Weekly Edition
Return to the Press page
| |||||||||||
Laptop With a Mission Widens Its Audience (The New York Times)
David Pogue reviews
the OLPC XO laptop for the New York Times. His take is quite
favorable, giving Linux two positive stories in as many days in a pretty
high-profile publication. "The truth is, the XO laptop, now in final
testing, is absolutely amazing, and in my limited tests, a total kid
magnet. Both the hardware and the software exhibit breakthrough after breakthrough —
some of them not available on any other laptop, for $400 or
$4,000."
(Log in to post comments)
Laptop With a Mission Widens Its Audience (The New York Times) Posted Oct 5, 2007 14:24 UTC (Fri) by richo123 (guest, #24309) [Link] All very positive! Good going NYT. The Ubuntu article was great too. Pretty realistic both articles in terms of the remaining niggles and the advice to use a Livecd in the other article.
Christmas Sensation? Posted Oct 5, 2007 15:56 UTC (Fri) by hathawsh (subscriber, #11289) [Link] What are the chances this laptop will become this year's technological sensation? Every December there is at least one new gadget that millions of people want yet the manufacturer can't deliver in sufficient quantity. $400 isn't very different from the cost of today's game consoles. Buyers might choose this over a game console because they will be doing something good both for their kids and for the world.
If more articles like this go mainstream, we might see unprecedented demand for the XO. Is the XO ready for that?
(My selfish side hopes the demand doesn't surge yet, because I hope to get one for *my* kid!)
Christmas Sensation? Posted Oct 5, 2007 22:16 UTC (Fri) by dlang (subscriber, #313) [Link] no, they are not ready for demand to surge, that's why this is going to be a limited offering.
that being said, these are being manufactured by the company that makes more laptops then anyone else in the world, so if the demand is there for this limited offering you can bet that there will be further cycles of this
Christmas Sensation? Posted Oct 6, 2007 2:25 UTC (Sat) by drag (subscriber, #31333) [Link] As I understand it part of the deal with those laptop makers is that they get rights to the patented technology if they build the OX for cheap. This is just what I've heard.
So it would not be cool for the OLPC to then turn around and undermine that agreement by competing against their own benefactors in the consumer electronics arena.
I expect that in the next couple years we are going to start seeing bits and peices of OLPC technology popping up in the consumer electronics arena under numerous different products.
Sooo... personally I am looking forward to getting one of those EEEs. Not OLPC offshoot per say, but it's decendant from Intel's response to the OLPC program.
So x86 is cool (either Intel's ULV or the low-end AMD Geode computer-on-a-chip stuff), but I think that the ARM proccessor has lots of potential to and it's actually much more open... not to mention that Debian has very good support for it (considuring it's a minority arch for PC-level machines)
http://www.linuxdevices.com/news/NS8620895791.html
Freedom-compatible hardware? Posted Oct 7, 2007 12:35 UTC (Sun) by zooko (subscriber, #2589) [Link] It's interesting that you think of ARM as more "open" than x86. I kind of feel that way too, because just about every chip manufacturer on earth is an ARM licensee, which means a large ecosystem of compatible competition.
On the other hand, the ARM architecture is heavily patented. I suppose other architectures are, too, but the ARM company's whole business model is to develop new architecture features, patent them, and license them to every chip manufacturer on earth. So on the downside, you can't just start a new company and jump into the ARM game without buying a licence from the company first, but on the upside, if you do that then you can safely bequeath the designs to your children because the patents will expire twenty-one years after they were applied for. (By the way, twenty-one years ago the very first ARM processors were shipping, so already any IP present in those ones is Free.)
For x86, by contrast, there are at least three competitive companies -- Intel, AMD, and Via -- although perhaps they have all cross-licensed lots of each other's patents for all I know.
