MIT $100 laptop project

Forum thread started by DLazlo on Fri, 2005-11-18 04:31

I'm not sure where the specs for it are yet, or when it will be availible, but if it's anywhere near compatible with Haiku, I will pledge the cost of the first as soon as they become availible to be donated to the Haiku development team.

This could be an enourmously important project worldwide, and the fact that it will be running Linux leaves a very nice gap for Haiku to offer it's self up as a replacement system possibly. I have no way of knowing yet what difficulties will be involved, but at least in this way I know that we can evaluate it as soon as possible.

I will be looking into aquirering one, or if someone thinks they can speed the process of getting one for the Haiku team, please email me.

Cale Lewis

note: It is a 500 MHz processor, so it will have plenty of power in that respect.

Comments

Re: MIT $100 laptop project

DLazlo wrote:
note: It is a 500 MHz processor, so it will have plenty of power in that respect.

I wouldn't be so sure of that...

My 800mhz Via C3 Ezra machine runs about on par with my P200MMX as far as performance...

as proven many times over, clock speed doesn't mean much.

I want to know what the specs are on the machine also - a wind-up-powered linux-compatible laptop with an LCD screen just screams "handy-consumer-device" -- i want one for my kids (heh)

Re: MIT $100 laptop project

DLazlo wrote:
This could be an enourmously important project worldwide, and the fact that it will be running Linux leaves a very nice gap for Haiku to offer it's self up as a replacement system possibly.

Yeah. There will almost certainly be open source drivers for all of its components, since it's meant to run Linux. It shouldn't be overly difficult to get working Haiku drivers.

And it'd be nice to give the children using the laptops (they are meant to be used by schoolchildren in third world countries) a choice of operating systems. Perhaps even worm our way into getting Haiku as a default operating system on a model of this sometime in the future. Haiku, after all, should have BeOS' ease of use, which would probably make it better suited than Linux, once it becomes a capable enough OS.

However! The laptop is not going to be for sale to the general public. They are being purchased directly by the governments of the countries in question, for distribution to schoolchildren. It may be very difficult to get one, like you suggest. (Especially difficult to get one legally.)

umccullough wrote:
My 800mhz Via C3 Ezra machine runs about on par with my P200MMX as far as performance...

For what it's worth, the computer I used BeOS on back when it was my main operating system was a 350 MHz AMD K6-2. It's hard to say, though. I haven't found any official information, but I've heard that the processor it's using is an AMD Alchemy. This is a MIPS32 chip, which would needs mean an entire port to this architecture is necessary. Quite a task, but by no means impossible if we can find motivated programmers with access to MIPS32 hardware.

The operating system would probably require a large amount of stripping down to run reasonably. The laptop is running a "skinny" version of Linux, as it is. Hard to say how reasonable this is for Haiku.

Re: MIT $100 laptop project

Benjamin Mullins wrote:
For what it's worth, the computer I used BeOS on back when it was my main operating system was a 350 MHz AMD K6-2. It's hard to say, though.

BeOS ran extremely well on my PII 350 also :) - my point was more along the lines of: clock speed doesn't count if it's a crappy processor... being an AMD, it may not be too bad though right?

Quote:
I haven't found any official information, but I've heard that the processor it's using is an AMD Alchemy. This is a MIPS32 chip, which would needs mean an entire port to this architecture is necessary. Quite a task, but by no means impossible if we can find motivated programmers with access to MIPS32 hardware.

Well, theoretically, if the laptops end up becoming available to "general public" - it shouldn't be hard to get them into developers hands...i mean - $100 is small beans to donate for a porting project ;) And considering they're already running linux, there's probably already a GCC compiler that generates the code for the processor, it's just a matter of retooling the kernel, and porting the drivers in theory (I know it's still a lot of work, but not like building something from the ground up with no specs right?)

