How to increase the value of a $100 PC to $250 and make it unaffordable

[Peter P.]

I think that WinCE and other 'squeezed' versions of Windows will always be
larger than 'squeezed' versions of Linux. The original article I linked to was a
piece of tendentious FUD imho. Of course the slot in the OLPC was not put there
for M$'s benefit, it was put there to facilitate data exchange with media
devices (such as cameras and media devices) and memory extension from the
beginning, when the design was outlined. The fact that the device can run
software directly off that media is an advantage that is unrelated to M$.

And this is in fact something that M$ abhors because the ability to run software
from external media without installing it also means the ability to make an
unlimited number of illegal copies of said software. In the open software
movement, this is a given advantage. For M$, it may mean that the OLPC will be a
one-diskette installation (in the sense that they will only need to leak one
copy of a working system, the replication will be autonomous after that). 

Given the cheek they have I would not be surprised if they would later lobby to
impose a M$ tax on OLPCs since 'they could be used to run illicitly copied M$
products' (said in FUD tense in front of Congress by a concerned looking
businessman). Riight. Then, the mpaa and riaa will be able to raid preschooler's
OLPCs looking for illegally downloaded nursery rhymes and condemn the kids to
watching captain copyright cartoons for no less than an hour a day. And M$ and
the riaa/mpaa would make it mandatory (and the DHS would chip in with them) to
implement unique serial numbers for the machines so they could attribute each
illegally downloaded nursery rhyme to one particular toddler.

Someone was flogging a compressed version of Windows on a 170MB usb stick. This
simply means the usual thing: that they need three+ times the space and five+
years to do the same thing Linux has been doing for years now. Linux distros
with full GUI and X11 and a reasonable set of applications have been fitting on
50MB 'business card' cds since about 1999, and there was at least one minimal
system that ran a X server from two floppies (or was it three) before that. As
always, M$ are emulating only the best, and they have a long way to go even if
reporters pitch in heavily on their side. I am surprised that nobody reacted to
the tone of the article so far. While reporters have to eat, they also have to
sleep soundly at night, I think.

As to the power of that system, I have a i386 powered monochrome Compaq laptop
from the 1990s. It runs Linux with X11 in 8MB ram. Yes, it's slow. It's a 33MHz
386SX, see, that's why. And yes, I am still using it, mostly to drive
microprocessor programmers with its parallel and serial ports, and because it is
small and light. And yes I avoid using it for heavy things. So a 200MHz ARM
should be almost ten times faster, and a 400MHz one twenty times faster
(assuming that the bottleneck for graphics speed is the system clock - most
graphics operations in a text window environment are memory-to-memory copy
operations). Imho, the makers of certain well-known word processors and
spreadsheets (and also web applications) which make 3GHz machines appear as slow
as a PC XT running Word Perfect on a monochrome Hercules screen have the right
to shut up when quipping about the 'low resources' of the OLPC. The resources
are not low, they are deemed sufficient for the role of this machine.

I have the 'pleasure' to use the newest computers with DSL networking, routers,
ethernet etc, daily only to find it normal to wait 5-6 seconds until a web
service like 'Yahoo' sends off an email. Obviously upgrading to a 25GHz
Petapentium would not improve things. The OLPC has a place as is and I hope that
it will become common and a commodity. I have plans to use it as a hardware
terminal and control unit for embedded systems especially, running Linux and
other open source operating systems. I hope that it is 'hackable' at the
hardware port level. There are others on this list who will likely try similar
things with it. 

Then, it is likely that Mr. Negroponte did not consider the need for a stable
and available commodity laptop-sized platform for industry and government
services. That might yet prove to be the biggest cash earner for his OLPC
system. Todays wireless terminals used by meter readers, garagists, supermarket
stockists and so on are often less powerful and more expensive then the OLPC. I
feel that an open architecture industry standard system like the OLPC could open
a huge door here.

Peter


--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists