Slashdot has had a number of articles about the iTron OS that's being used in a lot of Japanese autos these days. That and linux are both good candidates for a stable, reliable onboard OS. Of course, behind the scenes, iTron and linux are busy "stealing" ideas from each other. It'll be interesting to see which will win in the long run, or if they'll slowly merge. Or maybe divide the territory by consciously heading toward different parts of the tradeoff space.
I wonder if anyone is looking at any of the *BSD kernels for this sort of embedded use? They have reliability records comparable to iTron and linux, and also come with all the source code.
Re:except the BMW 7-series is practically undrivea
on
If Microsoft Built Cars...
·
· Score: 4, Informative
Consumer Reports had an article recently that mentioned this general topic. Their general suggestion was to avoid buying the new computerized luxury cars for a few years, until they get the UI right. It seems that in general their testers couldn't make much sense out of the menu-based centralized controls, and they considered these controls to be hazardous in the extreme under normal driving conditions.
BMW's newer models were not nearly the only bad examples.
Trying to discover where they've hidden some control in an N-level-deep menu tree is extremely distracting. You don't want this when you're driving.
Of course, if you screw up, the manufacturer will just call it an "operator error".
Then never complain... when your stuff gets downloaded.
But I have a very good complaint: My web site has my music on it. If this goes through, any Canadian downloading my music from my web site will be paying a tax to the recording industry. So, while I won't get any income from those downloads, someone else with no rights to my music will.
It's bad enough that the recording industry can force "standard" contracts on musicians that give all rights and profits to the recording company, and claim that this is "voluntary". Yeah; it's voluntary; you always have had the choice of nobody hearing your music because you can't get it distributed without signing one of these contracts.
But this sort of tax gives them profit from my music when I haven't signed any contract at all.
Microsoft OLE DB Provider for ODBC Drivers error '80040e57'...
It isn't obvious what this means, but they appear to be claiming that there is some Microsoft software running on port 80 on your machine, not linux or apache.
So I tried my home machine, which is RH linux with apache, and got:
Problem communicating with remote server: [my hostname]...
So in my case, my OS/server simply isn't being counted.
These two samples don't really impress me with their competence. But then, they do state right at the top that they are only counting the "top commercial" web sites. So their results aren't in conflict with netcraft, because they're counting different things.
Hey, I run the linux box that is the firewall between our home LAN and the cable modem. I provide internet access to my wife (and our parrots, when they play with the keyboard). I do all the troubleshooting of things like email and web-access problems. So can I am obviously qualify as an ISP, right? It'd be fun to be able to sue the spammers for DoS of our cable modem and wasting our time.
What I imagine is a new "customer convenience" feature: To buy something, all you have to do is pick it up and walk out of the store. The RFID sensors would identify you and the things you're buying, and automatically charge your account.
At least that's what the advertising would tell you.
What would also happen is that the RFID sensors would also identify the articles of clothing that you're wearing from their embedded chips, and would charge you for those, too. And every time you went out of that store, you'd be charged again for everything you're wearing.
After a while, of course, customers would wise up to this and raise a stink. So the stores would fix the bug. They'd only pick one or two such items, and only when you're carrying a lot of new purchases. That way, you probably wouldn't notice the extra charges, unless you were really a stickler for checking every purchase. And even then, you'd face taking off a day of work and trying to prove to them that they'd included an extra charge.
And if you tried to pay for something with cash, you might find yourself also being charged via the RFID. This has a precedent here in Massachusetts. The local toll roads have a "Fast Lane" electronic toll collection scheme. Usually it works well. But something that some people have found: In heavy traffic, sometimes you can't reach the Fast-Lane tollbooth. If you decide to go through a cash-only booth, there's a good chance that the sensors will detect your car and charge your account in addition. Sure, you can challenge it, and you'll probably win easily. But you'll have to take at least one day off work, and that's a steep price to pay to get maybe a $1 refund.
Several decades ago, American phone companies started centralizing their "information", typically in rural midwestern places where the accent wasn't as deviant (;-) as in the even cheaper South. A real problem ever since then has been the odd pronunciation of many place names, which the phone-company people generally can't guess.
This is partly because of all the place names taken from abo/native/Indian languages. But not entirely. Thus, here in New England, we have towns like Reading, pronounced as if it were "Redding". I grew up in the Seattle area (which I've heard pronounced "seat"+"ull" by Easterners). One of the fun place names there is Puyallup, (mis)pronounced by the locals as "pyu-Al-up".