Quote:
The operating system would probably require a large amount of stripping down to run reasonably. The laptop is running a "skinny" version of Linux, as it is. Hard to say how reasonable this is for Haiku.

I would say that Haiku probably has the potential to be "skinnier" than Linux for the most part - the architecture is modern and modular. There aren't enough dependencies to create problems. The GUI architecture is not a HUGE behemoth like it is in Linux... I'd say this is exactly where Haiku has great potential!

Re: MIT $100 laptop project

umccullough wrote:
BeOS ran extremely well on my PII 350 also :) - my point was more along the lines of: clock speed doesn't count if it's a crappy processor... being an AMD, it may not be too bad though right?

Oh, certainly. I'm not entirely sure why I made that comment, since I do understand that processor frequency is not the end-all measure of speed. But we can hope that it's beefy enough. After all, it does apparently run Linux decently enough.

umccullough wrote:
Well, theoretically, if the laptops end up becoming available to "general public" - it shouldn't be hard to get them into developers hands...i mean - $100 is small beans to donate for a porting project ;)

I think the best way to get a laptop (or a few) would probably be to make a request to the folks at MIT developing them for a few units for software development purposes. I'm not sure how likely they are to ever become available for sale, especially in western countries where most people would be turned off by a laptop that can't run Windows (MIPS strikes again).

Of course, ports to other architectures is really something to start thinking about after it works properly on x86. :p

Re: MIT $100 laptop project

Benjamin Mullins wrote:
After all, it does apparently run Linux decently enough.

Heh, that's what I thought when I bought this Via C3 Ezra machine -- it came with ThizLinux pre-installed.. wow, that was the absolute slowest thing I've ever used. I couldn't believe they actually shipped a machine in that state.

Quote:
I'm not sure how likely they are to ever become available for sale, especially in western countries where most people would be turned off by a laptop that can't run Windows (MIPS strikes again).

I wouldn't be so sure... $100 = consumer device to most americans... a decent embedded OS could probably go over well... Cell phones, PDAs, MP3 players, etc are all examples of non-windows running devices with somewhat-advanced embedded OSes (I realize PDAs mostly run Windows CE now, however). With a decent selection of free software already available for these laptops, I would expect people to jump all over them for that price.... maybe.

Anyone know the physical dimensions and weight of these things? It just occurred to me that I don't know how big they are... they appear to be smaller than most laptops currently on the market... ?

Re: MIT $100 laptop project

umccullough wrote:
I wouldn't be so sure... $100 = consumer device to most americans... a decent embedded OS could probably go over well... Cell phones, PDAs, MP3 players, etc are all examples of non-windows running devices with somewhat-advanced embedded OSes (I realize PDAs mostly run Windows CE now, however).

Hmm. That's a good point. I hadn't really thought of it as anything besides a normal (if somewhat simple) laptop.

umccullough wrote:
Anyone know the physical dimensions and weight of these things? It just occurred to me that I don't know how big they are... they appear to be smaller than most laptops currently on the market... ?

Quite small indeed: http://laptop.media.mit.edu/images/laptop-handside.jpg

processor

AMD now owns both the Geode (x86 based) and the Alchemy (MIPS based). As to which is going to be used I don't know. I do know that much more work has been done on the Geode, especially on low-power versions than I've been able to find on the Alchemy.

Re: MIT $100 laptop project

Benjamin Mullins wrote:
umccullough wrote:
I wouldn't be so sure... $100 = consumer device to most americans... a decent embedded OS could probably go over well... Cell phones, PDAs, MP3 players, etc are all examples of non-windows running devices with somewhat-advanced embedded OSes (I realize PDAs mostly run Windows CE now, however).

Hmm. That's a good point. I hadn't really thought of it as anything besides a normal (if somewhat simple) laptop.

Yes in fact, I forgot to list game consoles... PSP and Gameboy DS come to mind. The pic with the kid's hands on the laptop definitely shows how small it is... I think it looks like a very "palm-top"-sized machine.

