"My point wasn't that there's an exact parallel between the past and future, just that the idea that anonymity has been continuously diminishing is false."
Yeah, I guess you're right. I guess my point is that anonymity is not on a scale, either more or less, but that it is *qualitatively* changing.
No, supermakret cards are totally different than you knowing the store owner in the country.
If you lived in the country, where everyone knows everyone, there is a symmetry of non-privacy. Gus the storeowner knows you, knows your dirt. You know him and know his dirt. When your purchases are recorded on a supermarket scanner, you have no idea who sees this data, and you certainly aren't entitled to see the records of *their* shopping habits.
In the country, everyone knows everyone, and you have at least a minimum of a personal relationship. Everyone has a vested interested in doing good by everyone, at least to maintain their reputation, and heck, they might even like you and care about you.
Your modern supermarket purchases, however, probably aren't even read by an individual. They are just used to manipulate your purchases so the store can maximize the money they make off of you. Any individual who sees your purchase data won't actually be there to help you repair your car on a Sunday afternoon. Also, the data that is collected is probably scanned by law enforcement. Gus the storeowner might have a good idea about what you purchase, but he's not regularly spilling the beans to the feds about it.
The 'technology' involved in everyone knowing everyone is the hard-wired human person and group tracking. This has probably been going on since before Homo sapiens. The technology involved in supermarket purchase scanning is qualitatively different, and can be used in ways that 'knowing everybody in a small town' cannot be used.
What about size differentials? If you have a human with a sword going up against a dragon, how is that human going to hit anything other than a toe or foot? From your response, it sounds like this is not codified at all, it's just left up to the discretion of the GM. With some creatures, it seems like you could never logically get a critical hit, unless you had some kind of prop to use, like a nearby rope.
I was just browsing through the screenshots and I saw a sword-wrangler taking a swipe at a giant's shins. Does this kind of kill the suspension of belief that is to easy to make when playing with paper & pencil?
So how realistic is the fighting anyway? I think in 'real life', if you managed to strike a sword across a naked shin *just once*, you have a disabled giant. Do you have to make 40 hacks and then the giant vaporizes? Or do the D&D rules already cover this?
You make a task where the participant gets a reward if it is completed succesfully with a partner. Give them something they want, which, for the child and chimpanzee, is not money. Probably candy for the child and fruit for the chimp. Have them run through the test successfully several times with the researcher, so they get the idea that they get the reward if the test goes well. Then, replace the researcher with a child or a chimp.
I'm going to go ahead and say they look good without make up.
It's like the old saying, "you can't polish a turd". The women that are selected for promotion to celebrity already have good bone structure, symmetrical features, etc. If you really want to see ugly people, just walk down the street.
Also, those are really bad photographs. They are not a good indication of how pretty the person comes off as in real live, face-to-face.
There's the run-down: Christina Appelgate -- looks like she didn't get much sleep. Photograph snapped at the end of a word, which makes her neck look weird. Pamela Anderson -- I don't see anything wrong with this. Cameron Diaz -- This is the worst picture of her? She looks gorgeous. Alicia Silverstone -- OK, this picture was snapped when she had her mouth open and food in it. No wonder it's no flattering. Also, she may have been a little heavier than in the picture on the left. It certainly doesn't make her ugly. Britney Spears -- I just don't get it, she doesn't look bad. She's cute.
Of course, no woman is as beatiful as a woman with make-up on. The prettiest woman in the world couldn't compete against a woman wearing mascara and all the other tricks.
You want to see phenomenal growth? How about making phenominal music, not phenominal marketing campaigns. Otherwise, the only music you're going to sell is people buying their favorites in the new format, like they did from vinyl -> tapes, tapes -> CDs, CDs -> mp3s.
Find some really talented artists, not hyper-hyped hot chicks, make it available only digitally and watch the dollars roll in.
From what I got from the conversation, it wasn't the graphics that was the problem. It was the extensive word and phrase prediction that drive dasher:
> [Me] Do you think the slowness problem comes from the smooth animation
> of the letters scrolling?
[Chris] No, I think it comes directly for the probability calculations needed to determine the size for each Dasher node. The CPU on the 770 really isn't that high-spec.
I'm very keen on a true tablet PC system ( I hate Palms and other similiar devices -- they don't really seem designed to fully utilize a stylus input), but I don't have $1800-$2500 to spend. I have a 770 on backorder.
I was emailing Chris Ball, one of the developers of dasher, which is a very novel and efficient method for character and word input. Unfortenately, I was dismayed to learn that:
We finished the port. Problems:
It runs too slowly. We did some basic optimisation, but it's still too slow to feel good.
The hildon-input-method dynamic library is closed-source, so we can't get it working as an input method. This pretty much removes any further motivation we had for spending our time on it; if you care, I suggest complaining to Nokia.
So I don't think we're planning a release.
What a shame. I thought that with the maemo platform being open-source, this would be a killer device.
