I'm certainly not an expert on the anthropic principle, but let me see if I can at least make an analogy here.
Suppose you have four bags. One bag contains ten pennies, one a hundred pennies, one a thousand pennies, and one has ten nickles. (If you want the sensible version of this analogy, imagine that each bag has some number of pennies and some number of nickles, in interesting variations. I'm making it really simple.)
Suppose I pull a penny out of a bag, but you don't see which, and I say, which bag did I pull the penny out of, the ten-cent bag, the fifty-cent bag, the dollar bag, or the ten-dollar bag?
The only thing probability tells us conclusively is that I didn't get it out of the fifty-cent bag, which has only nickles.
Of the remaining bags, just because one of them has more pennies, that doesn't mean it's more likely I pulled a penny from that bag.
Now, you might think, well, but if instead of me picking a penny from a bag, that's a bad analogy, the better analogy to the anthropic principle is to talk about the point of view of the pennies. If we average across all pennies, we're more likely to see the point of view of a penny in the ten-dollar bag.
The real problem is this: only one of the four bag exists, and we have no way of knowing which one exists, and no basis for assigning any probabilities to them. Probabilities are best understood as the statistical properties of potential things, e.g. if the probability of a die rolling a 2 is 1/6, I'm saying, if I rolled the die a lot of times, the number of times would come up 2 would approach 1/6 of the attempts.
But we don't get to choose from multiple possible universes to accumulate statistics from them. That's meaningless. So you can't meaningfully assign probabilities to various alternate possible universes.
Let me make that explicit. I show you the penny, and I ask you which bag I took it out of. And you say, 1/3 probability the ten-cent bag, 1/3 probability the dollar bag, and 1/3 probability the ten-dollar bag, and 0 probability the fifty-cent bag.
Now, I reveal, that I made a conscious choice not to draw from the ten-cent bag, and I flipped a coin to decide between the other two. Does that make the probability 1/2 for the dollar and ten-dollar bags? Or is the truth just that I pulled it from the dollar back with probability 1, and from the others with probability 0, and there's really nothing more to be said?
Here's a similar thing to tangle with: suppose on Monday it rains, and on Tuesday it doesn't rain. The weatherman predicts a 70% chance of rain on Monday, and a 30% chance of rain on Tuesday. Was he right? What does that even mean? How would you go about measuring the accuracy of weather predictions? You could keep count, and find out for each percentage whether they actual number of times it rained matched that percentage. But, say over the course of the year it rains on average 1/5 of the time. A weatherman who predicts a 20% chance of rain every day will come out perfect, but be the most useless prediction.
Personally, I think the "matrix" agument has other more important flaws since you can pick a single fixed universe to make it work in. The substrate-independence argument seems a little weird when applied to a single machine computing 10M people's brains--it's not clear to me that anything approaching 10M independent experienceable consciousnesses would occur. But most importantly, I think outcome (2) is the most likely; people wouldn't be that interested in running such detailed simulations (of human brains sufficient to generate substrate-independent consciousness) for long periods of time; what's the point in simulating results you're never going to see?
In college I completed the requirements for an undergraduate philosophy degree, but this kind of angels-on-the-head-of-a-pin business is exactly why I decided to stop spending any time thinking about non-concrete crap.
It's hilarious how many times (like the original 'why ohh why') post people complain about "all the plot holes" in the original Matrix, and they all only ever name one plot hole, and it's the same one, the people-as-batteries thing. I'd think if it had lots of them, there'd be some variation in which one people complain about.
And then people like corporate mofo insist on over-intellectualizing the explanation. Just like parent:
But I think this discussion misses the real reason that they went with the power rational with The Matrix. I feel that they wanted to make a metaphorical statement about how people fuel "the Matrix" in reality.
The power rationale in The Matrix is just a strict conversion of the relationship between people and cows in real life into a relationship between machines and people. We're cows to the machines. That's it.
(The parent makes a basic mistake: the people don't power The Matrix. They power THE MACHINES. Powering the matrix is a small consequence of it, but not the primary goal. The purpose of the matrix is to distract the people from their role as cows. If their roles as cows were to power the matrix, the whole thing would be pointlessly redundant.)
Now, I admit, this cow analogy is a bit of a stretch, since there's nothing in the movie literally that represents it. Smith doesn't say humans are cows; he calls us a virus.