Sometimes this causes serious problems when trying to communicate with the phone-company person. They just can't map your pronunciation to anything in their database, and you can't guess how they expect place names to be mispronounced. My wife is from the Hudson River valley, where there are a lot of place names pronounced in ways that really surprise outsiders. But the locals think that's how they are pronounced, and are really upset when the phone-company person thinks there's no such town.
There has been a bit of a move away from this centralization on the part of some phone companies. They also have software that can take a semi-phonetic spelling and match it (sometimes). But it's still an ongoing problem.
Possibly the wierdest example was when I called an Arizona info line and asked for a person in Phoenix ("fee"+"nicks"). The person on the other end couldn't find that town in her listing.
Or we could even preserve half the terminology, and use master/disciple. The religious folk should like that.
There are lots of alternatives.
The dom/sub pair is a good one, though. It does go along with the usual meaning of master/slave in a computing context.
I wonder how long it would take to 1) get all standards orgs to decide on the replacement terms, and 2) convert all existing standards documents to use the new terms.
A bit of satirical anti-PC humor that I've seen in Scottish environments is to have the restrooms labelled with icons of people wearing either a dress or a kilt.
Heh, yeah. In fact, I do remember one article on the topic that talked about this. One suggestion was two classes of flight. One would be truly free-fall, with the appropriate warnings. The other kind would be powered the whole trip. This would use more fuel, so it would cost more. But it would be faster, and wouldn't be free-fall.
How practical this would be, I don't know. It is a bit disappointing that the airline industry doesn't seem to have taken it at all seriously.
I'd think the vomit comet sounds like a fun ride, as long as I can get a window seat.
Back when the Concorde first came out, and occasionally after that, there were articles comparing supersonic flight to suborbital flight.
The main benefit of supersonic flight is that, since you're in the atmosphere, you don't need to carry your oxidizer with you. Also, you can maintain cabin pressure with a lighter shell, since the outside pressure is nonzero. But you have to push your way through all that air, so you have to be powered the whole trip.
The general benefits of the suborbital path is that you spend most of your travel time coasting above the atmosphere, not using fuel at all, and every place on Earth is at most 90 minutes away, plus the time it takes to get up and down, for a max of 2 hours. But you need to carry at least some of your oxygen (both for fuel and for breathing), and the cabin needs to be strong enough to hold pressure in a near vacuum.
The engineering calculations concluded that the crossover point in fuel consumption was at about 1500 km (1000 miles); at longer distance the suborbital flight would use less fuel than the supersonic flight.
All this was basically engineering estimates, though; nobody seems to have seen a motive for seriously developing the suborbital approach. This is probably because "space flight" is generally considered way out and unfeasible, no matter what the engineers say.
It'd be interesting to read some up-to-date calculations on this topic. I haven't seen any for a few years. Google doesn't seem to find any (or I'm not guessing the right keywords).
Anyone know of any good, recent writing on the topic?
Actually, from his history, it's fairly clear that Bill Clinton does have one very important job skill: He's good at campaigning. He does seem at a bit of a loss when he has nothing to campaign for. Even his worst enemies have repeatedly said what a personable guy he is, and how they always find themselves leaving the room thinking he's on their side and is their best friend.
It's still not clear what he's going to do with the rest of his life.
Maybe he should sign on as an acolyte to Jimmy Carter.
... we'll have a complete paper record of who voted for who.
Great. We have an unelected president who has clearly stated that anyone who doesn't support him will be considered a terrorist. He is holding hundreds of people incommunicado, indefinitely and without charges already. We're installing voting equipment built by a company whose CEO has promised that he can deliver the next election to Bush. And we're also going to make sure that the voting equipment can keep a record of who I've voted for.
This really gives me a warm, fuzzy feeling about voting...
Ah, the old false dichotomy between a democracy and a representative republic. The US is, of course, both. Check your dictionary.
But the electoral failure three years ago was a result of something else that few people other than historians ever mention: The US Electoral College was in fact set up by the Founding Fathers as an explicit check on the power of the masses. They were afraid of a popular demagogue winning an election and overthrowing the established order. Not an irrational fear, as illustrated by several cases in the 20th century where this happened in some other countries. So they devised that peculiar scheme whereby the voters choose "electors", presumably well-to-do members of the established parties, and those electors then decide amongst themselves who should be the president. This system can overturn the wishes of the masses, and that's exactly what it was designed to do.
