Re:Java is finished for most open source work
on
Java vs .NET
·
· Score: 1
It should be noted, however, that it's still really hard to write stand-alone SWT apps. I do *really* look forward to when that's possible, though.:)
Re:Java is finished for most open source work
on
Java vs .NET
·
· Score: 3, Insightful
Well, I suggest you actually read the licenses you agree to when you download Java and the Java specification from Sun. Both the language and its implementations are tightly controled, in a way that practically no other language or platform is.
In that you can't create an implementation and call it "Java" without Sun giving you the okay. BFD. I can understand why Sun would do this... look what happened with Microsoft and their "implementation" of Java. That doesn't stop anyone from creating their own, compatible language, VM, etc, and calling it Espresso or something. Again, gcj, GNU classpath, etc, proves this to be the case.
People can't. Read the licenses.
Tell that to the FSF... not to mention IBM.
You are damned right it's hard. But it's not technically hard, it's "hard" in the same sense that a Windows clone like Wine is hard: Java is a platform controled by a consortium, driven in such a way that people can't successfully create third party implementations.
No, WINE is hard because the APIs are a moving target and most of them are undocumented. Moreover, WINE is hard because the Win32 APIs are NOT controlled by a consortium! They're controlled by a single entity, and hence are subject to change, revision, addition, etc, without anyone knowing about it, making it even harder to write a compatible version.
Java, OTOH, is highly stable and moves slowly (as has been noted by many others in the comments for this article) which in fact makes it *easier* to create competing versions, since you don't have a rapidly changing platform to remain compatible with.
What planet are you from? Maybe there is some form of "JVM" on their machines, but it's completely unpredictable what version it is (1.1? Microsoft? 1.3? 1.4?). I used to be able to use Java applets on my site (for SSH and other services), but that's become pretty much pointless these days.
Okay, let me revise my statement a little. Most people who use free/OSS platforms have a reasonably modern JDK installed. OTOH, you're absolutely right, if you're attempting to target Windows (although you can blame Microsoft for that). If you can name another language that approaches Java's install base (particularly amongst OSS folks) but meets the same needs, I might conceed the point.
I'd select Java for cross-platform development as well. But most of my development is not cross-platform.
So then who cares about Java and it's platform support? That was, after all, one of your objections regarding Java (and, no, I highly doubt Sun would drop support for either Linux or Windows).
The KDE developers were headed for a legal disaster: they had created a desktop whose license was incompatible with the toolkit they were using. If Troll Tech hadn't changed their license to dual-license, KDE would have been dead, in particular because it's not clear they could have legally created an open-source version of Qt. The KDE project was oblivious and ignorant of the license problems they were getting themselves into. And so are many open source Java developers.
Heh, please... those licensing issues were purely philosophical. It certainly didn't stop MANY people from using the platform... after all, those issues were well known right from the beginning. As a result, I *highly* doubt the project would have died. Hell, they could have just changed to a BSD license and been done with it. *shrug*
However, that's completely irrelevant. Why? Because there are no "licensing problems" with Java. This is just something you seem to have invented! I can create all the Java apps I like... Sun's license on the JDK and their control of it plays no part in the license I choose for my software. Basically, it's a *completely* different issue. In this case, my project can be as "free" as I want... it's simply the platform which (arguably) isn't.
Now, to make the comparison more apt, let's say TrollTech's license was compatible, but
Re:Java is finished for most open source work
on
Java vs .NET
·
· Score: 2, Insightful
So do dozens of other languages. Among them, Java is pretty mediocre and has numerous limitations and design flaws. For example, the array conformance is just broken, and so is the genericity implementation. The only thing Java has going for it over the competition is name recognition.
And ubiquity. There are, IMHO, very few languages which fill the same niche as Java while still being common. And, no, I don't think Python, et al, fit the bill.
Besides, if you are willing to put up with the patent and intellectual property problems that Java has, why not simply use Mono?
Simple question: huh? I've *never* heard of these "patent and intellectual property problems" you're referring to. Ever.