Certainly the winner of the "Freedom-compatible CPU architecture" category has to be SPARC! The CPU architecture is GPLed.
http://blogs.sun.com/webmink/entry/links_for_2007_09_26
http://swik.net/sparc+License:GPL
Of course, the technical questions of performance and power usage and so-on are a separate matter. SPARC has not historically been used much in embedded designs. Note that the Linux-based Pepper Pad started out with ARM and then switched to AMD Geode for the Pepper Pad 3. I personally attribute the necessity of that switch to their unfortunate decision to use lots of Java software.
Freedom-compatible hardware? Posted Oct 7, 2007 12:42 UTC (Sun) by zooko (subscriber, #2589) [Link] Oh, I just stumbled upon this academic project for what they call an "embedded" SPARC design:
http://www.iaik.tugraz.at/research/vlsi/01_projects/01_is...
The important thing about stuff like this is that their design is GPL'ed. In the long run, the combination of the GPL and Sun's continued efforts
http://en.wikipedia.org/wiki/UltraSPARC_T2
might lead to an ecosystem of compatible competition which is also full of Free designs.
Freedom-compatible hardware? Posted Oct 7, 2007 13:36 UTC (Sun) by drag (subscriber, #31333) [Link] Ya I suppose your right with the point up there.
After I said that I noticed that another website said that per-core the ARM is probably the most expensive to license.
Now I don't believe that patents with hardware are in the same league of 'evilness' as software patents. Two different things, not equivelent.
I just got that impression about the arm being open because everybody and their mom has their own paticular variation. (which is just terrific platform for a very portable OS like Linux)
If you think dealing with ARM is tough Posted Oct 7, 2007 12:58 UTC (Sun) by khim (subscriber, #9252) [Link] Intel's license is separate for 800Mhz bus and 1033MHz bus! Even if you have license for 800MHz bus and you hardware supports 1033MHz (typically the case) - you still need separate license. And a lot of other things are similarly licensed. Basic x86-architecture is decades old so the patents expired and AMD/Intel have cross-license agreements, but it's MUCH worse the ARM. Sparc is champion of openness, of course - but it's the case of "to late": early Sparc models were totally closed and when new, open Sparc was introduced all niches were already occupied by x86 and ARM...
Laptop With a Mission Widens Its Audience (The New York Times) Posted Oct 5, 2007 17:21 UTC (Fri) by boerner (subscriber, #4247) [Link] Another excellent article from David Pogue. Make sure you watch the video too. He has a great sense of humor.
And the best part is... Posted Oct 5, 2007 20:08 UTC (Fri) by Max.Hyre (subscriber, #1054) [Link] So the laptop uses a new battery chemistry, called lithium ferro-phosphate. It runs at one-tenth the temperature of a standard laptop battery, costs $10 to replace, and is good for 2,000 charges — versus 500 on a regular laptop battery.The heck with the laptop. For $400 I'll send both to the third world if I can get that battery.
And the best part is... Posted Oct 5, 2007 22:20 UTC (Fri) by dlang (subscriber, #313) [Link] unfortunately, on a normal laptop the batteries the size they are building for the OLPC would not last you very long.
that being said, they do have the patent on the new battery technology, and there is interest, so once they get past the initial release of the OLPC they can look at licensing deals for several of the technology advances they have made, and that can help them fund future things.
remember, the olpc orginization is only 15 people, they are limited oin how many different things they can pay attention to at any one time.
And the best part is... Posted Oct 7, 2007 14:48 UTC (Sun) by sbergman27 (subscriber, #10767) [Link] """that being said, they do have the patent on the new battery technology, and there is interest, """
Let's hear it for locking up useful technology under patents! As much as I admire what the OLPC guys are doing, I don't particularly enjoy seeing them getting a head lock on important technology which would likely have been made available by others in a short time, anyway. Patents, more correctly "limited monopolies" are for those ideas which "We, The People" would have been deprived the benefits of, for a substantial period, otherwise.