In this case, it did have some help from a court that ordered a halt to the vote counting, so that one state could "choose" the desired set of electors. This is something that the Founding Fathers apparently didn't anticipate, and has thrown a major monkey wrench into the works. But this isn't the first time; check out the 1876 election for a precedent. ("Rutherford Tilden election" is a good set of keywords for a search site.)
Now we have the have the phenomenon of new voting equipment being widely installed, from a company whose CEO has brazenly promised one party that he can deliver states to them in the next election. Information about this equipment supports his claim fully.
So maybe we can truly remove the US from the list of democracies...
They certainly are. Maybe it's not broadcasting (unless the listener happens to be a broad), but when you let someone else listen to a tune on your iPod, you are distributing it to them. And you are copying it into their brain. This wasn't part of the license that they sold you.
Re:The CIS majors must know something the CS don't
on
In Search of Stupidity
·
· Score: 3, Informative
Back in the 1970's, a bunch of guys at Rand implemented a very successful handwriting input package. It even included an editor that used a lot of the standard editor's marks to make changes.
They commented that there was a serious misunderstanding of the difficulty of handwriting recognition: If you want the computer to take a page of handwriting and recognize everything on it, that is nearly impossible. But if the computer can follow the handwriting as it is written, the job is fairly easy.
They also made the point that their software was typically only around 90% accurate (counting characters) when a person started using it. But it improved quickly. This wasn't because the software learned your writing; it didn't. It was because, when it drew the wrong character, you did it over until the software got it right. This trained the person to write in a way that the software could recognize. One side effect was that users of their system had noticably better handwriting after a few days.
But reading a sheet full of handwriting is still a very difficult task for a computer. Is there any software that does it well enough that you don't have to edit the results?
Sadly, I think this had the effect of driving up the long term costs to maintain their tech which they're still paying for today. Like business, they have simply accepted that IT has to be expensive, but it's a necessary evil to stay competitive.
You're probably right. But they may well be wrong. After all, this was the same reasoning that, back in the 1970's and early 1980's, led many schools to install IBM equipment. Those schools then found themselves paying exorbitantly to stay in what turned out to be a backwater. The future was in desktop machines, not mainframes.
It's risky trying to be a prophet, but there's a good chance that schools using this reasoning will find themselves in a similar situation. As with biology, the computing future has pretty much always belonged to the small, agile beasts. The big and powerful might look dominant, but they eventually lose out to the little guys. Some of those little guys evolve to big, powerful beasts, and then the pattern repeats.
A school that wants its students to be winners should always bet on the smallest, most versatile computers. This especially means computers which are programmable by individuals, not by huge IT departments. This should be easy, because these are usually the cheapest computers. But this probably won't happen in many schools.
But there's an ongoing question of how much actual education most of our schools are really interested in. There's a lot of evidence to support the theory that schools' basic function is socialization, and education is mostly a side effect. One of the better pieces of evidence is the widespread use of closed, proprietary computer systems, which don't qualify as "educational" except in the most minimal sense.
Oh, well; computer education has always been mostly at the college (and trade school) level. Very few high-school graduates have ever been able to write even a "Hello, world" program in any language. But we've managed to turn out a small population of computer wizards anyway. Maybe it doesn't matter all that what the public schools do.
I would think that the fall freshman CS class would be better qualified to write an article than Howard Strauss based on the completely erroneous comments he gave in his editorial.
Yeah, but he'd probably argue that they don't count, because they are a tiny minority, and they're all the sort of weirdos that modify source code.
But then, I probably shouldn't be allowed to comment, either. After all, I get paid to modify source code. I do this nearly every day. Sometimes I even write source code from scratch.
I'd bet there even a few others like me among the/. readers.
In the many years I spent in various US schools, I never saw a choice about any educational material. It was extremely rare for the teacher to do the choosing. The universal approach is that the school administration makes all such choices, and both the teachers and students have to attempt to use what "choice" was handed to them.
Suggesting that school student are "choosing" Microsoft (or sometimes Apple) is one of the most cynical comments that I've read here in a very long time. They no more chose their computers than they choose their textbooks. It's rare that they can even choose their teacher. A very few school systems allow a "choice" among a few schools, but this decision is almost always made by the parents, not the student.
