The protests were about a risk, that is, an accident that could have occurred with a certain probability.
Mod parent up.
Even if you disagree with these protestors, be glad that they are protesting. If there's no political pressure to "get it right", then somebody somewhere will get it wrong, with far-ranging consequences...
the movie ditches the one thing that made the series great: specialists cooperating
Good call... I had never quite fingered what was disappointing about the first MI movie, but I think that's it: the entire team is gruesomely killed off in the first couple of minutes.
And oh yeah, the product placement of powerbooks was annoying...
Why should I accept mail claiming to be from hotmail that originates from a Starbucks somewhere?
because many potential customers (like me) use their free web based account wherever they happen to be on what ever computer is nearby.
When you use your Yahoo web-based account, the email still originates on Yahoo's mail servers, even if you are sitting in a Starbucks. DomainKeys and SPF work on the level of the mail protocal itself and would not affect you.
This would actually take risk out of your business... a shoulder-surfing hacker next to you in Starbucks would not be able to forge emails from nflnk@yahoo.com without compromising something in Yahoo's system [at additional time/cost/risk to the hacker].
If the companies are also releasing the code to the purchasers, then the companies aren't violating or "ripping off" the open source apps.
If you're just re-branding an open source app, that's somewhat lame, even if you are respecting the license agreement. Legal dues aside, the are still some social dues to be paid.
True, but a legitimate question is "should corporations be allowed to hold monopolies on culture?". I don't think copyright should be abolished but this "world leader in the production of creative works" BS ignores the fact that mass-produced culture is not very authentic culture. Weaker copyrights might make a stronger society... it's an option worth exploring anyways.
That day 0 copy of the next StarWars can and should land you in jail plain and simple.
Go to jail for copying a song? Sending somebody to jail is not cheap. Not only do you have to pay the (relatively trivial) cost of incarceration, you also lose that individual's productive input to society. Also, the more people you put in jail, the more construction activity must go into jails (instead of say, libraries or offices). The more people in jail, the more guards/wardens/police/judges/lawyers you need. Human potiential that could be put towards space, towards knowledge, towards medicine...
As a general rule, when "everybody" is breaking the law, it's time to revise the law, not throw more people into jail.
how wrong it is to think that the Christian God would not reveal evidence of His existence to men... Romans 1:18-20
The Bible certaintly conveys that impression. Consider Luke 11:9-10: "...ask and it will be given to you; seek and you will find; knock and the door will be opened to you. For everyone who asks receives; he who seeks finds; and to him who knocks, the door will be opened.".
My original comment was facetious... I was pointing out that the distinct lack of evidence contradicts what we would expect the Christian God to do if he existed.
I'm not saying that this *disproves* the existence of the Christian God (or any other gods), but it is consistent with what we would expect from a purely social institution.
I believe God has kept people from checking out the location. Cool stuff, to me.
Yeah... I thought it was cool when I was a Christian too. Seriously! A kind of curse-of-the-ark thing. It poses a difficult question for theologians though: God loves you enough to torture his own son to death, but he's pretty meticulous about sealing off any avenue of empirical verification.
Let's face it, business is about money. If some companies believe in "doing no evil", it's simply because, in the opinion of the founders/managers, "doing no evil" is a good way to make money.
Let's face it, life is about pleasure. If some individual believes in "doing no evil", it's simply because, in the opinion of the individual's intellect, "doing no evil" is a good way to make pleasure. On the long term, pissing everybody off is probably not a viable pleasure strategy.
Seems like a good summary to me. Humans can do evil too. Why do many slashdotters seem to almost excuse corporations for being ruthless in the pursuit of their goals? Why it is presumed that the same social criticism we apply to individuals cannot be levied at corporations to pressure them to conform to acceptable business practices?
Perhaps we need some sort of "afterlife for corporations" concept to keep them on the straight and narrow...:-)
I'd just prefer it if search engines would have enhanced rules for the robot.txt file so a webmaster could tell them more specifically how they want to be searched.
Brainstorm a list of all the things you would like to put into a robots.txt (or robots.xml, if we have to go that route).
Find like-minded people who can help you refine and prioritize your list of "wants".
Next, get a developer to help you figure out what an effective specification would look like. The spec should provide a clean syntax that will be easy for non-programming web developers to use, and it should incorporate most or all of your "wants".