Anyway, I don't use Mono because, again, it's not ubiquitous. Most people have a JVM on their boxes these days. Can you say the same about Mono? Hardly... and yes, I know you can compile to executable, but I'm unaware if those executables are standalone (versus requiring various runtime libraries to operate, ala VB). Other than that, the only reason I don't use Mono is purely philosophical. *shrug* And no, I don't have those problems with Sun.
You will give a damn, when Sun goes out of business, or when Sun does something stupid with Java2D (like basing it on OpenGL), or when they don't support your favorite operating system.
Please... it's not like people can't create their own competing version. GNU gcj, anyone? Kaffee, on the VM side? There's just been no push toward it, mostly because it's *hard*.
As for OS support, Java is still *far FAR* superior to.NET, so given the choice, I'd select Java for cross-platform development.
Yes, and there are lots of people who also write open source Windows software. But they don't represent the mainstream of open source, and as a general approach to open source, it's self-defeating in the long run.
Talk about a non-point. They don't represent the mainstream because it's not "cool" to develop OSS Windows apps. Moreover, the people who'd use your neato free Windows tool won't bother, since they're running a free OS. AND, Windows doesn't have as much room for new development, since there are a gazillion apps already available for the platform (hence no sense of a "niche" to fill).
Besides, KDE developers certainly represented a segment of "main stream... open source", and they had no problems using a closed toolkit. And their project is quite healthy, and I'd contend they'd still be doing just fine, even if Qt hadn't opened their toolkit...
Re:You think Java's the bee's testicles?
on
Java vs .NET
·
· Score: 1
Oh, I've already done that... in fact, I wrote a small FastCGI implementation in Smalltalk to facilitate some ideas I had for using Smalltalk as a server-side language (which, BTW, is being done by a number of people). It's quite nice. However, Smalltalk is NOT FAST! That's not to say Java is fast, but Smalltalk is really bloody slow. The commercial Smalltalks are better, but Squeak is a dog (the main reason it's able to pull of multimedia is by tying C extensions in using Squeaks (excellent) extension mechanisms).
As a result, I feel Smalltalk fills a different niche than Java. Also, because (at least, Squeak) uses an image-based model it's not really appropriate for application development, since you can't write stand-alone apps that aren't huge. AND, it's not exactly as ubiquitous as Java. Oh, and it's also not easily portable across implementations.
OTOH, Smalltalk is a fantastic language (in fact, I love it more than any other OOP language I've touched to date). Frankly, if I could have Java's model (stand-alone apps, etc) with the Smalltalk language, I'd be in heaven.:)
Re:Java is finished for most open source work
on
Java vs .NET
·
· Score: 3, Insightful
Yeah, except that some people actually just, *gasp*, *like* to program in Java. It provides a nice, strongly typed, compiled, object oriented language (a niche none of those scripting languages fill) that's available almost anywhere while being FAR nicer to develop in than either C (too low-level and error-prone) or C++ (to damned friggin' ugly).
Moreover, Java provides a massive class library which makes it easy to write new applications without having to reinvent the wheel every time (and yes, I know about the STL, but it's still not entirely standard, it creates bloody HUGE application binaries, it takes FOREVER to compile, it's *impossible* to debug compile errors, and it isn't even as extensive as the Java APIs!)
Now, I would agree that Java is pretty well toast on the desktop, barring a toolkit revolution. I certainly wouldn't write an app using Swing. However, for other types of applications (like, say, Freenet), it can do an excellent job. And if you go with SWT, you can write some pretty damn nice GUI apps (just look at Eclipse, which is written entirely in Java).
But to dismiss Java out of hand because it doesn't mesh with your and some other zealots ideals is pretty narrow minded. Personally, I don't care if my app is somehow "tied" to Sun. As long as I can release my source code, who gives a damn if it runs on a proprietary substrate. That certainly didn't bother the KDE guys (or their users) back when Qt was non-free. Sure, there were people who complained, but there were also plenty of people who didn't give a damn (myself included).