Hey, if you're a hotshot programmer, you're probably fullly aware of how any kind of talent, good ideas, or cool implementations can be stymied by management. My guess is they had a lot of great names attached, but the studio just wanted to get this out the door so they could cash in on the LOTR in December phenomenon.
Read this post for my reply to another poster who asks the same question. Basically, GPG is actually *simpler* than the crappy HTTP session authentication hacks going around. Also, wikipedia is not repsonsible for the creditials of the entire wikipedia community.
"So as soon as someone corrects a spelling-mistake, the whole section is marked as untrustworthy?"
No. Remember, my original suggestion says 'whole' or 'part'. It doesn't say how small a part is.
You could easily design a system that check whether a non-dictionary string becomes a *very similiar* dictionary word, consider that a spellcheck, and not change the ranking of the article.
"My feeling is that if major banks, credit card providers, and brokerages get away with simple login/pass systems for account access with potentially billions of dollars at stake..."
This is good enough for financial institutions because they have other authentication mechanisms for fraud detection. Just for example, they have hueristics to examine whether this transaction makes sense given the history of the customer. And for non-personal accounts, they don't allow username/password web access for making transactions. Billions are actualy *not* at stake, because they go way beyond simple username and password for checking out transactions.
With wikipedia or any other bulletin board system, username/password is *all they have*.
"I'd suggest you drop the GPG aspect, because it just confuses and obfuscates the explanation of what's fundamentally a good idea."
I would argue that what is a good idea is reputation. A GPG system creates a reputation system that *persists* across the web. This is the fundamental good idea. If we could rank information *anywhere*, not just on Wikipedia, that would be great.
Read this post for my repsonse to someone else asking about the value of GPG.
Boils down to this:
. GPG authentication is *simpler* than the 1000 different crappy authentication-over-HTTP-sessions schemes going around. . Wikipedia isn't responsible for maintaining the authentication credentials for the entire community. It's good not to have all your eggs in one basket. . In the long run, GPG could replace any crappy authentication system in any bulliten board system anywhere. It would *simplify* the web.
1. It's harder to steal someone's GPG identity.
2. You're not putting all your eggs in one basket like you do with logins. If wikipedia had a catastrophic server failure, they might lose all the authentication data. Goodbye wikipedia community. With GPG keys, there isn't such a large risk.
Here's a feature you may be overlooking: GPG keys are *universal* username/password credentials. Any bulliten board system could use GPG signed messages. That would do away with everybody re-inventing this authentication system and site security.
I would argue that GPG authentication is actually simpler than a username/password over HTTP security system. If that's the case, how can you call it overengineering, especially if any other bulliten board can drop their lousy HTTP authentication mechanism and use this one? That reduces complexity for site admins all over the world.
About a year ago, I posted a discussion to some part of Wikipedia advocating digitally signing articles with GPG keys.
The plan was that each author, editor, and reader signs off for or against part or the whole of an article. The fallout should be that some articles get nearly universal positive sign offs, some get nearly universal against votes, and some are recorded as controversial. With GPG keys, we can also start ranking authors and editors -- are they generally agreed with, are they controversial, are they trolls. This is a codification of the skepticism that proponents of Wikipedia claim that any internet user should employ.
Something else I thought would be good would be to have branching articles. For instance, the entry for Hitler would have the main entry, which is the most agreed upon, a white-supremacist/neo-nazi version which stirs a lot of controversy, and maybe a David Icke version, which, while against Hitler, involves space reptiles and is therefore also controversial. Using the ranking and reputation system, a casual user can see how agreeable or controversial an article is.
Slashdot Mirror
"My point wasn't that there's an exact parallel between the past and future, just that the idea that anonymity has been continuously diminishing is false."
Yeah, I guess you're right. I guess my point is that anonymity is not on a scale, either more or less, but that it is *qualitatively* changing.
No, supermakret cards are totally different than you knowing the store owner in the country.
If you lived in the country, where everyone knows everyone, there is a symmetry of non-privacy. Gus the storeowner knows you, knows your dirt. You know him and know his dirt. When your purchases are recorded on a supermarket scanner, you have no idea who sees this data, and you certainly aren't entitled to see the records of *their* shopping habits.
In the country, everyone knows everyone, and you have at least a minimum of a personal relationship. Everyone has a vested interested in doing good by everyone, at least to maintain their reputation, and heck, they might even like you and care about you.
Your modern supermarket purchases, however, probably aren't even read by an individual. They are just used to manipulate your purchases so the store can maximize the money they make off of you. Any individual who sees your purchase data won't actually be there to help you repair your car on a Sunday afternoon. Also, the data that is collected is probably scanned by law enforcement. Gus the storeowner might have a good idea about what you purchase, but he's not regularly spilling the beans to the feds about it.
The 'technology' involved in everyone knowing everyone is the hard-wired human person and group tracking. This has probably been going on since before Homo sapiens. The technology involved in supermarket purchase scanning is qualitatively different, and can be used in ways that 'knowing everybody in a small town' cannot be used.