And the best part is... Posted Oct 7, 2007 18:26 UTC (Sun) by elanthis (subscriber, #6227) [Link] *bzzt*
Patents are an intentional limited monopoly which allows a company to recoup the costs of development and make profit for a short while before the technology goes on to better the human race. Most large companies would not bother investing heavily in R&D if the technology would immediately just get ripped off and sold for a cheaper price by a competitor.
Software patents are not "evil," they are just a very poor fit due to the speed of development. Software changes daily. Electronics in general change at a fast pace, faster than many other fields. I wouldn't mind patents on software or electronics if the length of the patents were reduced to be more in line with the pace of the fields. Pharmaceutical patent gets years eaten off their life due to the time they spend in testing and research before they're safe for the market, and auto and mechanical creations tend to be years in the making with patents held as early as possible. Software projects and certain kind of electronics can often be created, distributed, and end-of-lifed before the patent approval process even completes.
If software patents had a term of two years** and electronics a term of 8-10, I wouldn't see a huge problem with them in general. The current term length just doesn't make any sense for how fast these fields move, though.
** I really would prefer zero, though, since software shouldn't be patentable at all. I'd take two years as an acceptable compromise over what we have now without hesitation, though.
And the best part is... Posted Oct 8, 2007 15:36 UTC (Mon) by nix (subscriber, #2304) [Link] Software patents are also a poor fit due to extreme vagueness (i.e.they're far too vague to be *useful* to practitioners of the art) and because the complexity and interwoven nature of the underlying systems is such that it's impossible to write anything without violating multiple patents, and it's impossible to determine which patents you might be violating.
These are all problems even if the patent examination system wasn't broken
And the best part is... Posted Oct 9, 2007 1:30 UTC (Tue) by niner (subscriber, #26151) [Link] Fact: in most parts of the world software patents don't even exist. And yet these parts still have a very well working software industry. So it seems safe to assume, that software patents aren't necessary _at all_.
The "most large companies would not bother investing heavily in R&D" line is nice. It's easy to swallow and just sounds right and you here it everywhere, so it just must be true, mustn't it? But is it actually proven in any way? Has there ever been even a strong indication, that it might be true?
Companies would be really dumb to stop R&D, just because there's no patent protection. They might not have 20 years, but unless their "invention" is really trivial (in which case, it shouldn't be patentable anyway), it would take their competitors at least some time to re-engineer and copy it. So in fact, they would still have a monopoly, just one dependent on the complexity of the invention. And on top of that, they may copy their competitor's inventions, too. So I'd rather expect a much more accelerated development, instead of a stopped one.
And a last note to everyone's favourite, the pharmaceutical industry: this industry spends much more (orders of magnitude) on marketing, than on R&D. In fact, nowadays most research is done and sponsored by universities and thereby by the state. Also, most new drugs on the market, are just new combinations of old ones. The cash cows have stayed essentially the same for decades. Not really a good example of a need for patent protection at all...
And the best part is... Posted Oct 9, 2007 2:50 UTC (Tue) by dlang (subscriber, #313) [Link] you are mixing issues.
there is a fairly substantial difference between traditional patents and software patents.
you will fins many people who question the need for software patents (most people in the software field)
however traditional patents (or if you wish, hardware patents) have much more support for a number of reasons. one of which is that patents protect the little guy. once a product is invented, bugs are worked out, and demand for the product is proven, it's frequently much cheaper for an existing manufacturer in a related field to produce it then for the inventor, but the inventor has to pay all the development and marketing costs to get it to that point.
even software patents wouldn't be that big a problem if the patent office actually followed the rules in granting them. there are two big rules that the patent office has not been enforcing.
1. the invention must not be obvious to someone skilled in the field.
no need to comment on this one :-)
2. the patent must describe the invention clearly enough to be able to be implemented by someone ordinarily skilled in the field.
too many patents have vague claims that cannot be understood well enough to implement the invention, you aren't supposed to have both secrecy and patent protection.