I'm sure I'm not the only one who has noticed this...
...basically schools used to use only Macs and clung to them until recently when they started switching to PCs.
When true, this is rather damning for a school.
Schools exist to educate their students, after all. A "black box" system whose inner workings aren't available to students would be the last thing you'd expect a school to use. It would be proof that the school isn't interested in teaching their students to understand the computers.
It used to be that the Mac was the most extreme black-box computer, with "no user-servicable parts". But Macs have been converted to a unix base in recent years. Not just unix, but a rather open and accessible version of unix. So the Mac has gone from being useless for education to being one of the best.
If a school reacts to this by switching from Macs to MS systems, it is clear proof that they don't want their students to become knowledgable about computers. If they were truly trying to educate, why would they switch to what is now the most closed system on the market?
This would be much like, say, a school offering driver's ed, but not auto shop. Or if they had auto shop, not allowing the students to open the hood and start taking a car apart.
People with kids in school should investigate, and if your school system is using closed, proprietary computers, you should complain loudly. Keeping the students ignorant of how computers work (hardware or software) is not a good way to prepare them for the world that is coming.
Once again, the main technical problem lies with Windows.
True, but it's really not limited to them. Thus, if you install mozilla, it comes with java and javascript enabled by default. These may be a lot more secure that MS's scripting schemes, but they are channels for software that you might not want installed.
We really need education to help people understand why it's never a good idea to let software download code and run it automatically. Any scripting facility should be turned off. It should be off by default, no matter how convenient such things may be for users.
Windows is a special problem, because much of the scripting facility has been moved into the "system". None of the unixoid systems do this, so they are inherently a lot safer. But on any system, this problem can be introduced by apps that have the ability to download code and run it without asking you.
Slashdot Mirror
Slashdot has had a number of articles about the iTron OS that's being used in a lot of Japanese autos these days. That and linux are both good candidates for a stable, reliable onboard OS. Of course, behind the scenes, iTron and linux are busy "stealing" ideas from each other. It'll be interesting to see which will win in the long run, or if they'll slowly merge. Or maybe divide the territory by consciously heading toward different parts of the tradeoff space.
I wonder if anyone is looking at any of the *BSD kernels for this sort of embedded use? They have reliability records comparable to iTron and linux, and also come with all the source code.
Consumer Reports had an article recently that mentioned this general topic. Their general suggestion was to avoid buying the new computerized luxury cars for a few years, until they get the UI right. It seems that in general their testers couldn't make much sense out of the menu-based centralized controls, and they considered these controls to be hazardous in the extreme under normal driving conditions.
BMW's newer models were not nearly the only bad examples.
Trying to discover where they've hidden some control in an N-level-deep menu tree is extremely distracting. You don't want this when you're driving.
Of course, if you screw up, the manufacturer will just call it an "operator error".
Then never complain ... when your stuff gets downloaded.
But I have a very good complaint: My web site has my music on it. If this goes through, any Canadian downloading my music from my web site will be paying a tax to the recording industry. So, while I won't get any income from those downloads, someone else with no rights to my music will.
It's bad enough that the recording industry can force "standard" contracts on musicians that give all rights and profits to the recording company, and claim that this is "voluntary". Yeah; it's voluntary; you always have had the choice of nobody hearing your music because you can't get it distributed without signing one of these contracts.
But this sort of tax gives them profit from my music when I haven't signed any contract at all.
Somehow, I'm not too happy with this idea.
That isn't properly called "security through obscurity".
It's called a "honeypot".
There's a big difference in the security implications.
I tried your site, and I got:
...
...
http://www.isthatdamngood.com
Microsoft OLE DB Provider for ODBC Drivers error '80040e57'
It isn't obvious what this means, but they appear to be claiming that there is some Microsoft software running on port 80 on your machine, not linux or apache.
So I tried my home machine, which is RH linux with apache, and got:
Problem communicating with remote server: [my hostname]
So in my case, my OS/server simply isn't being counted.
These two samples don't really impress me with their competence. But then, they do state right at the top that they are only counting the "top commercial" web sites. So their results aren't in conflict with netcraft, because they're counting different things.
Hey, I run the linux box that is the firewall between our home LAN and the cable modem. I provide internet access to my wife (and our parrots, when they play with the keyboard). I do all the troubleshooting of things like email and web-access problems. So can I am obviously qualify as an ISP, right? It'd be fun to be able to sue the spammers for DoS of our cable modem and wasting our time.