But there is a general trend in the movie of trying to humanize the villains; Smith wants the codes to Zion because he wants out of the matrix; he's just as trapped there as any of the people. Sure, he doesn't have a real life body to go back to (strictly talking about the first movie here), so it's not entirely the same, but it is voiced explicitly that he wants out.
And from there, I don't find it hard to view the humans-as-power-supply thing as something meant to make the machines more sympathetic because their treatment of humans isn't much different from humanities' treatment of animals.
<PretentiousBullshitParody>Or, more likely, given the Buddhist themes of the movie, and the fact that many Buddhists are vegetarians, the Wachowskis have designed The Matrix as a morality play to teach people that places like Cowschwitz are bad and that they should either be vegetarians or hunt their own food. Isn't it obvious?</PretentiousBullshitParody>
I haven't looked at the video to see the graphics glitch of which you speak, but I am a graphics guy, and the core technology used in Doom 3 for shadowing, called stenciled shadows, doesn't work with translucent/transparent surfaces.
Basically, the algorithim uses the Z values stored in the Z-buffer to determine whether each pixel is in shadow or not. (The Z-buffer is used for hidden surface elimination, and normally stores the single nearest opaque-surface.)
When you draw transparent surfaces, you end up with multiple surfaces visible at each pixel--the nearest opaque surface, and all closer transparent surfaces. But there's only a single value in the Z-buffer, so the checks to determine shadow determine whether that particular point (back-projecting that pixel to that depth) is in shadow.
So either transparent surfaces pick up the shadowing of the surfaces you see through them, or you turn off shadowing for transparent surfaces (and maybe do something else for them, like raycast one or more points on the surface to the light sources and use that info for shadowing the whole surface or each vertex).
The facts in your first two paragraphs are probably related.
If you read netflix' documentation, you'll find that if your top entry is in "wait", they don't send out a new disc immediately; they wait up to 12 hours for your top choice to come in. If it doesn't, then they send you your next one.
So if your top N entries (on an N-out account) are all in "wait" mode, it probably means you always add a 12-hour delay to every movie you return. If you only return one at a time, one is all it takes. 12 hours might be enough to add a day, due to quantization. (It was for me.)
So move your non-now entries out of the top of your queue periodically and see if it brings you back to the 2-day delivery.
What it means to "send it if it comes in during that 12 hour period" when you have availability "long wait" and someone else has "now", I don't know.
This is probably false; this case study cites a cost of $0.37 per DVD, each way, in 2002, so a total cost of $0.74 cents shipping/rental.
Thus, they wouldn't lose money on shipping until well over 20 DVDs per month.
All their other costs work out to an equal amount of overhead in that report, for a total cost of $1.50/DVD. That's still more than double "five per month" to break even.
I don't know what the sources are for that study, but obviously it makes some sense--it ought to cost Netflix less per rental than it costs a traditional brick-and-mortar store.
DRM is a tool that can and will be used to abrogate fair use (without messing with the legal right to fair use, i.e. doing an end-run around the law).
Ubiquitous deployment of DRM just does away with the opportunity of anyone to vote with their (virtual) wallet. It gets us in a position where we have no choice.
Ditto ubiquitous deployment of "trusted computing" in hardware. First get it in everybody's hands, then in a few years start using it, then in a few years start requiring it. See also 'copy-protected' CDs.
Doom 2 was a retail release, not shareware--and it sold something like 10x as many copies as there were registrations for Doom 1.
Quake was released as retail as well, although it was a little more complicated than that.
For recent examples of successful shareware games, I'd look more at things like Bejeweled and such from PopCap, or the Exile/Avernum games from Spiderweb Software.
Of course, there's still plenty of unsuccessful shareware these days; I've written some myself, but I'll spare you the link.
I can understand the thinking that went into the exception clause--they're allowed to include a patented, non-royalty-free technology in a standard if basically everybody (including the public?) agrees. The idea here is if something so crucial, so important comes along, well, we'd want it.
The problem is, announcing this up front could still encourage companies (who want to make money) to try patenting things and springing them further down the road (before it becomes an official standard but after everybody's used to it), hoping that they'll end up with something that everybody desperately wants and thus make royalties. (Of course companies involved in the standards-making process are supposed to disclose, but that can be messy, as we know, and I'm also thinking of companies not involved in the process itself.)