If at all possible, the programmer will get involvement from people at Google and other search engines to see how this approach will be perceived by them. Since adoption is voluntary, it is critical that this new standard realistically interoperate with their business... if everybody is going to say "only check my site on Wednesdays", then search engines that adopt the new conventions will have unused bandwidth on other days of the week; this would mean that they would have to purchase a lot more bandwidth to achieve the same level of service.
Desiging a good spec (and communicating the design features that make it a good spec) will be challenging. Once you've done all of this, submit it to W3C if possible and let them take the ball and run with it. If the W3C won't take it, try the IETF. If they don't take it, try a few other standards organization, or whip up your own pretentious one.
While the spec is being "blessed" by the standards organization, have your programmer create a reference implementation and open-source it under something super-friendly to rapid adoption (not the GPL in this case, but something like the clarified BSD license that is compatible with the GPL). At the same time, prepare materials (tutorials, examples, etc.) to help promote your new standard.
IIRC, the XML SAX API is a pretty good example of how a group of people created a standard w/o going through a standards body. That approach might be effective too.
The subject of software piracy is never mentioned because Slashdot is made up of a lot of programmers and developers. Since software piracy would affect them, it's bad, right?
Actually, the BSA has about the same reputation as the MPAA and RIAA here on SlashDot. A big portion of the open source movement itself is about addressing the problems that occur when you start locking information down to create an artifical scarity of knowledge.
I have to admit that I'm much more respectful of software copyrights than I am of movie/music copyrights. I'm willing to concede hypocrisy in this regard, but I do wonder if it is immoral to so restrict the retransmission of "cultural" artifacts.
Personally, I hope to contribute--over the entire span of my life--ideas back to the community in the form of software and art, both in my formal work and as a hobby. In all situations, it would be gratifying to see other people take these ideas and develop them further.
One thought experiment: what if we completely rescinded copyright on future works of software and art? (Yes, yes... this would hurt the GPL too, blah, blah. I don't really think we should rescind copyright... we should just ask "what if"?) As an economic premise, I would say that people would still have needs for entertainment. That's still going to happen somehow, even if the money is not going to movies. And software would still be built based on business need. Hmmm... following the implications of all that might be an extensive exercise.
Humans are kinda in a rush to assign blame, or rather... attribute intention for events with socially relevant consequences. If it impacts us socially, our brains insist that somebody must have intended it, even if we also have a readily available mechanical explanation.
Or at least, that's the premise of Religion Explained. It's a fascinating read.
I'll tell you what... do you have a car with cruise control? Next time you're on the highway, set it, and watch how closely it's sticking to the set speed, even when you go around bends in the road, up and down hills, etc.
Umm... the cruise control on my 2002 Honda Accord is pretty lame, actually. It generally sticks within +/- 3 MPH, but I can do much better if I am controlling it myself.
The part that sucks is that sometimes the cruise control decides to gun the engine when going up a hill. It's unacceptable for a human driver to spike the tachometer to 4500 RPM near the top of the hill just because he's 5 MPH under his desired speed.
Of course, as you point out, the human has more information to work with than the car. Still, I think the grand-parent post has a point: Real Life is notorious for finding unanticipated circumstances to throw at "autonomous" devices.
One day machines will exceed human performance, but it's going to be a long road (pardon the pun). It's not just a matter of having faster reflexes... it's a matter of having superior judgement and reasoning. If a situation starts to unfold on the interstate, how much time should be spent looking for escape paths? If you try breaking hard, will the driver behind you have time to react? Are you fscked already? If so, is there something you can do to minimize damage (e.g,. hit a car instead of a tractor trailer?).
What's going to be impossible is to have BOTH human drivers and machine drivers on the road at the same time. This is a lot more difficult than just having machine drivers, because it requires that machine drivers be able to interpret social cues (and perhaps even fake them). Can the machine analyze another driver's face and tell that they are distracted/tired/busy? How will right-of-way psychology unfold when one of the participants is non-digital? If you're waiting for the traffic signal at a shady intersection and you start to get a Bad Feeling about the thugs approaching the car, will it be smart enough to do a risk evaluation and run the red light?
There's a lot of stuff to think about and be addressed before this is viable...
The reason why Linux, and many of the Open Source solutions that grew up around Linux are so damn difficult is the whole "not invented here" syndrome.
NIH syndrome exist everywhere, not just in OSS. I think both get darwinistically filtered by the marketplace to some degree.
