>If it is ever successful, it will create sidewalk traffic jams to match those made by cars.
Strange argument. Like Yogi Beara, "Nobody goes there anymore, its too popular." So you're arguing that a) it will never be successful anyway but b) if it is successful, it cannot be succesful.
At the point where people find them useful enough to cause traffic jams, then we will be able to put in additional infrastructure to support it, because we will have replaced enough of the current transportation infrstrucutre to sucure right-of-way. So this would be a success story.
>The real problem is the technological approach. Instead of trying to do what I already can do as a human being, walking and running, it should do what I can't do. Taking the latter approach is why technologies like the light bulb, the airplane, the automobile, the refrigerator, and the computer were successful.
Another strange argument, IMO. These things didn't do what people couldn't allready do, they just did some of those things marginally better.
Light bulb: bright light without the hasle, smoke and danger of lighting a candle.
Airplanes: "Whats wrong with taking the train?
Automobiles: "They aren't any faster than a horse, who needs 'em?"
Refrigerator; "My ice box works just fine, thank you, and the iceman is a friend of mine."
Computer: Why spend all of the company's money on a big expensive contraption that isn't going to do anything that we can't allready do?
The point is that these all seemed to be relatvily incremental changes when they were first introduced. They were revolutionary in some sense, but in terms of adoption and to the majority of people at the time, they often seemed like solutions in search of a problem. It is only in hindsight that _most_ people realized their obvious utility.
Which brings me to something I thought the last time this issue out. What the hell is wrong with all of you/.'rs? Did your weekly claven come out with a ruling that the HT just wasn't cool enough for you? Sonce when did geeks worry about what was cool? Aren't we supposed to be the one's who are interested in dorky things that no one else understands until much later. Or are you all a bunch of posers? Thats what I'm thinking. There are probably very few real nerds left in this business.
Slashdot Geekdom is getting so twisted and dare I say it..mainstream. When did any true geeks ever care what other people thought? Not meaning to get persomal, but what the hell is "Unorthodox" about pointing at people and laughing at them? Seemed pretty f*&(ing common at my jr.high in the late 70s.
You are right that replacing walking with this would be a pretty bad outcome. Definetly the biggest downside. But hopefully, if it meant craeting more walking style infrstructure and more walking friendly areas, as well as more interaction in pedestrain space (as opposed to car scape) it would actullay encourage people to walk more. Social forces work in strange ways.
Yea, I think thats a really good idea. People could use them for exploring once they got to a general location on mass transit.
Re:What happens when it stops working?
on
This is IT?
·
· Score: 1
They've thought of that. (Does anyone actaully read anything before having an "opinion"?)
The system ahs huge redundencies built into it. Redundent gyroscopes, double redeundent in both the controllers and motor. So the only thing that will hapen is a failure id that you'll come to a nice controlled halt.
Man, what a bunch of whiners...
on
This is IT?
·
· Score: 1
...when did tech people become such a bunch of negative finger-waggers. "Its just a scooter. I'll hit my head. It doesn't run Linux. Its slower than a car. Its not much better than walking. Its all hype. Waa waa waa." so it must not be cool.
Glad you all weren't around in the Altair days. "Its just a bunch of blinking lights. It doesn't do large batch jobs. What good is it? Who really needs a computer anyway? Whats all this hpe about computers? Waa waa waa."
Sorry if I'm rude but Bah! When did the industry become so full of (young) crumudgeons? You guys are no fun!:-)
One question: is try/finally an expensive or risky thing to do, or is it commonly used?
Nope, its not expensive..the cost is almost entirely in the catch case, and as stated, the code should be written so that the catch cse is rare.
Not risky either, the whole point of using it is to manage risk..methods are often written so that you are forced to trap for exceptions that might occur. For instance, if you open a file you have to trap for an IOException. Of course a programmer can still choose to ignore what's caught--there's still no way to legislate good coding.
So yep, try, catch, finally is used all over the place, but probably not quite as often as it should be in non-commercial code.
C# suppots Goto: after 30 years of people trying to get rid of it this misrable programming construct, Microsoft include it in their 'clean sheet' language. How pathetic is that?
Anyone that tries to pretend that C# isn't a Java ripoff is just completely full of it, and doesn't know a thing about Java. The gall is unbelievable. Its the big lie.
have to confess that one thing I would really like to kill in Linux is the sheer profusion of licences. I never can be bothered to read the things anyway. Maybe we could do with a license rating system with (100% = GPL, 0% = Microsoft Shrink wrap licence).
I think that's a great idea..then people can decide for themselves how strongly they want to adhere to OSS. In practice, it might be hard to implement though..