And the best part is... Posted Oct 11, 2007 7:31 UTC (Thu) by sepreece (subscriber, #19270) [Link] I think there are some software patents that do deserve protection, but the vast majority of the ones that actually turn up in court are very bad patents - things patented long after the idea was well known or patents that are being claimed to apply to things very different from what they originally described.
The current (until the Supreme Court recently revisited it) interpretation of "obviousness" was badly broken. The rules will change now, but it's not clear exactly what they will be. Almost certainly better, though.
The term of most patents today is too long. For software patents it's way too long. If we could get fundamental change in the law, it might be interesting to have a relatively short default term for all patents, but allow the examiners to consider applicants' request to extend it up to a statutory maximum based on documentation of development costs and duration or other factors - say a five year default with a maximum of 15 years from date of issue of the patent, or something like that. Opens a huge can of worms, though, and the courts have consistently tried to make the rules narrow and reduce interpretation.
And the best part is... Posted Oct 6, 2007 0:10 UTC (Sat) by rsidd (subscriber, #2582) [Link] What on earth does "one-tenth the temperature" mean? If a regular laptop battery runs at 320 kelvin, will this one run at 32 kelvin? Temperature isn't a quantity you can divide up.
And the best part is... Posted Oct 6, 2007 1:43 UTC (Sat) by arcticwolf (guest, #8341) [Link] I'm not sure, but applying common sense, I would assume that this refers to the increase in temperature the battery's exhibiting compared to the surrounding room's temperature. So if you're in a 20°C environment and a normal laptop battery would heat up to - just to make up a random number - 40°C, this one would supposedly heat up to 22°C only under the same conditions.
Of course, I'm just guessing...
And the best part is... Posted Oct 6, 2007 4:46 UTC (Sat) by tialaramex (subscriber, #21167) [Link] Actually absolute temperatures of the sort you quote (320 kelvin) really are a quantity and you can divide them up. Specifically the absolute temperature is an aggregate measure of kinetic energy in moving particles. There's a conversion equation between temperature and energy, which I can't be bothered to look up.
However you're right that it's obviously not what these people were talking about, I guess we could hope that they mean something like "the energy wasted as heat is ten times less, proportionally" e.g. conventional battery might be providing 10W of electrical power while wasting 5W as heat while this new one would waste just 0.5W as heat for the same power output.
But I suspect that we're due for a disappointment and it's actually more like the old one provides 10W for 5W wasted while the new one provides 2W and wastes 0.5W, thus allowing uncritical journalists who don't know any better to report that it's "one-tenth the temperature" and ignore the fact that it's also only one-fifth the power output, and thus not ready for power-hungry conventional laptops.
* All the above figures are arbitrary, I have no idea whether existing batteries are 10% efficient or 90% efficient.
And the best part is... Posted Oct 6, 2007 6:52 UTC (Sat) by rsidd (subscriber, #2582) [Link] Actually absolute temperatures of the sort you quote (320 kelvin) really are a quantity and you can divide them up. Specifically the absolute temperature is an aggregate measure of kinetic energy in moving particles.It is not an "aggregate measure" of kinetic energy. What you may be thinking of is that the average kinetic energy of a particle at thermal equilibrium is proportional to the temperature. But twice as many particles don't have twice the temperature. There's a conversion equation between temperature and energy, which I can't be bothered to look up. You must be thinking of Boltzmann's constant, but it's not a conversion equation. The relationship between energy, entropy and temperature is rather subtle. Temperature is not energy, and it is not average energy per particle; roughly speaking, it measures how sensitive entropy is to changes in energy.
And the best part is... Posted Oct 6, 2007 8:11 UTC (Sat) by richo123 (guest, #24309) [Link] Yes temperature is proportional to the kinetic energy PER MOLECULE. It isn't proportional to energy because if the energy is put into spinning the molecule rather than making it move that isn't counted in the temperature.
