'm selfishly asking
for help without providing anything back.

OTOH, if you are interested in a faster boot time, there could be side
advantages.

> (1) Wouldn't the doctors be better off getting a thick printed book?

Certainly, if you can fund the printing + shipping to various
countries around the world of a book that contains drug information +
anatomy + generics drug pricing (etc, quite heavy), including the
costs of reshipping reprinting book everytime something is updated.

> If you're really set on doing this with a computer, how about a US$300
> Dell Mini 10v, with a 120GB hard drive, 1GB RAM, and modern processor?
> Or a US$330 Acer Aspire One 10.1" with 160GB and similar specs?

Or what about 3x $100 OLPC, with mesh networking to propagate the book
updates for free, and a simple USB key being mailed to people who are
in too remote areas to update their system?

With the same amount of money, I could fund 3 times more deployment,
and also make sure the laptops are not resold or stolen, being too
weird looking.

Makes more sense IMHO.

> It should be able to resume from disk-based hibernation in seconds, and
> easily be programmed to hibernate when the lid is closed.  On my Acer,
> stock Ubuntu 9.04 resumes from hibernation in about 35 seconds,
> including 2 sec of grub menu delay, and there's lots of upstream
> interest if you find ways to speed it up.  Or if you suspend to RAM,
> it wakes up in 4 seconds, but takes more power while it's closed.

I see nothing that can't be done on the OLPC, and maybe even done
better (DCON vs S2R) in some cases.

<rant>

Please, do not see anything personal this comment, but from the
beginning, since I've joined the list, I've noted something close to
arrogance from various people who were and are still sure they know
better what must be done than the people they want to help.

Why don't doctors close the lid? Many reasons, but in the end I don't
want to try (and certainly fail) to change their habit, just provide a
tool they'll be tempted to use and not keep in a closet. So if they
want a quick boot, they will get it.

Why don't I use debxo on a SD? Because the last time I tried DebXO, it
was still quite slow and got a serious keyboard problem which I
couldn't easily fix. In any case the SD is so hard to introduce/remove
from the XO without breaking either that end users should not be
expected to manipulate it - it should only be an option. I also
remember of SD corruptions problems on suspend. An SD aren't giveaways
- I have to purchase them. One is cheap, but multiply that by the
amount of OLPCs to be deployed. And they would be easily to
sell/repurpose in digital cameras, etc.

I don't want to open the "SD" can of worms unless I'm forced to.

If you are not interested in what I want to do in the carribean, don't
give a hand. It may seem bitter, but I'm no longer even expecting
help, just looking for guidance when possible.

In any case, we are on the terrain, we know what people specifically
asked for, so will do things our way and serve their needs. This means
no sugar, no python, no sd if possible, and the simplest possible web
browser there is.

</rant>

As for speedups, I see 2 different ways :
 a) using a SD with a fat partition + ext2 filesystem
 b) using the nand with a fat partition + ubifs - this requires 2.6.29
which is not ready yet.

It's hard to chose at the moment. I guess I'll stick to b) and hope
2.6.29 makes it better, and if it doesn't go for a)

Mitch says there's very little time to gain and provides an excellent
analysis. I just have a final question there : regarding the 2 seconds
SPI flash slowdown, is there a way to boot from the NAND (without
reading the full SPI) if there's a special partition at the beginning,
or if there isn't or if a special key is pressed at boottime, go back
to SPI OFW?

To sum up what I've read in this thread, what should be done in any case :
a) discarding jffs2
b) discarding initrd
c) storing the kernel uncompressed in an uncompressed small partition

And yes, this seems trivial to do - the low hanging fruit.

Personally, I will do my best to do that and improve boot times.
I hope it will be helpfull to the OLPC in general - with Sugar or without Sugar.

If it isn't, never mind. That's just on my free time. There are just
different need for different people.

-- 
Dr. Guylhem Aznar, MD PhD

Unité d'Analyse Médico-Économique
Service de Santé Publique et d'Économie de la Santé
Pôle SPSSR
CHU de Fort de France
BP 632
97261 Fort De France Cedex
Martinique, France

Tel : 05 96 55 23 47
Fax : 05 96 75 84 57