>>Is it just me or does anybody think that Transmeta is using Linux+Linus as a way to sell their chips?
Well, duh!:-)
Even with all the free (as in beer) software discussions out there I don't think that there are that many companies that hire people hoping that they won't somehow help move more product.
I think they recognized that adding him to their roster would give them a little extra cachet, though I doubt they realized how much it would help them in some people's eyes. As you'd prob agree, that is more the fault of the maniupulee than the manipuler.
But I also think tat they felt that his expertise would genuinly contribute to their linux side of things, and I think their obvious commitment to the Linux platform cannot be undersold. If anything, the whole Linus phenomonen is just a sign that Linux has extrodinary mind-share, and thats largely a good thing, isn't it?
But ultimatly, the reason that it was a good move on Transmeta's part is that people trust that Linux wouldn't put his 'thumbs up' on a box 'o shit, and that's why his opinion is worth something.
Eww..its alway's so nerve-racking listening to the questions of 'mainstream' media people. They have negative clue technology wise--just enough knowledge to ask incredibly inane questions.
Like the CNN guy who asked wether the processor would help Window's run 'better' (more stable.) Uh yeah dude, sure. They thought they'd back engineer Windows because they had a little extra time to kill. He made clear he didn't understand the technology _at all_ by asking a question like that.
The rest of the stuff on MSNBC (of course) and news.com isn't much better.
Its always fun to watch a primary source and then read the news; you find out how much is distorted or abstracted away.
Would they send a reporter to Bosnia who didn't have a clue about european politics or history? Oh..wait a minute..they probably would.
You are assuming that time is linear, and that there is only one time 'thread'. What if, for example, multiple time 'threads' could spawn without bound in any direction? I admit that this doesn't neccessarily solve all contradictions or objections to the possiblity of information travel backwards in time, but FFT.
Traditionally the dharma [Buddhist understanding] and teachings about the dharma should not be sold. In fact IIRC, in the past it was considered a really bad move to sell liturgies.
Of course, in the modern world, where most ideas are conveyed through books, and it would be almost impossible to convey the dharma without having people sell books, its no longer considered a problem. There are many very good Dharma book publishers who have contributed greatly to the spread of Buddhism and made a good living at the same time.
I don't really think Java is a landmine, but as AC argued, it makes sense to treat it carefully.
I think the big danger of adopting it too early would have been that the Be msg would get confused and lose focus. So its more of a marketting msg than even a developer msg. Now that Be has convinced everyone that it has a life of its own, it can afford to let Java into the mix.
Not to mention that I bet that by playing it cool cannny ol' Gasee put himself into a nice negotiating position as far as licensing and development help from Sun go. If you think about it, it'd be pretty ridiculous for Be's appliance offereing not to have Java, and so from this perspective, Gasee did a superb job of playing hard to get.
Don't know anything about KDE's implementation, to be honest. I have to point out to that while Be's APIs are relatively simple, they're also incredibly powerful and well and consitently thought out. I would hope KDE's the same way, but I just don't know. You can get all the API docs and stuff at the Be site. It might be worth it to take a glance.
I also like ObjC. Much better than C++. Actually, a real decent language. Too bad it never really caught on; but I actually know a lot fo people that work with it today for specialized stuff. I wonder if C++ might be a little better for low level stuff, but I don't know.
What it comes down I guess is that any language can be used or abused. Good C++ code is still better than crappy Java code!
"Does this mean that you do not feel that JBED can work well?"
Nope. Perhaps I sacrificed a little clarity for brevity. What I should have siad is that overall you probably couldn't really build a decent responding desktop style OS in Java right now. While Java has great computatoional speed in some areas (I get better than C++ for some of my stuff) it still may not have the broad perfromance neccessary, esp. when dealing with graphics and device drivers, and any implementation that we did now that was all Java would take a long time to come to fruition, whereas BeOS is here now.
As someone who has done a fair amount of Swing development, I don't think you could quite do a full-on desktop that would have the perfromance that users expected and still leave some cycles for other tasks. Of course, if you take away the intermediate levels between the vm and the graphics, etc.. you might very well get good performance.
As far as JBED goes, remember that !(Real Time lt==gt Performance) All a Real Time system does is provide guaranteed performance; it doesn't specify what that perfromance actually is. One could call a system "real-time" that did one add op every second, as long as it refused to guarantee that it could do the op any faster.
