And
in the event that all of this fails, we're going to need the much-maligned national missile defence folks. When you don't know in advance what's coming, you have to be able to
protect yourselves!
Umm, what's the likelihood that "what's coming" is a missile versus something like a truck or van loaded with explosives? Maybe we could look at some statistics from recent terrorist attacks. Let's see, how many involved long-range ballistic missiles?
I bought Space Empires IV (Malfador Machinations), and while some aspects of that game are appealing, the AI is just hopelessly inadequate compared to that in good ol' MOO2. I now can look forward (with fingers crossed) to the challenge of MOO2 in a whole new package.
One of the best parts of SE IV is its configurability. That'd be nice to see in MOO3.
Because hardware IA-32 emulation in the Itanium seems currently lame, and perhaps doomed to be irretrievably lame, Transmeta technology seems like it might be an alternative that would allow the ostensibly high-power native Itanium ISA do IA-32 emulation at reasonable performance levels. Thus for AMD to control Transmeta (and all their IP) would not only help AMD directly with their own product lines, but indirectly by closing down that route for Itanium.
There's a heck of a lot of x86 software floating around out there, so well-performing emulation (in my opinion) will be critical to acceptance by large businesses.
I'm sure it's just poetic license, but Apollo 11 landed on the moon in mid-summer. What were you doing in school?:-)
As the parent of an obsessed three-year-old who knows the names of more Apollo astronauts than virtually any adult, the sense of loss is brought home to me every time he says "Michael Collins" or "Jim Lovell" and the other adults in the room ask what the heck he's talking about.
The idea of gathering people up and keeping or selling them as slaves has been around for quite some time. It wasn't new when Europeans started exploiting the African slave machine in the sixteenth century, and it wasn't new when the African slave machine got rolling. The idea that when Our Side wins we get to keep individuals from Their Side and force them to do what we want is very, very old. It's just a short jump from that to slave-gathering as a business, which is also quite old.
Modern Western slavery as in the US seems so repulsive to us because it stands out so glaringly against progress of the Enlightenment and the avowed principles of the United States itself. However, the basic "lifestyle" of an American slave would have been readily understandable to a time-travelling Greek captured and sold by the Achaemenid Persions a millenium and a half earlier. The scale and Industrial Revolution efficiency of early 19th century slaving would have seemed terrifying (as it seems appalling to us), but that's an unremarkable consequence of carrying on the practice during the early Industrial Revolution.
Is the Itanium gcc comparable in optimization ability to Intel's compiler suite? Optimizing for a (relatively) exotic design like Itanium is not the same as optimizing for more traditional ISA like IA-32 or Alpha or Mips.
Benchmarking IA-32 code is very interesting, because it exposes Intel's confidence (and willingness to gamble) that the software world will adopt the new architecture aggressively. I don't think Microsoft has announced availability of Windows on Itanium yet; in any case I'd bet it'll be a while before they have the stability and performance that'd justify the server-class pricetag they're likely to ask for it. On top of that, there's the problem of selling IT organizations on something wholly new, complete with added complexity of installation management and versioning.
I wonder what's in store for Transmeta and their adaptive run-time technology? If the hardware emulation (which, to me, seems crazy) that Intel builds into the chips can't cut it, will Intel (or Microsoft?) be forced to deal with Transmeta? Is it conceivable that one or the other would consider owning Transmeta?
If you want a good example that everybody will immediately understand, go out and rent/buy/borrow a copy of Disney's The Aristocats. Tell your friend to bring their two-year-old kids over to watch the movie (which is pretty dumb, but really young kids like it) and start it up. But wait -- what's this?!? A trailer for Disney's Dinosaur, complete with (to a two-year-old) terrifying, snarling monsters and a loud realistic soundtrack? Quick! Grab the remote and skip to the DVD menu! Sorry, but THAT OPERATION NOT PERMITTED AT THIS TIME.
That's right, the sweet little marketing dwarves at Disney decided it'd be really cool to force customers to sit through 15 minutes of trailers at the front of the movie. How did they do that? CSS, of course!
I have yet to go through that scenario with a previously oblivious friend and not have one more fightin' mad anti-MPAA compadre in the end.