Re:If I were to choose...
on
Java vs .NET
·
· Score: 1
I won't reiterate the other points about cost of development, but another point is the error rate introduced by using a lower-level language. Without numbers in front of me, from my own experience, I've found that bugs are FAR more common in C/C++. This is primarily due to the need for explicit memory management (and the heisenbugs that can introduce), as well as the lack of strict typing (although C++ fixes that to some degree).
In addition, a language like Java or.NET provides a developer with a much larger set of support libraries which can cut down on development time significantly without having to rely on a ton of third party packages.
That is, hands down, one of the dumbest posts I've ever read. I actually feel like I lost something by reading it... Congrats, you've proven to me that anyone really can use a computer.
A few hours. You honestly think it'd take just a few hours to 1) take all the calls from a bunch of people who's net connection are shut down and 2) instruct them (and potentially walk them through) how to disinfect and patch their systems? Really. Frankly, I find it remarkable how naive you are...
I absolutely agree with the original poster... if some idiot doesn't patch his box, I shouldn't suffer. If anything, set up rules at the upstream router to shut down his, and only his, connection (hell, you could automate this if you wanted). But don't you dare touch mine.
I know it was just bad writing to try and mask an obvious continuity problem,
I'd say the writing was, if anything, pretty obviously the writers poking fun at the Star Trek franchise and some of it's continuity issues. Frankly, I thought it was hilarious, which I suspect was the point.
Gah! I *completely* disagree. First of all, they totally fucked up the timeline with that episode. In TNG, the Enterprise D crew is supposed to be the first humans to encounter The Borg. The fact that the first Enterprise supposedly ran into them runs completely counter to this timeline.
Worse yet, they yet again decided fall back on The Borg as a plot device, evidently because they couldn't come up with a more intriguing enemy-type to deal with. It was pathetic... didn't the writers have anything better to do than create yet-another-Borg-episode?
This is primarily because Thunderbird has to be shipped with all the accompanying Mozilla technologies. ie, Gecko, XPCOM, etc, etc. Why is this? It's because the GRE isn't done. The GRE is an attempt to create a separate runtime library distribution containing just the core Mozilla libraries which other Mozilla-based projects can leverage, rather than having each project come with it's own copy. Not only would this save disk space, but also runtime memory usage as well (since Thunderbird and Firebird could use the same shared libraries).
Which is, typically, what happens in Science. Although, it should be noted that rarely do you need to completely toss out prior theories. For example, traditional Newtonian physics works just fine in most cases, even though General Relativity is, in fact, more accurate. Similarly, just because String Theory is a more fundamental theory, doesn't mean that Quantum tbeory (and the ideas it's based on) isn't still useful.
Absolutely. For example, gravitational lensing, one of the predictions of General Relativity was demonstrated by observing a star as it was eclipsed by the Sun.
1) They aren't any more effective than human personal, and unless they have an unrealistically low false-postive rate, they actually generate more work than they're worth
Hardly. You expect every security officer to memorize the faces of all the suspected terrorists in the US? Moreover, you're saying this would be *less* error-prone? You and I both know that humans are subject to bias. Hell, just imagine the racial profiling that would happen:
Security Officer: "Hmm... that guy sorta looks like one of those terrorists on the watch list, I think. Eh, he's middle-eastern. Close enough. bring 'im in."
2) As anyone who's been to (or worked in) a DMV knows, there's basically 2 human reponses to computerized systems like this: a) you ignore the computer whenever it disagrees with you or b) you always obey the computer, no matter what.
This is a training problem. The tool is still useful, even if your personnel are idiots. However, if they're properly trained to understand the tool's role in their job and how to use it effectively, this shouldn't be a problem. Besides, I'd expect the personnel working for airport security would be *slightly* more intelligent/diligent than your average DMV clerk.
Having seen the picture (which, BTW, looks more like a Decepticon to me:), I still find it amusing. Why? Because it's interesting how the human brain is so capable of recognizing faces/shapes/etc that it can pull patterns from randomness.