One advantage of OSS is that the software does not have to be easy to learn. (!) Programming is hard, even before you get to UI. There are many different design tradeoffs that must be made. Proprietary vendors are FORCED to make sacrifices for usability up front, early in the product's evolution. Effort towards stability/security/conceptual integrity, etc. might be funnelled into UI because otherwise the product will fail in the market.
By contrast, OSS projects generally start out to satisfy an immediate need or curiosity. They don't have to "conquer the marketplace". Developers start by addressing the core problem. The result is a tool that does the job well. Then, as the tool becomes popular, other people focus on "making it usable". Some great examples include: cdrecord/cdbakeoven, fetchmail/fetchmailconf, mysql/tora, cvs/cervisia. Notice how these are all split at the CLI/GUI border. This split gives people the best of both the CLI and GUI worlds.
This phenomena seems to apply best when "the tool" is not a GUI application where domain logic and presentation are tightly integrated. For instance, sound editors (like Ardour), drawing tools (xfig, gimp), and 3D modelers (blender) can be very difficult to learn if usability is not given a front seat in development. On the other hand, operating systems, system software, databases, conversion utilities, etc. seem to work better if "easy to learn" is not initially a must-have.
Sun says "make this easy to use," and it gets done.
LMAO... how many years did sun shovel CDE down people's throats? A single driving vision can be a great thing, agreed, but not if people have the wrong vision.
9/11 style air attack won't work. You'll either get SAMed or the containment building will likely survive the impact.
Hmmm... I dont' buy it. Are their SAM systems actively protecting every nuclear site? Are they operated by staff with proper training? Do they get tested regularly? Can they fire w/o offsite authorization? Are there any relevant NRC regs you can cite?
I am told that the containment buildings for the nuclear plants I support can withstand a direct hint from a volkswagen going 100 MPH. Containment would probably provide some protection against larger objects too, but Boeings are much larger and faster than Volkswagens.:-)
the worst the criminal could do is shut the plant down and perhaps try and disperse the fuel with explosives
I'll one-up you: the worst the criminal could do is hijack a Boeing and crash it into the holding pool, which is not protected by containment like the reactor is. Release of radiation? Yes, probably. Heck... a contractor with access to the right materials might be able to drop an explosive device into the pool and cause some major damage to the plant w/perhaps some radioactive release. And definite public panic.
Armed assault will be met with armed resistance. The minute the attack starts, someone presses the panic button and the cavalry arrives.
Agreed, but the calvary is not going to be all that large. A modest paramilitary force would have a pretty good chance of accomplishing some serious evil. A large, trained paramilitary force would be unstoppable [if they could make it to the gate undetected... the logistics might give them a way weeks ahead of time].
Note that this post is just my speculation, gathered from the snippets I hear about plant security and happenings in the industry. Nuclear power is basically a good thing, but be wary of having too much trust in the system. (How many Pentagon workers killed 2001-09-11 would have bet that the military could have scrammed some jets fast enough to prevent an attack on the pentagon given an hour's notice?)
Why do we need to fix earth? Just so long as it lasts out until we manage to get some other planets started up.
Wow... I know there's a "disposal mentality" in our society, but throwing away this planet once we make it to others is spectacularly careless. We're going to want this planet long after it's no longer necessary for the survival of the species.
The dude who invented this principle phrased it this way (translated from the Latin): "Entities should not be multiplied more than necessary." But what entities are "necessary"? To Ockham, God was a necessary entity, yet you hear Ockham's Razor used to deny the existence of God.
I prefer to think of in terms of trying to curve-fit a set of points on the graph. Curve-fitting requires a sense of bias. For instance, suppose you have 30 (x,y) points from an experiment. If you want to interpolate/extrapolate other values for these points, you need to come up with a model (in the form of an equation). But what sort of model should you use? Something linear? Something parabolic?
It depends on what the graph looks like (and any "special" knowledge or expectations you may have about the situation, but that's a separate topic). For simple phenomena, a straight line will often work fine... just determine the magic numbers with a least-squares-fit.
Of course, those 30 points aren't going to fit exactly on the line. Some will be over, others under. Now suppose you have a colleague that points this discrepancy. He comes up with a new model (by solving for 30 simultaneous equations... go MatLab). The new model fits the data perfectly.
Which model is "correct"? Your 1-degree model that doesn't quite fit the data at hand (though you do have the excuse of measurement error...), or your colleague's 30-degree model, which is full of unexpected dips and surges b/t the collected points [because that's what tends to happen when you force-fit data in the manner described].