IMO it would be far better to say "no, we will never allow royalty-required standards"--discouraging corporations from thinking the previous paragraph describes a plausible strategy. If something becomes so important that obviously everyone would rather have it then not, then they can change the rules at a later date, but there's no need to announce that and encourage the "wrong" behavior.
"'Security through Obscurity' Considered Harmful" Considered Harmful
The basis of almost all security is the keeping of a secret--whether that be a login password or the passcode to access your private key for encryption. The whole premise of security is that users with this "secret knowledge" will keep it secret. You could even keep your private key in the clear on your hard drive, relying on the fact that nobody has physical access to that hard drive.
Keeping a secret is central. Keeping that secret secret. Obscuring that secret. Security through obscurity. So the slogan ("'security through obscurity' considered harmful") is just wrong.
Now I'm going to wander off-topic in a more straightforward scenario here, but what the slogan intends to address is that, if you, say, use a secret encryption algorithm to protect your data, you have to keep that encryption algorithm just as secret as you keep your secret key, if you expect that algorithm to buy you any extra protection over, say, triple DES. And that tends to be hard.
You have to describe and encode your algorithm somewhere, and keep that secret. You can measure the number of bits it takes you to describe and encode that algorithm, and add in the number of bits of your encryption key. That's the total number of bits you have to keep secret to protect yourself.
The real truth behind this sort of scenario of the anti-"security through obscurity" meme is that you get better security through putting the same number of bits into a secret encryption key for a well-tested public algorithm than you get for splitting those bits between a secret algorithm and a secret key, because even if your algorithm isn't inferior, it's still using fewer bits of key.
So, sure, security through obscuring your algorithm is bad, because there are better places to spend your secret bits. But "security through secrecy" is fundamental--and I think it's foolish to try to distinguish between "secrecy" and "obscurity".
Back when I was a user of The World, the wstd.* Usenet groups saw frequent complaints (my own among them) about false positives in World's spam blocking. How do you evaluate the tradeoff between blocking spam and accidentally blocking legitimate email?
I could imagine that dictionary-based spam "attacks" might incur a higher per-user cost if they're aimed at theworld.com than if they're aimed at aol.com--same absolute cost, but the difference in number of actual users changes the per-user cost. Is spam-fighting actually more expensive for smaller ISPs than large ones?
TCPA [...] can be turned off if you don't want it, but it can only be a Good Thing.
Do we have to go through the "guns don't kill people" argument here? Any defense you can offer of the hardware in TCPA and be offered for Palladium as well.
The reality is that getting ubiquitous hardware out there that can do TCPA will inevitably lead to superior DRM on that platform. Whether this will matter beyond DVDs (which already have broken DRM) or not, it's hard to say. But e.g. the music industry decides to finally get off their asses and sell music online, you can be sure it'll be DRM'd as best they can.
But since TCPA offers no real benefit--it's not going to stop Slammer, it's not going to allow you to tell the difference between an executable your friend sent you that you should run and an executable your friend sent you that you shouldn't run--it's hard to see any point to having it in the first place.
The previous ask slashdot post about this with the motherboard vendor (or was it chipset?) just amounted to "our customers want it". Why the hell do their customers (the people who are building machines) want it? They shouldn't give a damn unless maybe they're getting pressure from someone else who wants it, and their customers (you, me, and Joe Hexpack) aren't asking for it, so I don't know who that would be besides MS (who certainly have a history of playing hardball with PC vendors).
They also repeatedly point out that Watt and Boulton remained market leaders after the patent expired, observing that since competitors should have spent those 30 years reverse-engineering and catching up, then Watt and Boulton's continued success reflected some sort of economic advantage to them being first-to-invent entirely independent of their patent protection.
But I see little reason (from the story they provide) to conclude that this would have happened without patent protection. Perhaps it was market effects--Boulton and Watt were the big name you trusted since nobody else had been doing this for thirty years. Or maybe, since nobody else could make engines like this and sell them, the potential competitors didn't do that much reverse-engineering or build engines they couldn't sell, just for the experience; perhaps they focused on activities which could actually give them a profit.
The authors use this example twice in the first five pages and it really jumped out at me--and I'm totally anti-patent.