On a less amusing note, it's also interesting how seemingly rational people can become so unbelievably superstitious the minute something happens which shakes their world-view, causing them to grapple for something, anything, to explain what is happening in the world around them. Frankly, I think it's an interesting insight into the human mind.
Yes! Mod this up! These systems aren't meant to replace the security personnel already in the airport. It's simply a tool to make their jobs a little easier...
Funny how the tinfoil-hat crowd automatically assumes the worst when something like this pops up.
Yes, but that's because *measuring* gait and mannerism is infinitely difficult. First, how do you define these things? Then, how do you measure them given you are limited to a few 2-D cameras in a rather large space at a distance from the subject. Hell, getting a good face shot is hard enough. Now you want them to be able to measure the person's stride length and whether or not they talk with their hands?
As for needing human verification, these types of systems aren't intended to be fully automated. I expect they send alerts to security personnel who then verify the results and act accordingly.
Or you could realize that a photo doesn't contain any kind of secured information, and is thus far easier to forge than a passport with an untamperable component which contains the issuee's exact biometrics.
And yes, you could just store the photo, but biometrics, one-on-one, are fairly reliable and difficult to fool (in theory), unlike a passport agent who may not be good at telling apart, say, Southeast Asian men.
Also, here's another interesting point, possibly only tangentially related: I understand that, in trials, eyewitnesses are not considered particularly reliable (including but not limited to facial recognition, presumably).
Yes, but a human's memory is fallable in general, and can be further distorted by emotion, prejudice, and any number of other factors. That is why eye witness accounts aren't reliable. A computer, however, has a perfect memory and is not susceptible to any sort of bias.
No, the Aussie government thinks "facial recognition as the international biometric" is a good idea (as opposed to, say, retinal scans, finger prints, voice recognition, etc). The "move to implementation" part is what this article is regarding, and even that's not strictly true... the article you mention is about secure passports, not active scanning in airports.
Satan's Face in the WTC smoke? Wow, I never heard of that one... that's pretty funny.:) Heck, it's right up there with seeing Jesus' face in a burrito.
Okay, the entire group of machines is the cluster. If a subset of those machines are used to perform some task, they can rightly be referred to as a Beowulf cluster on their own. However they are stil "part of" the larger cluster. This is likely what the grandparent poster was referring to.
Similarly, if I have a hundred people, that's a crowd. If I take 50 of those people, they are still a crowd, however they are also a subset, and hence a "part of", the original crowd.
Not at all. I'm saying the commercial sector is as insecure, if not more so, without providing any benefit to the consumer.
Now your point about a "web-of-trust" is a reasonable one. However, I should point out that it simply isn't scalable. The reality is that without an organized hierarchy, trying to set up all the trust relationships is simply infeasible... well, assuming that the ISPs, media companies, etc, don't continue to merge.:)
Now, I agree the process behind it should be open. Perhaps government regulation of a number of private entities would be a middle ground? Sorta FDIC-style?
Now, this I could live with. However, I must point out that the government hasn't exactly proven itself when it comes to regulating public entities (this is especially true in the current, shall we say, business-friendly government in office). This is blatantly obvious in the case of ICANN (which is, only now, getting a *little* pressure, and even then, only a little), and don't get me started on privatization of other industries (look how well the privatization of power worked for California). The fact is, the US government seems quite reluctant to regulate private industry (hell, just look at the behaviour of the FTC lately), and until that changes, I'm not sure you can trust private companies with something as vital (and potentially "abusable") as identity verification.
'course, one could ask if you could trust the government in this case. However, I would contend that, with the lack of a profit motive, it's at least more likely that the government could be trusted with this kind of responsibility as compared to the private sector.
And how is putting this responsibility in the hands of the private sector any better? It's already been proven that the commercial sector has a great deal of difficulty with security (MS is the obvious example, but look at all the companies who are hacked and their credit card databases stolen?) At least in the hands of the government the process is (hopefully) transparent and auditable (unlike, say, ICANN), not to mention accessible to the common citizen (ie, I could get a cert without having to pay ridiculous fees).