This is where Occam steps in... the 1-degree model is more reasonable. But don't rely on a witty quote. Take more measurements... the 1-degree model will continue to be slightly off, but the 30-degree model will be completely off.
Of course, there are phenomena that are better modeled with high-degree polynomials or non-polynomial equations. Choosing the right model is not only a question of science... it's a matter of design, even art. In many cases, the correct model will depend on the end result you are trying to achieve.
As a non-theist, I feel uncomfortable using Occam's Razor in debate. Witty sayings are cheap when it comes to slinging mud at each other's worldviews, and simplicity alone does not make for truth. I can sympathize with the spirit of Occam's statement though.... the idea of "God's Will" is very slippery: anything you observe can be attributed to God by fiat, even if his intentions do not make sense (because God's Will is unquestionable, unknowable). For me, believing in God was like rewriting the equation every time a new data point was found. [[That was my conlusion. Yours may be different... research for yourself if you haven't already.]]
Moving to more neutral ground, I like to apply Occam's Razor to technology too... when evaluating alternative approaches, I look at the simpler one and ask "can we get away with this, or do we *really* need the extra features of the more complex approach?" Having shot myself in the foot on a few occasions trying to make things too fancy, I'm a big fan of keeping it simple. But sometimes simple is wrong or inadequate. And sometimes we have misconceptions about what is actually simpler (for instance, both CLI and GUI can be considered "simpler" depending on the circumstances).
network sniffers are useful for other things as well.
I was surprised to learn that a Windows port of Ethereal was packaged and deployed to our shared apps installation environment. Surprised because we're a pretty large and conservative company, and non-computer types are quick to find the potential downsides to a tool and categorize it as evil before the good side can be seen.
Anyways, Ethereal has been very helpful for exploring a variety of problems w/the different software we fool with. This gives us another way to "peer inside the box" of proprietary applications. For instance: Crystal Reports is a reporting package that has a "subreports" feature. A subreport can be an external file that Crystal pulls in every time you open the main report. The troublesome thing is that Crystal doesn't tell you where the external file is located. But w/Ethereal, we can see the files being pulled over the network. This was an invaluable shortcut to figuring out what the hell was going on.
I appear to be on a switched ethernet segment, so I'm not in danger of breaking anyone's privacy if I were to use the tool in promiscuous mode (like I sometimes did in college:-).
SVG has lost out to more proprietary options.... the market has made its' decision.
Two points:
Possesive "it" does not need an apostrophe.
Flash is the main alternative right now because it long predates SVG in terms of practical software support. No biggie... the Flash format is documented and open. Pre-render the SVG into Flash for web clients that don't know SVG. Meanwhile, SVG is so useful that it's going to find it's way into Internet Explorer, probably through Microsoft and, failing that, through Adobe [actually, Adobe already has an SVG plugin for IE, IIRC].
Great Accomplishment!
on
Gimp Hits 2.0
·
· Score: 3, Insightful
Why all the flamewars b/t Adobe fans and Gimp fans? The news here is that the Gimp team has delivered us a significant upgrade that addresses many long-standing problems of the software (especially in terms of usability). Congratulations!! That's awesome!
If you prefer Adobe to Gimp, that's great too... buy it and use it. Judge the tradeoffs in cost and functionality for yourself and choose the best tool for you.
I just believe that for everything there is a formula and that there's a kinda metaformula, which describes EVERYTHING in this world.... Maybe my theory is wrong.
It is wrong. Godel said so. The problem is that your metaformula describes all truths, but there are more truths than there are possible descriptions.
Re:Bruce Sterlings previous work has been weak
on
The Zenith Angle
·
· Score: 2, Insightful
I always thought Sterling was a bit weak in his storytelling and writing skills.
Try reading Schismatrix or Holy Fire (or even Distraction) as if it was a biography. Instead of looking for plot, look at how technology transforms the protagonist and his society. Instead of looking for one amazing technological device that drives the story (an 'artifact-come-plot-device' or 'Stargate', if you will), look for a glut of radical, unbalancing technologies, each treated with unnerving casualness.
Some of Sterling's ideas are quite ambitious, almost terrifying, and utterly convincing. In Schismatrix, you can almost feel the collective human existence struggling to instantiate itself into something that's far above intelligence. Not all of his works are great, but Schismatrix and Holy Fire were triumphs; they rank at #1 and #3 on my list of all-time-best sci-fi.
Why is it that so many Unix/Linux programs (and everything else, for that matter) do not provide simple screenshots on their products websites?