What I imagine is a new "customer convenience" feature: To buy something, all you have to do is pick it up and walk out of the store. The RFID sensors would identify you and the things you're buying, and automatically charge your account.
At least that's what the advertising would tell you.
What would also happen is that the RFID sensors would also identify the articles of clothing that you're wearing from their embedded chips, and would charge you for those, too. And every time you went out of that store, you'd be charged again for everything you're wearing.
After a while, of course, customers would wise up to this and raise a stink. So the stores would fix the bug. They'd only pick one or two such items, and only when you're carrying a lot of new purchases. That way, you probably wouldn't notice the extra charges, unless you were really a stickler for checking every purchase. And even then, you'd face taking off a day of work and trying to prove to them that they'd included an extra charge.
And if you tried to pay for something with cash, you might find yourself also being charged via the RFID. This has a precedent here in Massachusetts. The local toll roads have a "Fast Lane" electronic toll collection scheme. Usually it works well. But something that some people have found: In heavy traffic, sometimes you can't reach the Fast-Lane tollbooth. If you decide to go through a cash-only booth, there's a good chance that the sensors will detect your car and charge your account in addition. Sure, you can challenge it, and you'll probably win easily. But you'll have to take at least one day off work, and that's a steep price to pay to get maybe a $1 refund.
Several decades ago, American phone companies started centralizing their "information", typically in rural midwestern places where the accent wasn't as deviant (;-) as in the even cheaper South. A real problem ever since then has been the odd pronunciation of many place names, which the phone-company people generally can't guess.
This is partly because of all the place names taken from abo/native/Indian languages. But not entirely. Thus, here in New England, we have towns like Reading, pronounced as if it were "Redding". I grew up in the Seattle area (which I've heard pronounced "seat"+"ull" by Easterners). One of the fun place names there is Puyallup, (mis)pronounced by the locals as "pyu-Al-up".
Sometimes this causes serious problems when trying to communicate with the phone-company person. They just can't map your pronunciation to anything in their database, and you can't guess how they expect place names to be mispronounced. My wife is from the Hudson River valley, where there are a lot of place names pronounced in ways that really surprise outsiders. But the locals think that's how they are pronounced, and are really upset when the phone-company person thinks there's no such town.
There has been a bit of a move away from this centralization on the part of some phone companies. They also have software that can take a semi-phonetic spelling and match it (sometimes). But it's still an ongoing problem.
Possibly the wierdest example was when I called an Arizona info line and asked for a person in Phoenix ("fee"+"nicks"). The person on the other end couldn't find that town in her listing.
I remember back in grade school thinking that "black board" wasn't a very good term, because most of them were green.
;-)
Then, when the first "white boards" came out, many of which were beige or cream, but definitely not white, it seemed even sillier.
The terms "marker board" and "chalk board" make a lot more sense, because that tells you the significant thing about what sort of board it is.
(As long as we don't get into a debate over whether a piece of chalk is a "marker".
Well, we could call them leader/acolyte.
Or we could even preserve half the terminology, and use master/disciple. The religious folk should like that.
There are lots of alternatives.
The dom/sub pair is a good one, though. It does go along with the usual meaning of master/slave in a computing context.
I wonder how long it would take to 1) get all standards orgs to decide on the replacement terms, and 2) convert all existing standards documents to use the new terms.
A bit of satirical anti-PC humor that I've seen in Scottish environments is to have the restrooms labelled with icons of people wearing either a dress or a kilt.
...
It's perfectly obvious which you should use
Heh, yeah. In fact, I do remember one article on the topic that talked about this. One suggestion was two classes of flight. One would be truly free-fall, with the appropriate warnings. The other kind would be powered the whole trip. This would use more fuel, so it would cost more. But it would be faster, and wouldn't be free-fall.
How practical this would be, I don't know. It is a bit disappointing that the airline industry doesn't seem to have taken it at all seriously.
I'd think the vomit comet sounds like a fun ride, as long as I can get a window seat.
Some people really like roller-coaster rides
... low-earth-orbit ...
Possibly, or a suborbital arc.
Back when the Concorde first came out, and occasionally after that, there were articles comparing supersonic flight to suborbital flight.