Slashdot Mirror
It should be noted, however, that it's still really hard to write stand-alone SWT apps. I do *really* look forward to when that's possible, though. :)
Well, I suggest you actually read the licenses you agree to when you download Java and the Java specification from Sun. Both the language and its implementations are tightly controled, in a way that practically no other language or platform is.
In that you can't create an implementation and call it "Java" without Sun giving you the okay. BFD. I can understand why Sun would do this... look what happened with Microsoft and their "implementation" of Java. That doesn't stop anyone from creating their own, compatible language, VM, etc, and calling it Espresso or something. Again, gcj, GNU classpath, etc, proves this to be the case.
People can't. Read the licenses.
Tell that to the FSF... not to mention IBM.
You are damned right it's hard. But it's not technically hard, it's "hard" in the same sense that a Windows clone like Wine is hard: Java is a platform controled by a consortium, driven in such a way that people can't successfully create third party implementations.
No, WINE is hard because the APIs are a moving target and most of them are undocumented. Moreover, WINE is hard because the Win32 APIs are NOT controlled by a consortium! They're controlled by a single entity, and hence are subject to change, revision, addition, etc, without anyone knowing about it, making it even harder to write a compatible version.
Java, OTOH, is highly stable and moves slowly (as has been noted by many others in the comments for this article) which in fact makes it *easier* to create competing versions, since you don't have a rapidly changing platform to remain compatible with.
What planet are you from? Maybe there is some form of "JVM" on their machines, but it's completely unpredictable what version it is (1.1? Microsoft? 1.3? 1.4?). I used to be able to use Java applets on my site (for SSH and other services), but that's become pretty much pointless these days.
Okay, let me revise my statement a little. Most people who use free/OSS platforms have a reasonably modern JDK installed. OTOH, you're absolutely right, if you're attempting to target Windows (although you can blame Microsoft for that). If you can name another language that approaches Java's install base (particularly amongst OSS folks) but meets the same needs, I might conceed the point.
I'd select Java for cross-platform development as well. But most of my development is not cross-platform.
So then who cares about Java and it's platform support? That was, after all, one of your objections regarding Java (and, no, I highly doubt Sun would drop support for either Linux or Windows).
The KDE developers were headed for a legal disaster: they had created a desktop whose license was incompatible with the toolkit they were using. If Troll Tech hadn't changed their license to dual-license, KDE would have been dead, in particular because it's not clear they could have legally created an open-source version of Qt. The KDE project was oblivious and ignorant of the license problems they were getting themselves into. And so are many open source Java developers.
Heh, please... those licensing issues were purely philosophical. It certainly didn't stop MANY people from using the platform... after all, those issues were well known right from the beginning. As a result, I *highly* doubt the project would have died. Hell, they could have just changed to a BSD license and been done with it. *shrug*
However, that's completely irrelevant. Why? Because there are no "licensing problems" with Java. This is just something you seem to have invented! I can create all the Java apps I like... Sun's license on the JDK and their control of it plays no part in the license I choose for my software. Basically, it's a *completely* different issue. In this case, my project can be as "free" as I want... it's simply the platform which (arguably) isn't.
Now, to make the comparison more apt, let's say TrollTech's license was compatible, but
So do dozens of other languages. Among them, Java is pretty mediocre and has numerous limitations and design flaws. For example, the array conformance is just broken, and so is the genericity implementation. The only thing Java has going for it over the competition is name recognition.
.NET, so given the choice, I'd select Java for cross-platform development.
... open source", and they had no problems using a closed toolkit. And their project is quite healthy, and I'd contend they'd still be doing just fine, even if Qt hadn't opened their toolkit...
And ubiquity. There are, IMHO, very few languages which fill the same niche as Java while still being common. And, no, I don't think Python, et al, fit the bill.