MIT contact

I have contacted someone at MIT who is checking on the info I need to hopefully get a laptop for compatiblity testing as the project advances. I will relay all news I get on to the Haiku team so that they may decide how far they wish to become involved with it.

Re: MIT contact

DLazlo wrote:
I have contacted someone at MIT who is checking on the info I need to hopefully get a laptop for compatiblity testing as the project advances. I will relay all news I get on to the Haiku team so that they may decide how far they wish to become involved with it.

Cool, I'm definitely interested in the vital specs like video/audio/network chipsets/functionality, etc.

Also, I'm not sure if i've seen any postings that specify how much RAM the machine comes with - only that it has 1 gb flash storage.

MIT $100 laptop project

Here are some specs:

- 500Mhz AMD processor
- 1GB flash memory (no hard drive);
- SVGA 8" diagonal display (dual LCD Color/Black & White mode for power conservation and outdoor reading);
- 128MB of DRAM;
- AC Cord that doubles as carrying strap as the power source and a hand-crank (one minute of cranking gives enough power for 10 minutes of operation).

Came from this article:

http://www.tectonic.co.za/viewr.php?id=731

I'm curious as to how the MIT people will react to the proposal. Given that most of these are going to be going to kids in Africa who may have never seen a computer before-- will they really want to offer two versions with two completely different operating systems? Still I think its worth looking into-- fun if nothing else.

MIT toy

What would the requirements be in power management?

Also I heard the flash drive could become a problem. While flash drives are useful for occasionally storage, you can easily reach the read/write limitation on them when running an OS off of them.

Re: MIT toy

Katisu wrote:
Also I heard the flash drive could become a problem. While flash drives are useful for occasionally storage, you can easily reach the read/write limitation on them when running an OS off of them.

Yes, at least that's how I always heard it... each bit can only be flipped so many times... of course, that isn't necessarily a problem unless a file is being modified repeatedly.

I'm curious though - with "only" 128MB RAM - it seems like a linux swapfile on a 1gb flash might be a bad thing - I wonder if they intend to disable swapfile entirely which could really limit the complexity of the applications that can be run...

laptop offerings and swap file

My intention (and I said so to MIT) was not that Haiku be offered on the laptop as planned*, but that we be allowed to do some work and testing to try to get Haiku running well on the laptop "if" there is not too much of a mismatch in the hardware used and it would be too much of a drain on the main focus of Haiku. Then let them take a look at it and consider it as a 'possible' option for them.

I feel that there might be areas of use that we might have the edge on an embedded Linux, and it could get us some media and public recognition as well as helping to set up future oppourtunities for Haiku to be used. Unless it's a design with absolutely no thought to the future, it should not be too dificult to switch Linux and Haiku or any other compatible system out on these laptops.

I don't imagine they'll be using a swapfile, just like embedded Linux, BeIA, or most any other embedded system does that I know of.

*Unless they should decide that, after all, it will be a while before it actually goes into production. ;-)

MIT $100 laptop project

hi, the idea sounds good! it's also called a green machine (because of it's colour). i've seen it a couple weeks ago in a bbc rss feed.

windows CE is a pig. i think linux is the best choice for such a device.

latest news

Here's what I've got in the way of 'official' response so far:
=========================================
We appreciate your interest in the $100 Laptop initiative. Please know we
receive more than 100 emails per day regarding the project. We are a small
team working on making this project a reality and unfortunately we are
unable to answer each e-mail personally.

Please
note that the $100 laptops--not yet in production--will not be available
for sale. The laptops will only be distributed to schools directly through
large government initiatives.

The Massachusetts Institute of Technology (MIT) and the One Laptop per
Child (OLPC) non-profit association have launched a global program to equip
hundreds of millions of students and their teachers with ultra-low-cost,
individual, connected laptop computers to dramatically enhance the
children’s primary and secondary education. These will be fully powered,
general-purpose laptops, sized for children and adolescents, running Open
Source software, with wireless mesh-network connectivity.