It's not clear to me yet that they "picked" Opera as a "target".
The opera page points out that maybe the Opera page was custom made for Opera 6, and then they run the IE page through Opera 6 and show that it renders fine. Therefore, they say, there's no need for a custom Opera 6 page, therefore it couldn't have been a custom Opera 6 page.
But the right thing to test would be whether Opera 6 renders the "bad pages" acceptably or is similarly broken. I could easily picture a developer trying to customize things for Opera 6 (for some reason or another) and testing something and Opera6 handling it fine and it accidentally getting left in.
I mean, maybe this really totally is MSN screwing with Opera, but the omission of the above data point, and the way they say
there is no
technical reason for MSN's behavior
when testing the Opera 6 case makes me a tad suspicious.
Duff's Device is pessimal in the face of an optimizing compiler. By jumping into the unrolled loop, you prevent the compiler from moving code across the branch entry points. Duff's Device turns the unrolled code from a single long basic block to a sequence of basic blocks identical to the original non-unrolled code, which will optimize identically to the original non-unrolled code.
A lot of the value of unrolling comes not merely from reducing the loop overhead (which Duff's Device still does) but from the ability to rearrange the code within the loop, such as moving loads forward and stores later. On a modern processor, out-of-order execution reduces the importance of such simple rearrangements, but it's still the case that it prevents any significant optimizations across the basic blocks.
Phoenix (I haven't tried Mozilla) does something sort of like this for it's "Go" history. I didn't work it out exactly--either it's remembering all pages, or it's just interleaving all the "stacked" pages from all tabs, but either way it totally screws up my normal browsing process, since I tend to leave a "main page" where I can get back to it with two or three "back"s, and use Go to shortcut that. Under Phoenix, the Go list ends up filling with misc junk and my "main page" isn't visible anymore.
So instead I'm still using Netscape 4.
false positives acceptable?
on
ISP Chief on Spam
·
· Score: 2, Interesting
I used to have an account on Shein's ISP--I'm sure there are still pointers to buzzard@world.std.com on my own pages--but their attitude towards false positives was simply unacceptable to me and a lot of other people I heard who left "The World". Erring on the side of getting more spam and no false positives was clearly preferred by me and by other vocal customers, especially those who ran businesses from their World accounts.
But Barry's stance was that since the vast majority of cusomters just wanted all the spam gone, the right thing to do was to accept a certain level of false positives. Unannounced--no warning that you would have legitimate mail returned to your friends with the unhelpful '200 UCE not accepted' or even '200 No thank you' replies (I don't remember the actual number, sorry)--with no "opt-out-of-the-spam-blocking" option for other customers.
One theory I have for The World's problms is that spam-blocking doesn't scale with customers, so The World is hit by it worse than larger ISPs. It seems like the support costs of dealing with customer complaints would scale with customers, though. But, for example, there apparently is (was) a pattern of spammers taking a list of plausible user names and emailing every name on the list @ the target host. Since that list of names is the same length whether it's theworld.com or aol.com, but the number of customers is different, the cost-per-customer for dealing with that (bandwidth / etc.) is higher for the smaller ISP. But nobody at The World was willing to comment on this sort of customer scalability issue (although they mentioned that particular spam scenario because they had a fairly aggressive response to it to avoid bandwidth--they stopped accepting connections from that IP for an hour or two if it was detected, which meant legit mail from that IP was often delayed and sometimes bounced if it kept getting reblocked).
Anyway, the upshot is, I have very little sympathy for somebody who thinks it's a good idea to let legitimate email get blocked as spam because it reduces customer support costs. It's just moving the problems somewhere else where the customers don't know about them.
Re:Transparency isn't really new in quake.
on
Tenebrae Quake
·
· Score: 1
The original quake lacked transparent water. It was hacked in for glquake, but you had to revis.
Quake 2 had support transparent water even with software rendering.
What, bugs that need fixing? It doesn't run on your platform? Isn't that the whole point of Open Source?
Anyway, the games were developed on P4s under Win2K, running on GeForce4s. I did a lot of the development on an Athlon under Win98, with a Radeon. I wouldn't expect significant Win32 portability issues. (Performance will suck without a top-end machine, though.)