Besides, if you are willing to put up with the patent and intellectual property problems that Java has, why not simply use Mono?
Simple question: huh? I've *never* heard of these "patent and intellectual property problems" you're referring to. Ever.
Anyway, I don't use Mono because, again, it's not ubiquitous. Most people have a JVM on their boxes these days. Can you say the same about Mono? Hardly... and yes, I know you can compile to executable, but I'm unaware if those executables are standalone (versus requiring various runtime libraries to operate, ala VB). Other than that, the only reason I don't use Mono is purely philosophical. *shrug* And no, I don't have those problems with Sun.
You will give a damn, when Sun goes out of business, or when Sun does something stupid with Java2D (like basing it on OpenGL), or when they don't support your favorite operating system.
Please... it's not like people can't create their own competing version. GNU gcj, anyone? Kaffee, on the VM side? There's just been no push toward it, mostly because it's *hard*.
As for OS support, Java is still *far FAR* superior to
Yes, and there are lots of people who also write open source Windows software. But they don't represent the mainstream of open source, and as a general approach to open source, it's self-defeating in the long run.
Talk about a non-point. They don't represent the mainstream because it's not "cool" to develop OSS Windows apps. Moreover, the people who'd use your neato free Windows tool won't bother, since they're running a free OS. AND, Windows doesn't have as much room for new development, since there are a gazillion apps already available for the platform (hence no sense of a "niche" to fill).
Besides, KDE developers certainly represented a segment of "main stream
Oh, I've already done that... in fact, I wrote a small FastCGI implementation in Smalltalk to facilitate some ideas I had for using Smalltalk as a server-side language (which, BTW, is being done by a number of people). It's quite nice. However, Smalltalk is NOT FAST! That's not to say Java is fast, but Smalltalk is really bloody slow. The commercial Smalltalks are better, but Squeak is a dog (the main reason it's able to pull of multimedia is by tying C extensions in using Squeaks (excellent) extension mechanisms).
:)
As a result, I feel Smalltalk fills a different niche than Java. Also, because (at least, Squeak) uses an image-based model it's not really appropriate for application development, since you can't write stand-alone apps that aren't huge. AND, it's not exactly as ubiquitous as Java. Oh, and it's also not easily portable across implementations.
OTOH, Smalltalk is a fantastic language (in fact, I love it more than any other OOP language I've touched to date). Frankly, if I could have Java's model (stand-alone apps, etc) with the Smalltalk language, I'd be in heaven.
Yeah, except that some people actually just, *gasp*, *like* to program in Java. It provides a nice, strongly typed, compiled, object oriented language (a niche none of those scripting languages fill) that's available almost anywhere while being FAR nicer to develop in than either C (too low-level and error-prone) or C++ (to damned friggin' ugly).
Moreover, Java provides a massive class library which makes it easy to write new applications without having to reinvent the wheel every time (and yes, I know about the STL, but it's still not entirely standard, it creates bloody HUGE application binaries, it takes FOREVER to compile, it's *impossible* to debug compile errors, and it isn't even as extensive as the Java APIs!)
Now, I would agree that Java is pretty well toast on the desktop, barring a toolkit revolution. I certainly wouldn't write an app using Swing. However, for other types of applications (like, say, Freenet), it can do an excellent job. And if you go with SWT, you can write some pretty damn nice GUI apps (just look at Eclipse, which is written entirely in Java).
But to dismiss Java out of hand because it doesn't mesh with your and some other zealots ideals is pretty narrow minded. Personally, I don't care if my app is somehow "tied" to Sun. As long as I can release my source code, who gives a damn if it runs on a proprietary substrate. That certainly didn't bother the KDE guys (or their users) back when Qt was non-free. Sure, there were people who complained, but there were also plenty of people who didn't give a damn (myself included).
I won't reiterate the other points about cost of development, but another point is the error rate introduced by using a lower-level language. Without numbers in front of me, from my own experience, I've found that bugs are FAR more common in C/C++. This is primarily due to the need for explicit memory management (and the heisenbugs that can introduce), as well as the lack of strict typing (although C++ fixes that to some degree).