The machines will be built and sold at cost to ministries of education,
with a floating price targeted at $100 and thereafter floating down.
According to the present schedule, integrated prototypes of the Gen-1
laptop will be available Q3, ’06. Manufacturing begins Q4. The first 10 to
15 million Gen-1 pilot-phase machines will ship by or before Q1, ’07, to
five or six large, geographically and culturally diverse countries. Each of
these nations will identify three or more regions within their country
emphasizing, when possible, rural and remote areas where all students in
all primary and secondary grades receive a personal and connected machine:
One Laptop per Child.

An essential feature of the pilot program will be training, logistics and
an administrative initiative based in the host country. Here, in
partnership with local educational organizations or other groups to be
determined, MIT will help create the means to implement a carefully
designed, exponential process to impart the necessary technological and
pedagogical skills to classroom teachers over a period of approximately six
months.

OLPC technology incorporates design and engineering advances developed at
MIT's Media Lab, plus innovations in manufacturing and distribution to
bring costs in line with the host countries’ budget structures. The
laptops will be made available only to schoolchildren and their teachers,
through government programs and not through retail or commercial channels.
In parallel, OLPC is in discussion with a handful of well known large and
global companies to release a commercial version circa $200 at the same
time.

Hardware

Its hardware specifications, as of October 2005, are:
500 MHz CPU (Processor) by AMD (Advanced Micro Devices) Inc.
1 GB of flash memory (no hard drive)
SVGA 8" diagonal display (dual LCD Color/Black & White mode for power
conservation and outdoor reading); estimated cost $35
128 MB of DRAM
AC Cord that doubles as carrying strap as the power source and a
hand-crank (one minute of cranking gives enough power for 10 minutes
of operating).

The cost of materials and assembly for each laptop is estimated to be
approximately $100.

Software

All of the software on the $100 Laptop will be open-source. The projected
software as of November 2005 is:
Red Hat Linux variant as operating system
A web browser
A word processor
An email program
A programming system

Participating countries

Brazil, Thailand, Egypt, United States (Massachusetts), Cambodia, Costa
Rica and Tunisia have already committed in various ways. The laptops will
be sold only to governments, to be distributed through the ministries of
education willing to adopt the policy of “one laptop per child”.

We will begin with five pilot countries and one U.S. state. Once this
occurs, we hope to extend our reach to other nations.

Prototype

Nicholas Negroponte is scheduled to demo a working prototype of the device
on November 16, 2005 at a United Nations technology conference in Tunisia.
However, the device shown will be a rough estimation, as work is still
being done on the development of the device for both performance and
pricing issues. Negroponte estimated that the screen alone required three
more months of development.
==========================================

I'll add the 'unofficial' info I've heard when I get OK from it's source but don't hold your breath. It seems there's still quite a bit of details that are still up in the air on this one. But in the long run I suppose the longer it takes them, the better we get a chance to look as a valuable add-on for many end users.

MIT $100 laptop project

i read it was $200 for the public ...

MIT $100 laptop project

other extra trivia:

"Each laptop will act as a node in a mesh peer-to-peer ad hoc network"

"5 million to 15 million laptops for those markets towards the end of 2006. Come December 2007..."

"The laptop can be used in a variety of ways as a computer, an electronic book, a television and a writing or drawing tablet"

"The laptops will be ruggedized and probably made of rubber, he said. " (i hope they stick to green :D )

"four USB (Universal Serial Bus) ports, be Wi-Fi- and cell phone enabled and come with 1GB of memory." (and that's the whole drive i think.)

DLazlo wrote:
500 MHz CPU (Processor) by AMD (Advanced Micro Devices) Inc.

it might be geode which is x86 with 0.9 watts power consumption. and some people say it can run 1.5 "geode" ghz