It's possible the missing sprites is the problem--the only one I know Chris Hecker tested running without the sprites was RoboDOOM. I can't suggest any other fixes without a real bug report (what system are you trying it on, etc. etc.).
This is certainly an issue. Some of the games were not that thrilling from a game design innovation standpoint. But you've named the three worst cases; if you made the full list, they wouldn't all be like that, so you're making it look worse than it really is. Plenty were quite innovative: the firefighter game, Wrath, Flow, Dueling Machine.
And even for those three, it doesn't quite tell the whole story. I wrote the least significantly innovative one from a game design standpoint: Very Serious RoboDOOM. This was because I was one of the main engine authors, and I had to spend a lot of my time answering questions, or simply eavesdropping on what people were discussing with each other in case they were having trouble with something I knew how to fix.
We had a big list of 20-some game ideas that we had come up with in the early stages of the engine development--but all the participants were coming up with those ideas so none of them were getting written; I decided to try to write something off that list. Robotron was simple, and hopefully wouldn't be too hard to task switch on, and in fact, it was fairly easy to implement. I didn't end up innovating the game design, but I was able to use the game to make (I think) a comment about the the game-design of shooters and implicitly of the industry--which I think is a fairly good accomplishment for one day's work where I spent most of the time helping other people.
Unfortunately, I made the "used the Doom 2 sprites" mistake, so it's an effort to see the result; as organizers of the event, our focus was so much on making the experience for the participants during the event optimal, we didn't consider the redistribution issue particularly well--it just wasn't our priority.
Slashdot Mirror
Suppose you have four bags. One bag contains ten pennies, one a hundred pennies, one a thousand pennies, and one has ten nickles. (If you want the sensible version of this analogy, imagine that each bag has some number of pennies and some number of nickles, in interesting variations. I'm making it really simple.)
Suppose I pull a penny out of a bag, but you don't see which, and I say, which bag did I pull the penny out of, the ten-cent bag, the fifty-cent bag, the dollar bag, or the ten-dollar bag?
The only thing probability tells us conclusively is that I didn't get it out of the fifty-cent bag, which has only nickles.
Of the remaining bags, just because one of them has more pennies, that doesn't mean it's more likely I pulled a penny from that bag.
Now, you might think, well, but if instead of me picking a penny from a bag, that's a bad analogy, the better analogy to the anthropic principle is to talk about the point of view of the pennies. If we average across all pennies, we're more likely to see the point of view of a penny in the ten-dollar bag.
The real problem is this: only one of the four bag exists, and we have no way of knowing which one exists, and no basis for assigning any probabilities to them. Probabilities are best understood as the statistical properties of potential things, e.g. if the probability of a die rolling a 2 is 1/6, I'm saying, if I rolled the die a lot of times, the number of times would come up 2 would approach 1/6 of the attempts.
But we don't get to choose from multiple possible universes to accumulate statistics from them. That's meaningless. So you can't meaningfully assign probabilities to various alternate possible universes.
Let me make that explicit. I show you the penny, and I ask you which bag I took it out of. And you say, 1/3 probability the ten-cent bag, 1/3 probability the dollar bag, and 1/3 probability the ten-dollar bag, and 0 probability the fifty-cent bag.
Now, I reveal, that I made a conscious choice not to draw from the ten-cent bag, and I flipped a coin to decide between the other two. Does that make the probability 1/2 for the dollar and ten-dollar bags? Or is the truth just that I pulled it from the dollar back with probability 1, and from the others with probability 0, and there's really nothing more to be said?
Here's a similar thing to tangle with: suppose on Monday it rains, and on Tuesday it doesn't rain. The weatherman predicts a 70% chance of rain on Monday, and a 30% chance of rain on Tuesday. Was he right? What does that even mean? How would you go about measuring the accuracy of weather predictions? You could keep count, and find out for each percentage whether they actual number of times it rained matched that percentage. But, say over the course of the year it rains on average 1/5 of the time. A weatherman who predicts a 20% chance of rain every day will come out perfect, but be the most useless prediction.
Personally, I think the "matrix" agument has other more important flaws since you can pick a single fixed universe to make it work in. The substrate-independence argument seems a little weird when applied to a single machine computing 10M people's brains--it's not clear to me that anything approaching 10M independent experienceable consciousnesses would occur. But most importantly, I think outcome (2) is the most likely; people wouldn't be that interested in running such detailed simulations (of human brains sufficient to generate substrate-independent consciousness) for long periods of time; what's the point in simulating results you're never going to see?