That said, I wouldn't be surprised if JBED does give quite decent performance, but it is not a desktop OS. I think small-embedded type OSs are a natural fit for Java right now. Ironically, embedded systems don't typically have the same performance challenges that desktop systems do. My impression is that they're typically designed such that the processor is underutilizzed if anything. The time taken to display a couple of lines of text on a cell-phone for instance is infintesimal compared to rendering a busy workstation desktop. At the same time, you never want that cellphone to take any perceived time to display something, so you design with that in mind. Its a much more controllable situation.
But yes, I do think we'll see more and more Java to the metal kind of implementations as the world rolls on, I just don't expect to see them tmrw. I also think diversity is healthy so I'd like to see and support both.
This is great news. We've been memeber of the BeOS developers program for a couple of years, but I haven't been able to do any 'real' development on it because our code has to be able to run on other machines.
BeOS itself is so slick..once you've used it, it is such a drag (literallly) to go back to NT, or even Linux. There is _no_ refresh lag, resources are used up smoothly and predicatbly, and the whole thing boots instantaneously (at least when compared to NT!) Also, with real file jouranling, you can even shut the machine off (not reccommended though!) and not worry that you've just corrupted your entire file system; when you shut down correctly, there is hardly any wait at all.
BeOS also has a commendably simple, clean and elegant API implementation. Actually, its the cleanest, most rational C++ implementation I've seen. (Hmmm, maybe thats not saying much. I guess I should say its a great implementation regardless of language.) Interestingly, they have explicitly avoided all the C++ 'enhancements' like STL, exceptions model, etc.., etc.. and are really writing to the bone; they're really using C++ like C with simple object extensions which is the only way I'd ever use it.
I love Java, but I'm not a completly fanatical idiot about it. Right now, you simply couldn't implement the low-level parts of a decent OS in Java. If this is a good Java implementation, with the BeOS finegrained threading, clean graphics mdel, etc.., this has the potential to be the best Java implementation of them all. Truly the best of both worlds.
Still trying hard to keep my expecations down so if its a half-assed implementaiton I won't Be Bummed..
Slashdot Mirror
"Microsoft Claimed it 'will recognize all your handwriting unless you can't read it yourself."
Well, then I'm screwed...
Tom's hardware story
damn cool; now with high-end x86 processors and better graphics:
5 /c ebit-08.html
Defietly the one thing I could see wanting; not a G4 cube, but a lot cheaper and more powerful.
http://www.tomshardware.com/business/02q1/02031
>If it is ever successful, it will create sidewalk traffic jams to match those made by cars.
/.'rs? Did your weekly claven come out with a ruling that the HT just wasn't cool enough for you? Sonce when did geeks worry about what was cool? Aren't we supposed to be the one's who are interested in dorky things that no one else understands until much later. Or are you all a bunch of posers? Thats what I'm thinking. There are probably very few real nerds left in this business.
Strange argument. Like Yogi Beara, "Nobody goes there anymore, its too popular." So you're arguing that a) it will never be successful anyway but b) if it is successful, it cannot be succesful.
At the point where people find them useful enough to cause traffic jams, then we will be able to put in additional infrastructure to support it, because we will have replaced enough of the current transportation infrstrucutre to sucure right-of-way. So this would be a success story.
>The real problem is the technological approach. Instead of trying to do what I already can do as a human being, walking and running, it should do what I can't do. Taking the latter approach is why technologies like the light bulb, the airplane, the automobile, the refrigerator, and the computer were successful.
Another strange argument, IMO. These things didn't do what people couldn't allready do, they just did some of those things marginally better.
Light bulb: bright light without the hasle, smoke and danger of lighting a candle.
Airplanes: "Whats wrong with taking the train?
Automobiles: "They aren't any faster than a horse, who needs 'em?"
Refrigerator; "My ice box works just fine, thank you, and the iceman is a friend of mine."
Computer: Why spend all of the company's money on a big expensive contraption that isn't going to do anything that we can't allready do?
The point is that these all seemed to be relatvily incremental changes when they were first introduced. They were revolutionary in some sense, but in terms of adoption and to the majority of people at the time, they often seemed like solutions in search of a problem. It is only in hindsight that _most_ people realized their obvious utility.
Which brings me to something I thought the last time this issue out. What the hell is wrong with all of you
Sorry, I meant they are brushless; don't know how I made the leap from brushless to not wound.
The motors aren't wound.
Slashdot Geekdom is getting so twisted and dare I say it..mainstream. When did any true geeks ever care what other people thought? Not meaning to get persomal, but what the hell is "Unorthodox" about pointing at people and laughing at them? Seemed pretty f*&(ing common at my jr.high in the late 70s.
You are right that replacing walking with this would be a pretty bad outcome. Definetly the biggest downside. But hopefully, if it meant craeting more walking style infrstructure and more walking friendly areas, as well as more interaction in pedestrain space (as opposed to car scape) it would actullay encourage people to walk more. Social forces work in strange ways.