The main benefit of supersonic flight is that, since you're in the atmosphere, you don't need to carry your oxidizer with you. Also, you can maintain cabin pressure with a lighter shell, since the outside pressure is nonzero. But you have to push your way through all that air, so you have to be powered the whole trip.
The general benefits of the suborbital path is that you spend most of your travel time coasting above the atmosphere, not using fuel at all, and every place on Earth is at most 90 minutes away, plus the time it takes to get up and down, for a max of 2 hours. But you need to carry at least some of your oxygen (both for fuel and for breathing), and the cabin needs to be strong enough to hold pressure in a near vacuum.
The engineering calculations concluded that the crossover point in fuel consumption was at about 1500 km (1000 miles); at longer distance the suborbital flight would use less fuel than the supersonic flight.
All this was basically engineering estimates, though; nobody seems to have seen a motive for seriously developing the suborbital approach. This is probably because "space flight" is generally considered way out and unfeasible, no matter what the engineers say.
It'd be interesting to read some up-to-date calculations on this topic. I haven't seen any for a few years. Google doesn't seem to find any (or I'm not guessing the right keywords).
Anyone know of any good, recent writing on the topic?
Because Clinton had no real job or real skills,
Actually, from his history, it's fairly clear that Bill Clinton does have one very important job skill: He's good at campaigning. He does seem at a bit of a loss when he has nothing to campaign for. Even his worst enemies have repeatedly said what a personable guy he is, and how they always find themselves leaving the room thinking he's on their side and is their best friend.
It's still not clear what he's going to do with the rest of his life.
Maybe he should sign on as an acolyte to Jimmy Carter.
Yeah, but big and shiny as it might be, the sun isn't nearly as complex as for example a fruit fly or a nematode.
Of course, we are a long way from really understanding them.
... we'll have a complete paper record of who voted for who.
...
Great. We have an unelected president who has clearly stated that anyone who doesn't support him will be considered a terrorist. He is holding hundreds of people incommunicado, indefinitely and without charges already. We're installing voting equipment built by a company whose CEO has promised that he can deliver the next election to Bush. And we're also going to make sure that the voting equipment can keep a record of who I've voted for.
This really gives me a warm, fuzzy feeling about voting
Ah, the old false dichotomy between a democracy and a representative republic. The US is, of course, both. Check your dictionary.
...
But the electoral failure three years ago was a result of something else that few people other than historians ever mention: The US Electoral College was in fact set up by the Founding Fathers as an explicit check on the power of the masses. They were afraid of a popular demagogue winning an election and overthrowing the established order. Not an irrational fear, as illustrated by several cases in the 20th century where this happened in some other countries. So they devised that peculiar scheme whereby the voters choose "electors", presumably well-to-do members of the established parties, and those electors then decide amongst themselves who should be the president. This system can overturn the wishes of the masses, and that's exactly what it was designed to do.
In this case, it did have some help from a court that ordered a halt to the vote counting, so that one state could "choose" the desired set of electors. This is something that the Founding Fathers apparently didn't anticipate, and has thrown a major monkey wrench into the works. But this isn't the first time; check out the 1876 election for a precedent. ("Rutherford Tilden election" is a good set of keywords for a search site.)
Now we have the have the phenomenon of new voting equipment being widely installed, from a company whose CEO has brazenly promised one party that he can deliver states to them in the next election. Information about this equipment supports his claim fully.
So maybe we can truly remove the US from the list of democracies
This definition needs a new entry:
c: a government in which the supreme power is vested in an unelected court that can decide which votes are to be counted.
They certainly are. Maybe it's not broadcasting (unless the listener happens to be a broad), but when you let someone else listen to a tune on your iPod, you are distributing it to them. And you are copying it into their brain. This wasn't part of the license that they sold you.
Back in the 1970's, a bunch of guys at Rand implemented a very successful handwriting input package. It even included an editor that used a lot of the standard editor's marks to make changes.
They commented that there was a serious misunderstanding of the difficulty of handwriting recognition: If you want the computer to take a page of handwriting and recognize everything on it, that is nearly impossible. But if the computer can follow the handwriting as it is written, the job is fairly easy.
They also made the point that their software was typically only around 90% accurate (counting characters) when a person started using it. But it improved quickly. This wasn't because the software learned your writing; it didn't. It was because, when it drew the wrong character, you did it over until the software got it right. This trained the person to write in a way that the software could recognize. One side effect was that users of their system had noticably better handwriting after a few days.