.NET provides a developer with a much larger set of support libraries which can cut down on development time significantly without having to rely on a ton of third party packages.
In addition, a language like Java or
That is, hands down, one of the dumbest posts I've ever read. I actually feel like I lost something by reading it... Congrats, you've proven to me that anyone really can use a computer.
Umm... I believe the phrase is "he's coming right *for* us". Sheesh. ;)
A few hours. You honestly think it'd take just a few hours to 1) take all the calls from a bunch of people who's net connection are shut down and 2) instruct them (and potentially walk them through) how to disinfect and patch their systems? Really. Frankly, I find it remarkable how naive you are...
I absolutely agree with the original poster... if some idiot doesn't patch his box, I shouldn't suffer. If anything, set up rules at the upstream router to shut down his, and only his, connection (hell, you could automate this if you wanted). But don't you dare touch mine.
Regarding this comment:
I know it was just bad writing to try and mask an obvious continuity problem,
I'd say the writing was, if anything, pretty obviously the writers poking fun at the Star Trek franchise and some of it's continuity issues. Frankly, I thought it was hilarious, which I suspect was the point.
Gah! I *completely* disagree. First of all, they totally fucked up the timeline with that episode. In TNG, the Enterprise D crew is supposed to be the first humans to encounter The Borg. The fact that the first Enterprise supposedly ran into them runs completely counter to this timeline.
Worse yet, they yet again decided fall back on The Borg as a plot device, evidently because they couldn't come up with a more intriguing enemy-type to deal with. It was pathetic... didn't the writers have anything better to do than create yet-another-Borg-episode?
Woah, wait... HIBT?
This is primarily because Thunderbird has to be shipped with all the accompanying Mozilla technologies. ie, Gecko, XPCOM, etc, etc. Why is this? It's because the GRE isn't done. The GRE is an attempt to create a separate runtime library distribution containing just the core Mozilla libraries which other Mozilla-based projects can leverage, rather than having each project come with it's own copy. Not only would this save disk space, but also runtime memory usage as well (since Thunderbird and Firebird could use the same shared libraries).
Which is, typically, what happens in Science. Although, it should be noted that rarely do you need to completely toss out prior theories. For example, traditional Newtonian physics works just fine in most cases, even though General Relativity is, in fact, more accurate. Similarly, just because String Theory is a more fundamental theory, doesn't mean that Quantum tbeory (and the ideas it's based on) isn't still useful.
Absolutely. For example, gravitational lensing, one of the predictions of General Relativity was demonstrated by observing a star as it was eclipsed by the Sun.
1) They aren't any more effective than human personal, and unless they have an unrealistically low false-postive rate, they actually generate more work than they're worth
Hardly. You expect every security officer to memorize the faces of all the suspected terrorists in the US? Moreover, you're saying this would be *less* error-prone? You and I both know that humans are subject to bias. Hell, just imagine the racial profiling that would happen:
Security Officer: "Hmm... that guy sorta looks like one of those terrorists on the watch list, I think. Eh, he's middle-eastern. Close enough. bring 'im in."
2) As anyone who's been to (or worked in) a DMV knows, there's basically 2 human reponses to computerized systems like this: a) you ignore the computer whenever it disagrees with you or b) you always obey the computer, no matter what.
This is a training problem. The tool is still useful, even if your personnel are idiots. However, if they're properly trained to understand the tool's role in their job and how to use it effectively, this shouldn't be a problem. Besides, I'd expect the personnel working for airport security would be *slightly* more intelligent/diligent than your average DMV clerk.
Having seen the picture (which, BTW, looks more like a Decepticon to me :), I still find it amusing. Why? Because it's interesting how the human brain is so capable of recognizing faces/shapes/etc that it can pull patterns from randomness.
On a less amusing note, it's also interesting how seemingly rational people can become so unbelievably superstitious the minute something happens which shakes their world-view, causing them to grapple for something, anything, to explain what is happening in the world around them. Frankly, I think it's an interesting insight into the human mind.