And the best part is... Posted Oct 6, 2007 10:58 UTC (Sat) by rsidd (subscriber, #2582) [Link] Still wrong. Rotational kinetic energy certainly depends on the temperature, in the same way translational kinetic energy does, for a classical system. (That is, on average it is 1/2 kT for each degree of freedom that appears quadratically in the energy, including rotational degrees of freedom. Here k is the Boltzmann constant.)
Again, temperature is not proportional to energy. Temperature is the inverse of (dS/dE) where the derivatives are partial derivatives with other state variables constant. This happens to be proportional to the average energy for a classical ideal gas, and for other systems where the equipartition theorem applies. For other systems this does not apply. And even if the laptop battery's temperature were proportional to its energy (what energy? Energy consumption? Energy emitted? Energy stored?) what does it mean to say "it runs at a tenth the temperature"?
And the best part is... Posted Oct 6, 2007 11:14 UTC (Sat) by nix (subscriber, #2304) [Link] It's obvious what it means... you can only use it on Neptune and pointsfurther out; any further in and it overheats.
HTH, HAND, etc, &c, usw...
And the best part is... Posted Oct 6, 2007 12:32 UTC (Sat) by drag (subscriber, #31333) [Link] > what does it mean to say "it runs at a tenth the temperature"?
It says: 'I am a typo'. Or at least that is what it says to me.
It probably means it runs at a tenth of the temperature when compared to other convientional laptop batteries. That wouldn't be hard to do as other laptops typically draw much more current and require much larger batteries. The faster you draw current the hotter the batteries get (and the more expensive then need to be to manage the power draw without damage). This combined with whatever chemistry or circuity improvement the XO does then one tenth the temperature would be easy to to imagine.
:)
And the best part is... Posted Oct 6, 2007 13:25 UTC (Sat) by richo123 (guest, #24309) [Link] Sorry you are correct. My statement is true only for a monatomic ideal gas.
And the best part is... Posted Oct 7, 2007 9:39 UTC (Sun) by tialaramex (subscriber, #21167) [Link] I'd argue that it can only mean that while an ordinary laptop battery is intended for use at around room temperature (say 290-330 K) these new batteries are intended for use at around 30 K, which means you will require a source of liquid helium to cool it in a practical way. No doubt this will be very inconvenient for its users in developing nations.
That is, it's nonsense, and it's a shame that
• the writer apparently didn't know any better
And the best part is... Posted Oct 7, 2007 10:05 UTC (Sun) by khim (subscriber, #9252) [Link] Actually "practically no-one reading will think there's anything odd about it" because there are nothing odd about it. Any non-geek reader will understood that the talk is about temperature difference between ambient temperature and temperature of battery, not about temperature of battery per se. There were a lot of talks about "hot" Prescott or "cool" Core2 - and temperature there is meant in exactly the some way, so it's not something new or strange. Some guys will note that this huge difference mostly stem from low power consumption of laptop, not from superiority of the battery itself (mobile phones also have pretty coll batteries while using conventional chemistry). And this will be all... Geeks, on the other hand, will nitpick to the death...
And the best part is... Posted Oct 7, 2007 11:20 UTC (Sun) by tialaramex (subscriber, #21167) [Link] “because there are nothing odd about it”
Nothing about it /appears/ to be odd if you're scientifically illiterate, which most of the readers and practically all the writers are. We're agreed thus far. Newspapers routinely publish nonsense about science or technology without any comment. Suppose they'd made a similarly significant mistake, but in a section of the paper that some of the readers know something about, like the Sports section.
“Tiger Woods scored two bogeys in a row, picking up the points needed to win the match.”
It wouldn't take five minutes for sports fans to mob the newspaper's corrections desk. “Tiger won with two birdies, not bogeys, you imbeciles. Call yourselves a newspaper? Has the guy writing your sports columns ever played golf?” etc.