In college I completed the requirements for an undergraduate philosophy degree, but this kind of angels-on-the-head-of-a-pin business is exactly why I decided to stop spending any time thinking about non-concrete crap.
And then people like corporate mofo insist on over-intellectualizing the explanation. Just like parent:
The power rationale in The Matrix is just a strict conversion of the relationship between people and cows in real life into a relationship between machines and people. We're cows to the machines. That's it.
(The parent makes a basic mistake: the people don't power The Matrix. They power THE MACHINES. Powering the matrix is a small consequence of it, but not the primary goal. The purpose of the matrix is to distract the people from their role as cows. If their roles as cows were to power the matrix, the whole thing would be pointlessly redundant.)
Now, I admit, this cow analogy is a bit of a stretch, since there's nothing in the movie literally that represents it. Smith doesn't say humans are cows; he calls us a virus.
But there is a general trend in the movie of trying to humanize the villains; Smith wants the codes to Zion because he wants out of the matrix; he's just as trapped there as any of the people. Sure, he doesn't have a real life body to go back to (strictly talking about the first movie here), so it's not entirely the same, but it is voiced explicitly that he wants out.
And from there, I don't find it hard to view the humans-as-power-supply thing as something meant to make the machines more sympathetic because their treatment of humans isn't much different from humanities' treatment of animals.
<PretentiousBullshitParody>Or, more likely, given the Buddhist themes of the movie, and the fact that many Buddhists are vegetarians, the Wachowskis have designed The Matrix as a morality play to teach people that places like Cowschwitz are bad and that they should either be vegetarians or hunt their own food. Isn't it obvious?</PretentiousBullshitParody>
Basically, the algorithim uses the Z values stored in the Z-buffer to determine whether each pixel is in shadow or not. (The Z-buffer is used for hidden surface elimination, and normally stores the single nearest opaque-surface.)
When you draw transparent surfaces, you end up with multiple surfaces visible at each pixel--the nearest opaque surface, and all closer transparent surfaces. But there's only a single value in the Z-buffer, so the checks to determine shadow determine whether that particular point (back-projecting that pixel to that depth) is in shadow.
So either transparent surfaces pick up the shadowing of the surfaces you see through them, or you turn off shadowing for transparent surfaces (and maybe do something else for them, like raycast one or more points on the surface to the light sources and use that info for shadowing the whole surface or each vertex).
If you read netflix' documentation, you'll find that if your top entry is in "wait", they don't send out a new disc immediately; they wait up to 12 hours for your top choice to come in. If it doesn't, then they send you your next one.
So if your top N entries (on an N-out account) are all in "wait" mode, it probably means you always add a 12-hour delay to every movie you return. If you only return one at a time, one is all it takes. 12 hours might be enough to add a day, due to quantization. (It was for me.)
So move your non-now entries out of the top of your queue periodically and see if it brings you back to the 2-day delivery.
What it means to "send it if it comes in during that 12 hour period" when you have availability "long wait" and someone else has "now", I don't know.
Thus, they wouldn't lose money on shipping until well over 20 DVDs per month.
All their other costs work out to an equal amount of overhead in that report, for a total cost of $1.50/DVD. That's still more than double "five per month" to break even.
I don't know what the sources are for that study, but obviously it makes some sense--it ought to cost Netflix less per rental than it costs a traditional brick-and-mortar store.
Netflix does share 40% of revenues (table 1, page 7) with the movie studios.
You do realize that if he'd stopped playing the game, the vibrator would have stopped.
He didn't stop playing.
Now there's a needs-attentive boyfriend.
Ubiquitous deployment of DRM just does away with the opportunity of anyone to vote with their (virtual) wallet. It gets us in a position where we have no choice.
Ditto ubiquitous deployment of "trusted computing" in hardware. First get it in everybody's hands, then in a few years start using it, then in a few years start requiring it. See also 'copy-protected' CDs.
"Crippleware" works, "traditional" shareware doesn't. One shareware author documented a 5-to-1 difference in the matter.
You can read more about Ambrosia's take on the matter here.