Should it be illegal to think about child molestation?
If I write a note wherein I describe my thoughts on the pleasures of child molestation, should I be arrested?
If I have in my posession a graphic, well-written description of child molestation that I find arousing, should I be arrested?
If I have in my posession a graphic, well-written description of child molestation that somebody else finds arousing, should I be arrested?
If I have in my posession a videotape of excerpts from the gymnastics competition at the 2000 Olympics, and a child molester finds the images of little girls appealing and arousing, should I be arrested? (Should women's gymnastics be banned from television?)
How exactly would you prove, or what proof would convince you, that images of children engaged in sex acts (photographs from life, VR images, pen and ink drawings) drive borderline sex offenders to commit crimes of child molestation that they otherwise wouldn't commit?
The laws about child porn discourage more vigourously, because people are much more disturbed by it.
I agree, but I don't think that's rational.
The main fear of child abuse/molestation, including child porn, is the psychological damage to the child that may not be reparable.
I'd say that the damage caused by murder is pretty irreparable. And of course not just to the murdered party.
Serial murderers certainly seem to repeatedly commit crimes without an outside party noticing. It's got one big advantage, crime-wise, over child abuse: the victim can't turn you in.
So if there's going to be a "special case", shouldn't we start with the worst crime we can think of? Is that really child molestation?
Frankly, I don't think legal means are the answer. I the only answer is to ostracize - or outright execute - people who perpetuate child porn.
Hmm... so you support the idea of such people being illegally executed? What other crimes are suited for such treatment? And let's keep the basic point clear: we're not necessarily just talking about people who create child pornography by abusing children (or who create virtual child pornography by wrangling bits; would that really call for summary execution?), we're talking about ownership of imagery. Should people who "perpetuate" child porn simply by trading images be shot too? How about people who trade pictures of murdered people? Can you give us a breakdown of those crimes that call for extra-legal execution, and the corresponding information-ownership crimes that call for like punishment on the grounds of "perpetuation"?
(How do you prove "perpetuation" in court, I wonder?)
Should they do this for all producers of all content that? What about when there are live-action VR shooter games where it looks exactly like real people are being blown apart by your plasma rifle? Should the producers give the same sort of evidence that they did not actually blow people up in order to make the game?
As a parent, I assure you that I consider child abuse to be a terrible crime, but I can't say that I think it's worse than murder. I find it odd that I can publish and collect as many images as I like depicting people being murdered. Some of those images are disturbing, but others are strangely fascinating. Hmm... maybe I'll invade Poland this afternoon. Boy, I'm really getting worked up; I've never felt so alive! If it hadn't been for all my murder books I'd still be a quiet recluse.
To my knowledge, the only crime of which it is illegal to possess photographs is the sexual abuse of a child. I don't know of any laws against possession of photographs of murder. Surely one or more episodes of Fox Most Heinous High Speed Chases or whatever would certainly have landed the producers in jail if so.
No, I don't want to automatically click through to see the latest in X10 video technology. What I want is a browser feature that - within web sites I choose - will automatically start background http GET transactions for every anchor tagged image on the page. It should dutifully throw away everything it gets back, of course, though maybe it would pay enough attention to load embedded images from the content. The idea would be to allow me to support web sites I like (for example, somethingawful.com) by boosting their click-through numbers, without me having to actually do any real click-throughs.
Of course, it would be only a small matter of time before the value of a click-through would be driven way down, since advertisers would realize that the feature was the online equivalent of an automatice TV commercial squelcher. But would the world be worse at that point?:-)
I guess what pisses me off the most are things like ASP pages that I can't view on netscape on my SGI.
If you can't view the page, it's not an ASP problem, it's just plain bad HTML. Again this has absolutely nothing to do with dynamic content. My dynamic content works fine on Netscape (and Opera) because I don't write code to generate bad HTML. Others are not very particular.
We've got dual-CPU PII machines (probably about
400-500 MHz) in a load-balanced farm, running
NT. (I'd love to run Linux or FreeBSD, but there are definite issues with the fast JVM from IBM on multi-processor machines; like, it doesn't work:-(
Oh, and 1GB of RAM apiece.