Yes! Mod this up! These systems aren't meant to replace the security personnel already in the airport. It's simply a tool to make their jobs a little easier...
Funny how the tinfoil-hat crowd automatically assumes the worst when something like this pops up.
Yes, but that's because *measuring* gait and mannerism is infinitely difficult. First, how do you define these things? Then, how do you measure them given you are limited to a few 2-D cameras in a rather large space at a distance from the subject. Hell, getting a good face shot is hard enough. Now you want them to be able to measure the person's stride length and whether or not they talk with their hands?
As for needing human verification, these types of systems aren't intended to be fully automated. I expect they send alerts to security personnel who then verify the results and act accordingly.
Or you could realize that a photo doesn't contain any kind of secured information, and is thus far easier to forge than a passport with an untamperable component which contains the issuee's exact biometrics.
And yes, you could just store the photo, but biometrics, one-on-one, are fairly reliable and difficult to fool (in theory), unlike a passport agent who may not be good at telling apart, say, Southeast Asian men.
Also, here's another interesting point, possibly only tangentially related: I understand that, in trials, eyewitnesses are not considered particularly reliable (including but not limited to facial recognition, presumably).
Yes, but a human's memory is fallable in general, and can be further distorted by emotion, prejudice, and any number of other factors. That is why eye witness accounts aren't reliable. A computer, however, has a perfect memory and is not susceptible to any sort of bias.
No, the Aussie government thinks "facial recognition as the international biometric" is a good idea (as opposed to, say, retinal scans, finger prints, voice recognition, etc). The "move to implementation" part is what this article is regarding, and even that's not strictly true... the article you mention is about secure passports, not active scanning in airports.
Satan's Face in the WTC smoke? Wow, I never heard of that one... that's pretty funny. :) Heck, it's right up there with seeing Jesus' face in a burrito.
Okay, the entire group of machines is the cluster. If a subset of those machines are used to perform some task, they can rightly be referred to as a Beowulf cluster on their own. However they are stil "part of" the larger cluster. This is likely what the grandparent poster was referring to.
Similarly, if I have a hundred people, that's a crowd. If I take 50 of those people, they are still a crowd, however they are also a subset, and hence a "part of", the original crowd.
You are claiming government security is better?
:)
Not at all. I'm saying the commercial sector is as insecure, if not more so, without providing any benefit to the consumer.
Now your point about a "web-of-trust" is a reasonable one. However, I should point out that it simply isn't scalable. The reality is that without an organized hierarchy, trying to set up all the trust relationships is simply infeasible... well, assuming that the ISPs, media companies, etc, don't continue to merge.
Now, I agree the process behind it should be open. Perhaps government regulation of a number of private entities would be a middle ground? Sorta FDIC-style?
Now, this I could live with. However, I must point out that the government hasn't exactly proven itself when it comes to regulating public entities (this is especially true in the current, shall we say, business-friendly government in office). This is blatantly obvious in the case of ICANN (which is, only now, getting a *little* pressure, and even then, only a little), and don't get me started on privatization of other industries (look how well the privatization of power worked for California). The fact is, the US government seems quite reluctant to regulate private industry (hell, just look at the behaviour of the FTC lately), and until that changes, I'm not sure you can trust private companies with something as vital (and potentially "abusable") as identity verification.
'course, one could ask if you could trust the government in this case. However, I would contend that, with the lack of a profit motive, it's at least more likely that the government could be trusted with this kind of responsibility as compared to the private sector.
but the modes of failure could be catastrophic.
And how is putting this responsibility in the hands of the private sector any better? It's already been proven that the commercial sector has a great deal of difficulty with security (MS is the obvious example, but look at all the companies who are hacked and their credit card databases stolen?) At least in the hands of the government the process is (hopefully) transparent and auditable (unlike, say, ICANN), not to mention accessible to the common citizen (ie, I could get a cert without having to pay ridiculous fees).