Open source software websites tend to be better about this than proprietary ones. They also (in the case of web apps) tend to have functional demos that you can test drive w/o the hassle of installing. It's very frustrating when any software maker (open source or proprietary) does not include screenshots, but at least on OSS sites, I can determine this quickly instead of getting lost in layer after layer of marketing crap. It's just another little joy of OSS.
Accountability tools are very good things when properly applied. The hard part is making sure they're not abused.
Some people would make you accountable for your private thoughts and intentions. If they could, they would impose controls and restrictions over how you think. They would strip you of your free will.
History shows that if you create an accountability tool that is cleanly abstracted from policy, the tool will be abused sooner rather than later. The best way to prevent abuse is to encode a (rather liberal) policy into the tools and force the policymakers to deal with problems reactively.
Autonomy is a very good thing when properly allowed to flurish. The hard part is making sure that the autonomus units don't collapse in on themselves by taking advantage of each other.
Is it a bad thing when a CS student wants to put his headphones in the computer lab to block out all the other converstations (sometimes in different languages) around him so that he can concentrate?
IMO, it's the conversations in my own language that tend to disrupt my conversation. With a few exceptions (like Cantonese), foreign languages generally sound like pleasant background music to me. (And bonus points if the speaker is French, Japanese, or female.)
There use to be these two Asian kids in my computer lab that would discuss their projects with each other using really broken english in tense, panicy voices. They were rather loud. That always shot my concentration to complete hell.
Slashdot Mirror
Mod parent up.
Even if you disagree with these protestors, be glad that they are protesting. If there's no political pressure to "get it right", then somebody somewhere will get it wrong, with far-ranging consequences...
Good call... I had never quite fingered what was disappointing about the first MI movie, but I think that's it: the entire team is gruesomely killed off in the first couple of minutes.
And oh yeah, the product placement of powerbooks was annoying...
I never went to see the second movie.
because many potential customers (like me) use their free web based account wherever they happen to be on what ever computer is nearby.
When you use your Yahoo web-based account, the email still originates on Yahoo's mail servers, even if you are sitting in a Starbucks. DomainKeys and SPF work on the level of the mail protocal itself and would not affect you.
This would actually take risk out of your business... a shoulder-surfing hacker next to you in Starbucks would not be able to forge emails from nflnk@yahoo.com without compromising something in Yahoo's system [at additional time/cost/risk to the hacker].
If you're just re-branding an open source app, that's somewhat lame, even if you are respecting the license agreement. Legal dues aside, the are still some social dues to be paid.
True, but a legitimate question is "should corporations be allowed to hold monopolies on culture?". I don't think copyright should be abolished but this "world leader in the production of creative works" BS ignores the fact that mass-produced culture is not very authentic culture. Weaker copyrights might make a stronger society... it's an option worth exploring anyways.
That day 0 copy of the next StarWars can and should land you in jail plain and simple.
Go to jail for copying a song? Sending somebody to jail is not cheap. Not only do you have to pay the (relatively trivial) cost of incarceration, you also lose that individual's productive input to society. Also, the more people you put in jail, the more construction activity must go into jails (instead of say, libraries or offices). The more people in jail, the more guards/wardens/police/judges/lawyers you need. Human potiential that could be put towards space, towards knowledge, towards medicine...
As a general rule, when "everybody" is breaking the law, it's time to revise the law, not throw more people into jail.
The Bible certaintly conveys that impression. Consider Luke 11:9-10: "...ask and it will be given to you; seek and you will find; knock and the door will be opened to you. For everyone who asks receives; he who seeks finds; and to him who knocks, the door will be opened.".
My original comment was facetious... I was pointing out that the distinct lack of evidence contradicts what we would expect the Christian God to do if he existed.
I'm not saying that this *disproves* the existence of the Christian God (or any other gods), but it is consistent with what we would expect from a purely social institution.
Yeah... I thought it was cool when I was a Christian too. Seriously! A kind of curse-of-the-ark thing. It poses a difficult question for theologians though: God loves you enough to torture his own son to death, but he's pretty meticulous about sealing off any avenue of empirical verification.
Let's face it, life is about pleasure. If some individual believes in "doing no evil", it's simply because, in the opinion of the individual's intellect, "doing no evil" is a good way to make pleasure. On the long term, pissing everybody off is probably not a viable pleasure strategy.