But reading a sheet full of handwriting is still a very difficult task for a computer. Is there any software that does it well enough that you don't have to edit the results?
Sadly, I think this had the effect of driving up the long term costs to maintain their tech which they're still paying for today. Like business, they have simply accepted that IT has to be expensive, but it's a necessary evil to stay competitive.
You're probably right. But they may well be wrong. After all, this was the same reasoning that, back in the 1970's and early 1980's, led many schools to install IBM equipment. Those schools then found themselves paying exorbitantly to stay in what turned out to be a backwater. The future was in desktop machines, not mainframes.
It's risky trying to be a prophet, but there's a good chance that schools using this reasoning will find themselves in a similar situation. As with biology, the computing future has pretty much always belonged to the small, agile beasts. The big and powerful might look dominant, but they eventually lose out to the little guys. Some of those little guys evolve to big, powerful beasts, and then the pattern repeats.
A school that wants its students to be winners should always bet on the smallest, most versatile computers. This especially means computers which are programmable by individuals, not by huge IT departments. This should be easy, because these are usually the cheapest computers. But this probably won't happen in many schools.
But there's an ongoing question of how much actual education most of our schools are really interested in. There's a lot of evidence to support the theory that schools' basic function is socialization, and education is mostly a side effect. One of the better pieces of evidence is the widespread use of closed, proprietary computer systems, which don't qualify as "educational" except in the most minimal sense.
Oh, well; computer education has always been mostly at the college (and trade school) level. Very few high-school graduates have ever been able to write even a "Hello, world" program in any language. But we've managed to turn out a small population of computer wizards anyway. Maybe it doesn't matter all that what the public schools do.
I would think that the fall freshman CS class would be better qualified to write an article than Howard Strauss based on the completely erroneous comments he gave in his editorial.
/. readers.
Yeah, but he'd probably argue that they don't count, because they are a tiny minority, and they're all the sort of weirdos that modify source code.
But then, I probably shouldn't be allowed to comment, either. After all, I get paid to modify source code. I do this nearly every day. Sometimes I even write source code from scratch.
I'd bet there even a few others like me among the
In the many years I spent in various US schools, I never saw a choice about any educational material. It was extremely rare for the teacher to do the choosing. The universal approach is that the school administration makes all such choices, and both the teachers and students have to attempt to use what "choice" was handed to them.
...
Suggesting that school student are "choosing" Microsoft (or sometimes Apple) is one of the most cynical comments that I've read here in a very long time. They no more chose their computers than they choose their textbooks. It's rare that they can even choose their teacher. A very few school systems allow a "choice" among a few schools, but this decision is almost always made by the parents, not the student.
I'm sure I'm not the only one who has noticed this
...basically schools used to use only Macs and clung to them until recently when they started switching to PCs.
When true, this is rather damning for a school.
Schools exist to educate their students, after all. A "black box" system whose inner workings aren't available to students would be the last thing you'd expect a school to use. It would be proof that the school isn't interested in teaching their students to understand the computers.
It used to be that the Mac was the most extreme black-box computer, with "no user-servicable parts". But Macs have been converted to a unix base in recent years. Not just unix, but a rather open and accessible version of unix. So the Mac has gone from being useless for education to being one of the best.
If a school reacts to this by switching from Macs to MS systems, it is clear proof that they don't want their students to become knowledgable about computers. If they were truly trying to educate, why would they switch to what is now the most closed system on the market?
This would be much like, say, a school offering driver's ed, but not auto shop. Or if they had auto shop, not allowing the students to open the hood and start taking a car apart.
People with kids in school should investigate, and if your school system is using closed, proprietary computers, you should complain loudly. Keeping the students ignorant of how computers work (hardware or software) is not a good way to prepare them for the world that is coming.
Once again, the main technical problem lies with Windows.
True, but it's really not limited to them. Thus, if you install mozilla, it comes with java and javascript enabled by default. These may be a lot more secure that MS's scripting schemes, but they are channels for software that you might not want installed.
We really need education to help people understand why it's never a good idea to let software download code and run it automatically. Any scripting facility should be turned off. It should be off by default, no matter how convenient such things may be for users.
Windows is a special problem, because much of the scripting facility has been moved into the "system". None of the unixoid systems do this, so they are inherently a lot safer. But on any system, this problem can be introduced by apps that have the ability to download code and run it without asking you.