Doom 2 was a retail release, not shareware--and it sold something like 10x as many copies as there were registrations for Doom 1.
Quake was released as retail as well, although it was a little more complicated than that.
For recent examples of successful shareware games, I'd look more at things like Bejeweled and such from PopCap, or the Exile/Avernum games from Spiderweb Software.
Of course, there's still plenty of unsuccessful shareware these days; I've written some myself, but I'll spare you the link.
The problem is, announcing this up front could still encourage companies (who want to make money) to try patenting things and springing them further down the road (before it becomes an official standard but after everybody's used to it), hoping that they'll end up with something that everybody desperately wants and thus make royalties. (Of course companies involved in the standards-making process are supposed to disclose, but that can be messy, as we know, and I'm also thinking of companies not involved in the process itself.)
IMO it would be far better to say "no, we will never allow royalty-required standards"--discouraging corporations from thinking the previous paragraph describes a plausible strategy. If something becomes so important that obviously everyone would rather have it then not, then they can change the rules at a later date, but there's no need to announce that and encourage the "wrong" behavior.
The basis of almost all security is the keeping of a secret--whether that be a login password or the passcode to access your private key for encryption. The whole premise of security is that users with this "secret knowledge" will keep it secret. You could even keep your private key in the clear on your hard drive, relying on the fact that nobody has physical access to that hard drive.
Keeping a secret is central. Keeping that secret secret. Obscuring that secret. Security through obscurity. So the slogan ("'security through obscurity' considered harmful") is just wrong.
Now I'm going to wander off-topic in a more straightforward scenario here, but what the slogan intends to address is that, if you, say, use a secret encryption algorithm to protect your data, you have to keep that encryption algorithm just as secret as you keep your secret key, if you expect that algorithm to buy you any extra protection over, say, triple DES. And that tends to be hard.
You have to describe and encode your algorithm somewhere, and keep that secret. You can measure the number of bits it takes you to describe and encode that algorithm, and add in the number of bits of your encryption key. That's the total number of bits you have to keep secret to protect yourself.
The real truth behind this sort of scenario of the anti-"security through obscurity" meme is that you get better security through putting the same number of bits into a secret encryption key for a well-tested public algorithm than you get for splitting those bits between a secret algorithm and a secret key, because even if your algorithm isn't inferior, it's still using fewer bits of key.
So, sure, security through obscuring your algorithm is bad, because there are better places to spend your secret bits. But "security through secrecy" is fundamental--and I think it's foolish to try to distinguish between "secrecy" and "obscurity".
Back when I was a user of The World, the wstd.* Usenet groups saw frequent complaints (my own among them) about false positives in World's spam blocking. How do you evaluate the tradeoff between blocking spam and accidentally blocking legitimate email?
I could imagine that dictionary-based spam "attacks" might incur a higher per-user cost if they're aimed at theworld.com than if they're aimed at aol.com--same absolute cost, but the difference in number of actual users changes the per-user cost. Is spam-fighting actually more expensive for smaller ISPs than large ones?
Do we have to go through the "guns don't kill people" argument here? Any defense you can offer of the hardware in TCPA and be offered for Palladium as well.
The reality is that getting ubiquitous hardware out there that can do TCPA will inevitably lead to superior DRM on that platform. Whether this will matter beyond DVDs (which already have broken DRM) or not, it's hard to say. But e.g. the music industry decides to finally get off their asses and sell music online, you can be sure it'll be DRM'd as best they can.
But since TCPA offers no real benefit--it's not going to stop Slammer, it's not going to allow you to tell the difference between an executable your friend sent you that you should run and an executable your friend sent you that you shouldn't run--it's hard to see any point to having it in the first place.
The previous ask slashdot post about this with the motherboard vendor (or was it chipset?) just amounted to "our customers want it". Why the hell do their customers (the people who are building machines) want it? They shouldn't give a damn unless maybe they're getting pressure from someone else who wants it, and their customers (you, me, and Joe Hexpack) aren't asking for it, so I don't know who that would be besides MS (who certainly have a history of playing hardball with PC vendors).