They do just fine in general. The slow pages are the ones that involve accessing big chunks of the database, though by being less stupid in our database implementation that's getting better.
Performance analysis (simple page timing logs) have shown that a considerable number of slow pages are the result of slow client links. That problem is exacerbated by the tendency of pages to be big (our pages anyway). That's one thing JSP sure makes easy!
Session migration over multiple load-balanced machines is touched upon lightly but no technical details were
offered. I would have liked to see a practical example on session persistence coupled with a use case of load
balancing a JSP website.
This seems like a pretty critical omission to me. Creating a fast website with lots of server-side computing with a stateless transaction model is tricky, and some description of how to introduce caching of any kind into a load-balanced world seems like the most important single contribution a book like this would make to my library.
But this is probably a moot
point since you gain i18n features directly from the Java language...
Yea, lots of managers think that too:-) Sorry, but internationalization cuts right down to the basics of your application design, unless you just want to create separate installations for each language.
So you don't like dynamic content because you don't like something that you know is different but seems superficially similar? I'm not sure I get it...
Most of the things that make pages slow are stupid counters and ad servers. A few judicious host file entries can really speed up your surfing experience.
I don't know how it would be implemented, but surely tough anti-spam laws and a regulatory body would be an attractive solution?
If that's actually a question, then no a regulatory body would not be an attractive solution. What the heck would they do, put meters on everybody's mail client?
Chances are that if you're down with non-free software on Linux, then you only use Linux because you have some irrational dislike of Bill Gates.
That's pathetic.
Entertaining and Slashdot-culturally-correct as it may be to spew righteous flameage at a strawman, that statement is just plain stupid. I'd say chances are most people who use Linux do so because they like it.
Idealistic endeavors like the Free Software Movement should be things that believers adopt for their own personal reasons. Browbeating other people for not sharing your ideals is just plain wrong-headed. If you believe in Free Software, great. Write some. Make it so compelling that it'll dominate its space and drive out non-free competition.
What we don't need is this stupid Free Software jihad mentality.
Parsing angle brackets is a non-issue, and does not make parsing hard. It's been a while since I played with it, but I'd imagine TeX as being quite a bit harder to parse than SGML/XML/HTML. XML, at least, is almost trivial.
There's no such thing as "HTML output"; the phrase does not even make sense. The statement is simply wrong anyway, since (for example) I could quite easily implement an HTML renderer that mapped HTML onto TeX statements and then invoked TeX for actual rendering. HTML is a markup language, and does not imply any specific implementation.
More important to me than user-visible calendar features would be a data architecture that allowed great flexibility in representing timed events, and a distributed implementation that allowed a wide variety of uses.
I've spent a considerable amount of time (under salary:-) thinking about how to represent timed events, and how to deal with the relationships between different schedules, and finally how to implement different kinds of services that work with that information. For whatever it's worth, what I decided was that just about everything can be reduced to the basic building block of an event cycle definition.
The ideas I had came from thinking about the ways audio synthesis works, or how similar systems like texture synthesis (KPT Bryce et al). If you think about it, basic event cycles (i.e., your bi-weekly staff meetings) are like waveforms. If you can describe your staff meeting as a waveform and capture that in a way that you can perform arithmetic between waveforms, then you can glue the building blocks together and capture a very rich variety of scheduling expressions. Using a consistent algebraic system makes it possible to implement all sorts of semantics (searching for coincident points on various different schedules, "masking" operations for holidays or other policies, and so on).
With an infrastructure like that, it'd be possible to put together a bunch of different kinds of clients that performed different kinds of scheduling. For example, there's no reason why the same system couldn't handle the tracking of scheduled IT operations as well as personal business schedules of employees.
The point is, it's not illegal to own photographs of somebody murdering a person. Murder is just as illegal as conducting a kiddie porn photo shoot, but photographs of murders are not illegal to own.
Why?
What the hell are you talking about? This topic has next-to-nothing to do with the Columbine incident. That was at most a catalytic factor triggering the outpouring of stories from kids who have to deal with a wide variety of stupid stuff. Take some time to read the testimonials about clueless parents and narrow-minded control freak school officials and forget about Klebold and Harris. They might as well be fictional characters.