Seems like a good summary to me. Humans can do evil too. Why do many slashdotters seem to almost excuse corporations for being ruthless in the pursuit of their goals? Why it is presumed that the same social criticism we apply to individuals cannot be levied at corporations to pressure them to conform to acceptable business practices?
Perhaps we need some sort of "afterlife for corporations" concept to keep them on the straight and narrow... :-)
Brainstorm a list of all the things you would like to put into a robots.txt (or robots.xml, if we have to go that route).
Find like-minded people who can help you refine and prioritize your list of "wants".
Next, get a developer to help you figure out what an effective specification would look like. The spec should provide a clean syntax that will be easy for non-programming web developers to use, and it should incorporate most or all of your "wants".
If at all possible, the programmer will get involvement from people at Google and other search engines to see how this approach will be perceived by them. Since adoption is voluntary, it is critical that this new standard realistically interoperate with their business... if everybody is going to say "only check my site on Wednesdays", then search engines that adopt the new conventions will have unused bandwidth on other days of the week; this would mean that they would have to purchase a lot more bandwidth to achieve the same level of service.
Desiging a good spec (and communicating the design features that make it a good spec) will be challenging. Once you've done all of this, submit it to W3C if possible and let them take the ball and run with it. If the W3C won't take it, try the IETF. If they don't take it, try a few other standards organization, or whip up your own pretentious one.
While the spec is being "blessed" by the standards organization, have your programmer create a reference implementation and open-source it under something super-friendly to rapid adoption (not the GPL in this case, but something like the clarified BSD license that is compatible with the GPL). At the same time, prepare materials (tutorials, examples, etc.) to help promote your new standard.
IIRC, the XML SAX API is a pretty good example of how a group of people created a standard w/o going through a standards body. That approach might be effective too.
Actually, the BSA has about the same reputation as the MPAA and RIAA here on SlashDot. A big portion of the open source movement itself is about addressing the problems that occur when you start locking information down to create an artifical scarity of knowledge.
I have to admit that I'm much more respectful of software copyrights than I am of movie/music copyrights. I'm willing to concede hypocrisy in this regard, but I do wonder if it is immoral to so restrict the retransmission of "cultural" artifacts.
Personally, I hope to contribute--over the entire span of my life--ideas back to the community in the form of software and art, both in my formal work and as a hobby. In all situations, it would be gratifying to see other people take these ideas and develop them further.
One thought experiment: what if we completely rescinded copyright on future works of software and art? (Yes, yes... this would hurt the GPL too, blah, blah. I don't really think we should rescind copyright... we should just ask "what if"?) As an economic premise, I would say that people would still have needs for entertainment. That's still going to happen somehow, even if the money is not going to movies. And software would still be built based on business need. Hmmm... following the implications of all that might be an extensive exercise.
Humans are kinda in a rush to assign blame, or rather... attribute intention for events with socially relevant consequences. If it impacts us socially, our brains insist that somebody must have intended it, even if we also have a readily available mechanical explanation.
Or at least, that's the premise of Religion Explained. It's a fascinating read.
Umm... the cruise control on my 2002 Honda Accord is pretty lame, actually. It generally sticks within +/- 3 MPH, but I can do much better if I am controlling it myself.
The part that sucks is that sometimes the cruise control decides to gun the engine when going up a hill. It's unacceptable for a human driver to spike the tachometer to 4500 RPM near the top of the hill just because he's 5 MPH under his desired speed.
Of course, as you point out, the human has more information to work with than the car. Still, I think the grand-parent post has a point: Real Life is notorious for finding unanticipated circumstances to throw at "autonomous" devices.
One day machines will exceed human performance, but it's going to be a long road (pardon the pun). It's not just a matter of having faster reflexes... it's a matter of having superior judgement and reasoning. If a situation starts to unfold on the interstate, how much time should be spent looking for escape paths? If you try breaking hard, will the driver behind you have time to react? Are you fscked already? If so, is there something you can do to minimize damage (e.g,. hit a car instead of a tractor trailer?).
What's going to be impossible is to have BOTH human drivers and machine drivers on the road at the same time. This is a lot more difficult than just having machine drivers, because it requires that machine drivers be able to interpret social cues (and perhaps even fake them). Can the machine analyze another driver's face and tell that they are distracted/tired/busy? How will right-of-way psychology unfold when one of the participants is non-digital? If you're waiting for the traffic signal at a shady intersection and you start to get a Bad Feeling about the thugs approaching the car, will it be smart enough to do a risk evaluation and run the red light?