Yea, I think thats a really good idea. People could use them for exploring once they got to a general location on mass transit.
They've thought of that. (Does anyone actaully read anything before having an "opinion"?)
The system ahs huge redundencies built into it. Redundent gyroscopes, double redeundent in both the controllers and motor. So the only thing that will hapen is a failure id that you'll come to a nice controlled halt.
...when did tech people become such a bunch of negative finger-waggers. "Its just a scooter. I'll hit my head. It doesn't run Linux. Its slower than a car. Its not much better than walking. Its all hype. Waa waa waa." so it must not be cool.
Glad you all weren't around in the Altair days. "Its just a bunch of blinking lights. It doesn't do large batch jobs. What good is it? Who really needs a computer anyway? Whats all this hpe about computers? Waa waa waa."
Sorry if I'm rude but Bah! When did the industry become so full of (young) crumudgeons? You guys are no fun!
And your point is...?
Hmm, maybe you ought to read Knuth's paper "structured programming with goto statements"
1. Just because its possible doesn't make it a good idea. You could right OO code in Forth, but I wouldn't want to maitain it.
2. The paper refers to structured methodolgies, and; I think its even more out of place now with OO methodologies and design.
3. So Knuth is only right 99.7% of the time..
One question: is try/finally an expensive or risky thing to do, or is it commonly used?
Nope, its not expensive..the cost is almost entirely in the catch case, and as stated, the code should be written so that the catch cse is rare.
Not risky either, the whole point of using it is to manage risk..methods are often written so that you are forced to trap for exceptions that might occur. For instance, if you open a file you have to trap for an IOException. Of course a programmer can still choose to ignore what's caught--there's still no way to legislate good coding.
So yep, try, catch, finally is used all over the place, but probably not quite as often as it should be in non-commercial code.
C# suppots Goto: after 30 years of people trying to get rid of it this misrable programming construct, Microsoft include it in their 'clean sheet' language. How pathetic is that?
Anyone that tries to pretend that C# isn't a Java ripoff is just completely full of it, and doesn't know a thing about Java. The gall is unbelievable. Its the big lie.
"then you really should get back in your little commercial box."
That's exactly the f*&#ed up attitude I'm talking about. I people don't agree with you, then they're completely committed to commercial software.
Eat me.
have to confess that one thing I would really like to kill in Linux is the sheer profusion of licences. I never can be bothered to read the things anyway.
Maybe we could do with a license rating system with (100% = GPL, 0% = Microsoft Shrink wrap licence).
I think that's a great idea..then people can decide for themselves how strongly they want to adhere to OSS. In practice, it might be hard to implement though..
>>Is it just me or does anybody think that Transmeta is using Linux+Linus as a way to sell their chips?
:-)
Well, duh!
Even with all the free (as in beer) software discussions out there I don't think that there are that many companies that hire people hoping that they won't somehow help move more product.
I think they recognized that adding him to their roster would give them a little extra cachet, though I doubt they realized how much it would help them in some people's eyes. As you'd prob agree, that is more the fault of the maniupulee than the manipuler.
But I also think tat they felt that his expertise would genuinly contribute to their linux side of things, and I think their obvious commitment to the Linux platform cannot be undersold. If anything, the whole Linus phenomonen is just a sign that Linux has extrodinary mind-share, and thats largely a good thing, isn't it?
But ultimatly, the reason that it was a good move on Transmeta's part is that people trust that Linux wouldn't put his 'thumbs up' on a box 'o shit, and that's why his opinion is worth something.
Eww..its alway's so nerve-racking listening to the questions of 'mainstream' media people. They have negative clue technology wise--just enough knowledge to ask incredibly inane questions.
Like the CNN guy who asked wether the processor would help Window's run 'better' (more stable.) Uh yeah dude, sure. They thought they'd back engineer Windows because they had a little extra time to kill. He made clear he didn't understand the technology _at all_ by asking a question like that.
The rest of the stuff on MSNBC (of course) and news.com isn't much better.
Its always fun to watch a primary source and then read the news; you find out how much is distorted or abstracted away.
Would they send a reporter to Bosnia who didn't have a clue about european politics or history? Oh..wait a minute..they probably would.
You are assuming that time is linear, and that there is only one time 'thread'. What if, for example, multiple time 'threads' could spawn without bound in any direction? I admit that this doesn't neccessarily solve all contradictions or objections to the possiblity of information travel backwards in time, but FFT.