That things were just as bad in the good ol' days is also irrelevent. Long-standing inequity is not justified simply on the basis of it being traditional.
Slashdot Mirror
Umm, what's the likelihood that "what's coming" is a missile versus something like a truck or van loaded with explosives? Maybe we could look at some statistics from recent terrorist attacks. Let's see, how many involved long-range ballistic missiles?
One of the best parts of SE IV is its configurability. That'd be nice to see in MOO3.
There's a heck of a lot of x86 software floating around out there, so well-performing emulation (in my opinion) will be critical to acceptance by large businesses.
As the parent of an obsessed three-year-old who knows the names of more Apollo astronauts than virtually any adult, the sense of loss is brought home to me every time he says "Michael Collins" or "Jim Lovell" and the other adults in the room ask what the heck he's talking about.
Modern Western slavery as in the US seems so repulsive to us because it stands out so glaringly against progress of the Enlightenment and the avowed principles of the United States itself. However, the basic "lifestyle" of an American slave would have been readily understandable to a time-travelling Greek captured and sold by the Achaemenid Persions a millenium and a half earlier. The scale and Industrial Revolution efficiency of early 19th century slaving would have seemed terrifying (as it seems appalling to us), but that's an unremarkable consequence of carrying on the practice during the early Industrial Revolution.
Benchmarking IA-32 code is very interesting, because it exposes Intel's confidence (and willingness to gamble) that the software world will adopt the new architecture aggressively. I don't think Microsoft has announced availability of Windows on Itanium yet; in any case I'd bet it'll be a while before they have the stability and performance that'd justify the server-class pricetag they're likely to ask for it. On top of that, there's the problem of selling IT organizations on something wholly new, complete with added complexity of installation management and versioning.
I wonder what's in store for Transmeta and their adaptive run-time technology? If the hardware emulation (which, to me, seems crazy) that Intel builds into the chips can't cut it, will Intel (or Microsoft?) be forced to deal with Transmeta? Is it conceivable that one or the other would consider owning Transmeta?
That's right, the sweet little marketing dwarves at Disney decided it'd be really cool to force customers to sit through 15 minutes of trailers at the front of the movie. How did they do that? CSS, of course!
I have yet to go through that scenario with a previously oblivious friend and not have one more fightin' mad anti-MPAA compadre in the end.
If I write a note wherein I describe my thoughts on the pleasures of child molestation, should I be arrested?
If I have in my posession a graphic, well-written description of child molestation that I find arousing, should I be arrested?
If I have in my posession a graphic, well-written description of child molestation that somebody else finds arousing, should I be arrested?
If I have in my posession a videotape of excerpts from the gymnastics competition at the 2000 Olympics, and a child molester finds the images of little girls appealing and arousing, should I be arrested? (Should women's gymnastics be banned from television?)
How exactly would you prove, or what proof would convince you, that images of children engaged in sex acts (photographs from life, VR images, pen and ink drawings) drive borderline sex offenders to commit crimes of child molestation that they otherwise wouldn't commit?
I agree, but I don't think that's rational.
The main fear of child abuse/molestation, including child porn, is the psychological damage to the child that may not be reparable.
I'd say that the damage caused by murder is pretty irreparable. And of course not just to the murdered party.
Serial murderers certainly seem to repeatedly commit crimes without an outside party noticing. It's got one big advantage, crime-wise, over child abuse: the victim can't turn you in.
So if there's going to be a "special case", shouldn't we start with the worst crime we can think of? Is that really child molestation?
Hmm... so you support the idea of such people being illegally executed? What other crimes are suited for such treatment? And let's keep the basic point clear: we're not necessarily just talking about people who create child pornography by abusing children (or who create virtual child pornography by wrangling bits; would that really call for summary execution?), we're talking about ownership of imagery. Should people who "perpetuate" child porn simply by trading images be shot too? How about people who trade pictures of murdered people? Can you give us a breakdown of those crimes that call for extra-legal execution, and the corresponding information-ownership crimes that call for like punishment on the grounds of "perpetuation"? (How do you prove "perpetuation" in court, I wonder?)