This is not about geeks, you don't need to be a technology "geek" to know that the sentence about temperature is nonsense. Remember this is not a blog, there's an editorial process that's supposed to screen out typos and stupid mistakes. But for Sci/Tech articles that process might as well not exist (sometimes it even makes things worse as a sub-editor who has never heard of the "Watt hour" changes it to "Watts per hour" for example).
The Language Log ran an article about this sort of phenomenon recently. Apparently one of the remaining unintegrated tribes doesn't have numeracy. When outsiders offer to teach it to them they can't see the benefit, even though they are traders and the inability to count must interfere constantly. The Language Log article argues that our culture as a whole has the same attitude towards useful post-numeracy skills, and our justifications are equally foolish.
And the best part is... Posted Oct 28, 2007 7:17 UTC (Sun) by kreutzm (subscriber, #4700) [Link] While we're at physics, could you quickly pass me the recipie to get *liquid* helium at 30 K? Thanks!
And the best part is... Posted Oct 28, 2007 11:15 UTC (Sun) by evgeny (subscriber, #774) [Link] In order to cool something down to 30K, one needs a cooling medium with a _lower_ temperature; so what's wrong with using liquid helium (< 4.2K) for that purpose??
And the best part is... Posted Oct 7, 2007 9:27 UTC (Sun) by tialaramex (subscriber, #21167) [Link] “twice as many particles don't have twice the temperature”
I certainly didn't intend to imply that. There are lots of aggregate functions, not only summing and in this case I was thinking of the mean (but I wasn't sure if it might not be something more subtle). Anyway my intention was only to make it clear that it is meaningful for something to be "one tenth the temperature" even though that's obviously not what they actually meant to say about this battery.
I'm afraid I was only taught the examples with ideal gases in which it looks pretty straight forward, particles go faster means higher temperature. But I'm not surprised that this all gets more complicated when your real particles disappointingly don't behave like tiny snooker balls.
And the best part is... Posted Oct 8, 2007 9:52 UTC (Mon) by salimma (subscriber, #34460) [Link] I'm assuming that by "one-tenth the temperature", they mean the heat production, thus the increase in temperature over ambient. I'm pretty sure my laptop battery runs at not much above room temperature!
And the best part is... Posted Oct 8, 2007 16:02 UTC (Mon) by dlang (subscriber, #313) [Link] actually, the battery in your laptop runs far above ambiant temperature where the reaction is taking place, the heat then spreads across the rest of the battery to be dissipated. the problem with laptops getting hot is not just the heat from the CPU, it's also the heat from the battery.
the 'exploding' laptops that have been hitting the news are where the batteries have not been able to dissipate their heat properly and ht heat has built up excessively. this new battery technology avoids this heat and so should not have this sort of problem
It's even better than that. Posted Oct 22, 2007 9:23 UTC (Mon) by hazelsct (subscriber, #3659) [Link] Lithium iron phosphate does not have any thermal runaway problems at all, so you can short the battery and drive a nail through it while it's discharging at maximum rate and it won't explode. (Don't try that with a traditional lithium cobalt/manganese oxide battery, unless you're trying to kill yourself!) Furthermore, it has higher power density, so you can safely get a lot more current out of these cells than traditional batteries, and charge them faster. That's something OLPC doesn't use, but the Prius plug-in conversion kits love it, because you can grab almost all of the braking energy even in panic stops, and accelerate quickly. The downside is that it has lower energy per unit mass and per unit volume than traditional lithium batteries, hence the "won't last long in a normal laptop" comment. Oh -- and the OLPC people don't hold the patent on this chemistry, that's held at the University of Texas (though don't tell A123 people that). OLPC may have patents on some of the surrounding technology in the laptop, I don't know.
Laptop With a Mission Widens Its Audience (The New York Times) Posted Oct 7, 2007 13:00 UTC (Sun) by zooko (subscriber, #2589) [Link] I'm extremely interested in getting an XO for my children. On the other hand, I have to weigh it against alternatives such as a Pepper Pad 3.
|
|||||||||||