Traditionally the dharma [Buddhist understanding] and teachings about the dharma should not be sold. In fact IIRC, in the past it was considered a really bad move to sell liturgies.
Of course, in the modern world, where most ideas are conveyed through books, and it would be almost impossible to convey the dharma without having people sell books, its no longer considered a problem. There are many very good Dharma book publishers who have contributed greatly to the spread of Buddhism and made a good living at the same time.
I don't really think Java is a landmine, but as AC argued, it makes sense to treat it carefully.
I think the big danger of adopting it too early would have been that the Be msg would get confused and lose focus. So its more of a marketting msg than even a developer msg. Now that Be has convinced everyone that it has a life of its own, it can afford to let Java into the mix.
Not to mention that I bet that by playing it cool cannny ol' Gasee put himself into a nice negotiating position as far as licensing and development help from Sun go. If you think about it, it'd be pretty ridiculous for Be's appliance offereing not to have Java, and so from this perspective, Gasee did a superb job of playing hard to get.
Don't know anything about KDE's implementation, to be honest. I have to point out to that while Be's APIs are relatively simple, they're also incredibly powerful and well and consitently thought out. I would hope KDE's the same way, but I just don't know. You can get all the API docs and stuff at the Be site. It might be worth it to take a glance.
I also like ObjC. Much better than C++. Actually, a real decent language. Too bad it never really caught on; but I actually know a lot fo people that work with it today for specialized stuff. I wonder if C++ might be a little better for low level stuff, but I don't know.
What it comes down I guess is that any language can be used or abused. Good C++ code is still better than crappy Java code!
"Does this mean that you do not feel that JBED can work well?"
Nope. Perhaps I sacrificed a little clarity for brevity. What I should have siad is that overall you probably couldn't really build a decent responding desktop style OS in Java right now. While Java has great computatoional speed in some areas (I get better than C++ for some of my stuff) it still may not have the broad perfromance neccessary, esp. when dealing with graphics and device drivers, and any implementation that we did now that was all Java would take a long time to come to fruition, whereas BeOS is here now.
As someone who has done a fair amount of Swing development, I don't think you could quite do a full-on desktop that would have the perfromance that users expected and still leave some cycles for other tasks. Of course, if you take away the intermediate levels between the vm and the graphics, etc.. you might very well get good performance.
As far as JBED goes, remember that !(Real Time lt==gt Performance) All a Real Time system does is provide guaranteed performance; it doesn't specify what that perfromance actually is. One could call a system "real-time" that did one add op every second, as long as it refused to guarantee that it could do the op any faster.
That said, I wouldn't be surprised if JBED does give quite decent performance, but it is not a desktop OS. I think small-embedded type OSs are a natural fit for Java right now. Ironically, embedded systems don't typically have the same performance challenges that desktop systems do. My impression is that they're typically designed such that the processor is underutilizzed if anything. The time taken to display a couple of lines of text on a cell-phone for instance is infintesimal compared to rendering a busy workstation desktop. At the same time, you never want that cellphone to take any perceived time to display something, so you design with that in mind. Its a much more controllable situation.
But yes, I do think we'll see more and more Java to the metal kind of implementations as the world rolls on, I just don't expect to see them tmrw. I also think diversity is healthy so I'd like to see and support both.
This is great news. We've been memeber of the BeOS developers program for a couple of years, but I haven't been able to do any 'real' development on it because our code has to be able to run on other machines.
BeOS itself is so slick..once you've used it, it is such a drag (literallly) to go back to NT, or even Linux. There is _no_ refresh lag, resources are used up smoothly and predicatbly, and the whole thing boots instantaneously (at least when compared to NT!) Also, with real file jouranling, you can even shut the machine off (not reccommended though!) and not worry that you've just corrupted your entire file system; when you shut down correctly, there is hardly any wait at all.
BeOS also has a commendably simple, clean and elegant API implementation. Actually, its the cleanest, most rational C++ implementation I've seen. (Hmmm, maybe thats not saying much. I guess I should say its a great implementation regardless of language.) Interestingly, they have explicitly avoided all the C++ 'enhancements' like STL, exceptions model, etc.., etc.. and are really writing to the bone; they're really using C++ like C with simple object extensions which is the only way I'd ever use it.
I love Java, but I'm not a completly fanatical idiot about it. Right now, you simply couldn't implement the low-level parts of a decent OS in Java. If this is a good Java implementation, with the BeOS finegrained threading, clean graphics mdel, etc.., this has the potential to be the best Java implementation of them all. Truly the best of both worlds.
Still trying hard to keep my expecations down so if its a half-assed implementaiton I won't Be Bummed..
..accorrding to findings, pp. 23