As a parent, I assure you that I consider child abuse to be a terrible crime, but I can't say that I think it's worse than murder. I find it odd that I can publish and collect as many images as I like depicting people being murdered. Some of those images are disturbing, but others are strangely fascinating. Hmm ... maybe I'll invade Poland this afternoon. Boy, I'm really getting worked up; I've never felt so alive! If it hadn't been for all my murder books I'd still be a quiet recluse.
To my knowledge, the only crime of which it is illegal to possess photographs is the sexual abuse of a child. I don't know of any laws against possession of photographs of murder. Surely one or more episodes of Fox Most Heinous High Speed Chases or whatever would certainly have landed the producers in jail if so.
Of course, it would be only a small matter of time before the value of a click-through would be driven way down, since advertisers would realize that the feature was the online equivalent of an automatice TV commercial squelcher. But would the world be worse at that point? :-)
If you can't view the page, it's not an ASP problem, it's just plain bad HTML. Again this has absolutely nothing to do with dynamic content. My dynamic content works fine on Netscape (and Opera) because I don't write code to generate bad HTML. Others are not very particular.
Oh, and 1GB of RAM apiece.
They do just fine in general. The slow pages are the ones that involve accessing big chunks of the database, though by being less stupid in our database implementation that's getting better.
Performance analysis (simple page timing logs) have shown that a considerable number of slow pages are the result of slow client links. That problem is exacerbated by the tendency of pages to be big (our pages anyway). That's one thing JSP sure makes easy!
This seems like a pretty critical omission to me. Creating a fast website with lots of server-side computing with a stateless transaction model is tricky, and some description of how to introduce caching of any kind into a load-balanced world seems like the most important single contribution a book like this would make to my library.
Yea, lots of managers think that too :-) Sorry, but internationalization cuts right down to the basics of your application design, unless you just want to create separate installations for each language.
Most of the things that make pages slow are stupid counters and ad servers. A few judicious host file entries can really speed up your surfing experience.
If that's actually a question, then no a regulatory body would not be an attractive solution. What the heck would they do, put meters on everybody's mail client?
Entertaining and Slashdot-culturally-correct as it may be to spew righteous flameage at a strawman, that statement is just plain stupid. I'd say chances are most people who use Linux do so because they like it.
Idealistic endeavors like the Free Software Movement should be things that believers adopt for their own personal reasons. Browbeating other people for not sharing your ideals is just plain wrong-headed. If you believe in Free Software, great. Write some. Make it so compelling that it'll dominate its space and drive out non-free competition.
What we don't need is this stupid Free Software jihad mentality.
I've spent a considerable amount of time (under salary :-) thinking about how to represent timed events, and how to deal with the relationships between different schedules, and finally how to implement different kinds of services that work with that information. For whatever it's worth, what I decided was that just about everything can be reduced to the basic building block of an event cycle definition.
The ideas I had came from thinking about the ways audio synthesis works, or how similar systems like texture synthesis (KPT Bryce et al). If you think about it, basic event cycles (i.e., your bi-weekly staff meetings) are like waveforms. If you can describe your staff meeting as a waveform and capture that in a way that you can perform arithmetic between waveforms, then you can glue the building blocks together and capture a very rich variety of scheduling expressions. Using a consistent algebraic system makes it possible to implement all sorts of semantics (searching for coincident points on various different schedules, "masking" operations for holidays or other policies, and so on).
With an infrastructure like that, it'd be possible to put together a bunch of different kinds of clients that performed different kinds of scheduling. For example, there's no reason why the same system couldn't handle the tracking of scheduled IT operations as well as personal business schedules of employees.
Just a thought.
The point is, it's not illegal to own photographs of somebody murdering a person. Murder is just as illegal as conducting a kiddie porn photo shoot, but photographs of murders are not illegal to own. Why?
Hmm ... do I get to pick the nurse?
That things were just as bad in the good ol' days is also irrelevent. Long-standing inequity is not justified simply on the basis of it being traditional.