But I see little reason (from the story they provide) to conclude that this would have happened without patent protection. Perhaps it was market effects--Boulton and Watt were the big name you trusted since nobody else had been doing this for thirty years. Or maybe, since nobody else could make engines like this and sell them, the potential competitors didn't do that much reverse-engineering or build engines they couldn't sell, just for the experience; perhaps they focused on activities which could actually give them a profit.
The authors use this example twice in the first five pages and it really jumped out at me--and I'm totally anti-patent.
The opera page points out that maybe the Opera page was custom made for Opera 6, and then they run the IE page through Opera 6 and show that it renders fine. Therefore, they say, there's no need for a custom Opera 6 page, therefore it couldn't have been a custom Opera 6 page.
But the right thing to test would be whether Opera 6 renders the "bad pages" acceptably or is similarly broken. I could easily picture a developer trying to customize things for Opera 6 (for some reason or another) and testing something and Opera6 handling it fine and it accidentally getting left in.
I mean, maybe this really totally is MSN screwing with Opera, but the omission of the above data point, and the way they say
when testing the Opera 6 case makes me a tad suspicious.A lot of the value of unrolling comes not merely from reducing the loop overhead (which Duff's Device still does) but from the ability to rearrange the code within the loop, such as moving loads forward and stores later. On a modern processor, out-of-order execution reduces the importance of such simple rearrangements, but it's still the case that it prevents any significant optimizations across the basic blocks.
So instead I'm still using Netscape 4.
But Barry's stance was that since the vast majority of cusomters just wanted all the spam gone, the right thing to do was to accept a certain level of false positives. Unannounced--no warning that you would have legitimate mail returned to your friends with the unhelpful '200 UCE not accepted' or even '200 No thank you' replies (I don't remember the actual number, sorry)--with no "opt-out-of-the-spam-blocking" option for other customers.
One theory I have for The World's problms is that spam-blocking doesn't scale with customers, so The World is hit by it worse than larger ISPs. It seems like the support costs of dealing with customer complaints would scale with customers, though. But, for example, there apparently is (was) a pattern of spammers taking a list of plausible user names and emailing every name on the list @ the target host. Since that list of names is the same length whether it's theworld.com or aol.com, but the number of customers is different, the cost-per-customer for dealing with that (bandwidth / etc.) is higher for the smaller ISP. But nobody at The World was willing to comment on this sort of customer scalability issue (although they mentioned that particular spam scenario because they had a fairly aggressive response to it to avoid bandwidth--they stopped accepting connections from that IP for an hour or two if it was detected, which meant legit mail from that IP was often delayed and sometimes bounced if it kept getting reblocked).
Anyway, the upshot is, I have very little sympathy for somebody who thinks it's a good idea to let legitimate email get blocked as spam because it reduces customer support costs. It's just moving the problems somewhere else where the customers don't know about them.
The original quake lacked transparent water. It was hacked in for glquake, but you had to revis. Quake 2 had support transparent water even with software rendering.
Anyway, the games were developed on P4s under Win2K, running on GeForce4s. I did a lot of the development on an Athlon under Win98, with a Radeon. I wouldn't expect significant Win32 portability issues. (Performance will suck without a top-end machine, though.)
It's possible the missing sprites is the problem--the only one I know Chris Hecker tested running without the sprites was RoboDOOM. I can't suggest any other fixes without a real bug report (what system are you trying it on, etc. etc.).
And even for those three, it doesn't quite tell the whole story. I wrote the least significantly innovative one from a game design standpoint: Very Serious RoboDOOM. This was because I was one of the main engine authors, and I had to spend a lot of my time answering questions, or simply eavesdropping on what people were discussing with each other in case they were having trouble with something I knew how to fix.
We had a big list of 20-some game ideas that we had come up with in the early stages of the engine development--but all the participants were coming up with those ideas so none of them were getting written; I decided to try to write something off that list. Robotron was simple, and hopefully wouldn't be too hard to task switch on, and in fact, it was fairly easy to implement. I didn't end up innovating the game design, but I was able to use the game to make (I think) a comment about the the game-design of shooters and implicitly of the industry--which I think is a fairly good accomplishment for one day's work where I spent most of the time helping other people.
Unfortunately, I made the "used the Doom 2 sprites" mistake, so it's an effort to see the result; as organizers of the event, our focus was so much on making the experience for the participants during the event optimal, we didn't consider the redistribution issue particularly well--it just wasn't our priority.