There's a lot of stuff to think about and be addressed before this is viable...
NIH syndrome exist everywhere, not just in OSS. I think both get darwinistically filtered by the marketplace to some degree.
One advantage of OSS is that the software does not have to be easy to learn. (!) Programming is hard, even before you get to UI. There are many different design tradeoffs that must be made. Proprietary vendors are FORCED to make sacrifices for usability up front, early in the product's evolution. Effort towards stability/security/conceptual integrity, etc. might be funnelled into UI because otherwise the product will fail in the market.
By contrast, OSS projects generally start out to satisfy an immediate need or curiosity. They don't have to "conquer the marketplace". Developers start by addressing the core problem. The result is a tool that does the job well. Then, as the tool becomes popular, other people focus on "making it usable". Some great examples include: cdrecord/cdbakeoven, fetchmail/fetchmailconf, mysql/tora, cvs/cervisia. Notice how these are all split at the CLI/GUI border. This split gives people the best of both the CLI and GUI worlds.
This phenomena seems to apply best when "the tool" is not a GUI application where domain logic and presentation are tightly integrated. For instance, sound editors (like Ardour), drawing tools (xfig, gimp), and 3D modelers (blender) can be very difficult to learn if usability is not given a front seat in development. On the other hand, operating systems, system software, databases, conversion utilities, etc. seem to work better if "easy to learn" is not initially a must-have.
Sun says "make this easy to use," and it gets done.
LMAO... how many years did sun shovel CDE down people's throats? A single driving vision can be a great thing, agreed, but not if people have the wrong vision.
Hmmm... I dont' buy it. Are their SAM systems actively protecting every nuclear site? Are they operated by staff with proper training? Do they get tested regularly? Can they fire w/o offsite authorization? Are there any relevant NRC regs you can cite?
I am told that the containment buildings for the nuclear plants I support can withstand a direct hint from a volkswagen going 100 MPH. Containment would probably provide some protection against larger objects too, but Boeings are much larger and faster than Volkswagens. :-)
the worst the criminal could do is shut the plant down and perhaps try and disperse the fuel with explosives
I'll one-up you: the worst the criminal could do is hijack a Boeing and crash it into the holding pool, which is not protected by containment like the reactor is. Release of radiation? Yes, probably. Heck... a contractor with access to the right materials might be able to drop an explosive device into the pool and cause some major damage to the plant w/perhaps some radioactive release. And definite public panic.
Armed assault will be met with armed resistance. The minute the attack starts, someone presses the panic button and the cavalry arrives.
Agreed, but the calvary is not going to be all that large. A modest paramilitary force would have a pretty good chance of accomplishing some serious evil. A large, trained paramilitary force would be unstoppable [if they could make it to the gate undetected... the logistics might give them a way weeks ahead of time].
Note that this post is just my speculation, gathered from the snippets I hear about plant security and happenings in the industry. Nuclear power is basically a good thing, but be wary of having too much trust in the system. (How many Pentagon workers killed 2001-09-11 would have bet that the military could have scrammed some jets fast enough to prevent an attack on the pentagon given an hour's notice?)
Wow... I know there's a "disposal mentality" in our society, but throwing away this planet once we make it to others is spectacularly careless. We're going to want this planet long after it's no longer necessary for the survival of the species.
I prefer to think of in terms of trying to curve-fit a set of points on the graph. Curve-fitting requires a sense of bias. For instance, suppose you have 30 (x,y) points from an experiment. If you want to interpolate/extrapolate other values for these points, you need to come up with a model (in the form of an equation). But what sort of model should you use? Something linear? Something parabolic?
It depends on what the graph looks like (and any "special" knowledge or expectations you may have about the situation, but that's a separate topic). For simple phenomena, a straight line will often work fine... just determine the magic numbers with a least-squares-fit.
Of course, those 30 points aren't going to fit exactly on the line. Some will be over, others under. Now suppose you have a colleague that points this discrepancy. He comes up with a new model (by solving for 30 simultaneous equations... go MatLab). The new model fits the data perfectly.
Which model is "correct"? Your 1-degree model that doesn't quite fit the data at hand (though you do have the excuse of measurement error...), or your colleague's 30-degree model, which is full of unexpected dips and surges b/t the collected points [because that's what tends to happen when you force-fit data in the manner described].
This is where Occam steps in... the 1-degree model is more reasonable. But don't rely on a witty quote. Take more measurements... the 1-degree model will continue to be slightly off, but the 30-degree model will be completely off.
Of course, there are phenomena that are better modeled with high-degree polynomials or non-polynomial equations. Choosing the right model is not only a question of science... it's a matter of design, even art. In many cases, the correct model will depend on the end result you are trying to achieve.
As a non-theist, I feel uncomfortable using Occam's Razor in debate. Witty sayings are cheap when it comes to slinging mud at each other's worldviews, and simplicity alone does not make for truth. I can sympathize with the spirit of Occam's statement though.... the idea of "God's Will" is very slippery: anything you observe can be attributed to God by fiat, even if his intentions do not make sense (because God's Will is unquestionable, unknowable). For me, believing in God was like rewriting the equation every time a new data point was found. [[That was my conlusion. Yours may be different... research for yourself if you haven't already.]]
Moving to more neutral ground, I like to apply Occam's Razor to technology too... when evaluating alternative approaches, I look at the simpler one and ask "can we get away with this, or do we *really* need the extra features of the more complex approach?" Having shot myself in the foot on a few occasions trying to make things too fancy, I'm a big fan of keeping it simple. But sometimes simple is wrong or inadequate. And sometimes we have misconceptions about what is actually simpler (for instance, both CLI and GUI can be considered "simpler" depending on the circumstances).
I was surprised to learn that a Windows port of Ethereal was packaged and deployed to our shared apps installation environment. Surprised because we're a pretty large and conservative company, and non-computer types are quick to find the potential downsides to a tool and categorize it as evil before the good side can be seen.
Anyways, Ethereal has been very helpful for exploring a variety of problems w/the different software we fool with. This gives us another way to "peer inside the box" of proprietary applications. For instance: Crystal Reports is a reporting package that has a "subreports" feature. A subreport can be an external file that Crystal pulls in every time you open the main report. The troublesome thing is that Crystal doesn't tell you where the external file is located. But w/Ethereal, we can see the files being pulled over the network. This was an invaluable shortcut to figuring out what the hell was going on.
I appear to be on a switched ethernet segment, so I'm not in danger of breaking anyone's privacy if I were to use the tool in promiscuous mode (like I sometimes did in college :-).
Two points:
If you prefer Adobe to Gimp, that's great too... buy it and use it. Judge the tradeoffs in cost and functionality for yourself and choose the best tool for you.
It is wrong. Godel said so. The problem is that your metaformula describes all truths, but there are more truths than there are possible descriptions.
Try reading Schismatrix or Holy Fire (or even Distraction) as if it was a biography. Instead of looking for plot, look at how technology transforms the protagonist and his society. Instead of looking for one amazing technological device that drives the story (an 'artifact-come-plot-device' or 'Stargate', if you will), look for a glut of radical, unbalancing technologies, each treated with unnerving casualness.
Some of Sterling's ideas are quite ambitious, almost terrifying, and utterly convincing. In Schismatrix, you can almost feel the collective human existence struggling to instantiate itself into something that's far above intelligence. Not all of his works are great, but Schismatrix and Holy Fire were triumphs; they rank at #1 and #3 on my list of all-time-best sci-fi.
Yah, if windows scripting is on, you'll need to block ".wsh", ".vbs", and ".js" too. Maybe others.
Open source software websites tend to be better about this than proprietary ones. They also (in the case of web apps) tend to have functional demos that you can test drive w/o the hassle of installing. It's very frustrating when any software maker (open source or proprietary) does not include screenshots, but at least on OSS sites, I can determine this quickly instead of getting lost in layer after layer of marketing crap. It's just another little joy of OSS.
Some people would make you accountable for your private thoughts and intentions. If they could, they would impose controls and restrictions over how you think. They would strip you of your free will.
History shows that if you create an accountability tool that is cleanly abstracted from policy, the tool will be abused sooner rather than later. The best way to prevent abuse is to encode a (rather liberal) policy into the tools and force the policymakers to deal with problems reactively.
Autonomy is a very good thing when properly allowed to flurish. The hard part is making sure that the autonomus units don't collapse in on themselves by taking advantage of each other.
IMO, it's the conversations in my own language that tend to disrupt my conversation. With a few exceptions (like Cantonese), foreign languages generally sound like pleasant background music to me. (And bonus points if the speaker is French, Japanese, or female.)
There use to be these two Asian kids in my computer lab that would discuss their projects with each other using really broken english in tense, panicy voices. They were rather loud. That always shot my concentration to complete hell.