"What I talk to an old friend is ABSOLUTELY none of the governments business."
Not even if you two are seriouly planning on flying planes into buildings or releasing sarin gas in a subway?
What about size differentials? If you have a human with a sword going up against a dragon, how is that human going to hit anything other than a toe or foot? From your response, it sounds like this is not codified at all, it's just left up to the discretion of the GM. With some creatures, it seems like you could never logically get a critical hit, unless you had some kind of prop to use, like a nearby rope.
I was just browsing through the screenshots and I saw a sword-wrangler taking a swipe at a giant's shins. Does this kind of kill the suspension of belief that is to easy to make when playing with paper & pencil?
So how realistic is the fighting anyway? I think in 'real life', if you managed to strike a sword across a naked shin *just once*, you have a disabled giant. Do you have to make 40 hacks and then the giant vaporizes? Or do the D&D rules already cover this?
So you can beat kids with soap wrapped in a towel?
You make a task where the participant gets a reward if it is completed succesfully with a partner. Give them something they want, which, for the child and chimpanzee, is not money. Probably candy for the child and fruit for the chimp. Have them run through the test successfully several times with the researcher, so they get the idea that they get the reward if the test goes well. Then, replace the researcher with a child or a chimp.
I'm going to go ahead and say they look good without make up.
It's like the old saying, "you can't polish a turd". The women that are selected for promotion to celebrity already have good bone structure, symmetrical features, etc. If you really want to see ugly people, just walk down the street.
Also, those are really bad photographs. They are not a good indication of how pretty the person comes off as in real live, face-to-face.
There's the run-down:
Christina Appelgate -- looks like she didn't get much sleep. Photograph snapped at the end of a word, which makes her neck look weird.
Pamela Anderson -- I don't see anything wrong with this.
Cameron Diaz -- This is the worst picture of her? She looks gorgeous.
Alicia Silverstone -- OK, this picture was snapped when she had her mouth open and food in it. No wonder it's no flattering. Also, she may have been a little heavier than in the picture on the left. It certainly doesn't make her ugly.
Britney Spears -- I just don't get it, she doesn't look bad. She's cute.
Of course, no woman is as beatiful as a woman with make-up on. The prettiest woman in the world couldn't compete against a woman wearing mascara and all the other tricks.
You want to see phenomenal growth? How about making phenominal music, not phenominal marketing campaigns. Otherwise, the only music you're going to sell is people buying their favorites in the new format, like they did from vinyl -> tapes, tapes -> CDs, CDs -> mp3s.
Find some really talented artists, not hyper-hyped hot chicks, make it available only digitally and watch the dollars roll in.
If only I had modpoints...!
What did I just get done asking you?
BTW it's Mjg59, not Mig59.
Argh! Please don't make fun of my slashdot skills.
Link?
Hey, I'm a PHP hacker. I have a liberal arts degree and I have a tough time with anything more advanced than binomial equations.
The source code is freely available. I would really enjoy it if you were able to make dasher run at a reasonable speed on the 770.
From what I got from the conversation, it wasn't the graphics that was the problem. It was the extensive word and phrase prediction that drive dasher:
> [Me] Do you think the slowness problem comes from the smooth animation
> of the letters scrolling?
[Chris] No, I think it comes directly for the probability calculations needed to determine the size for each Dasher node. The CPU on the 770 really isn't that high-spec.
Here's a paper that Chris directed me to.
I was emailing Chris Ball, one of the developers of dasher, which is a very novel and efficient method for character and word input. Unfortenately, I was dismayed to learn that:
We finished the port. Problems:
So I don't think we're planning a release.
What a shame. I thought that with the maemo platform being open-source, this would be a killer device.
Hey, if you're a hotshot programmer, you're probably fullly aware of how any kind of talent, good ideas, or cool implementations can be stymied by management. My guess is they had a lot of great names attached, but the studio just wanted to get this out the door so they could cash in on the LOTR in December phenomenon.
OK, you're right. I think we are getting at the same point.
Now I understand your point. It is a good one.
If you're planning to take over the world, do it one city at a time.
Read this post for my reply to another poster who asks the same question. Basically, GPG is actually *simpler* than the crappy HTTP session authentication hacks going around. Also, wikipedia is not repsonsible for the creditials of the entire wikipedia community.
"So as soon as someone corrects a spelling-mistake, the whole section is marked as untrustworthy?"
No. Remember, my original suggestion says 'whole' or 'part'. It doesn't say how small a part is.
You could easily design a system that check whether a non-dictionary string becomes a *very similiar* dictionary word, consider that a spellcheck, and not change the ranking of the article.
"My feeling is that if major banks, credit card providers, and brokerages get away with simple login/pass systems for account access with potentially billions of dollars at stake..."
This is good enough for financial institutions because they have other authentication mechanisms for fraud detection. Just for example, they have hueristics to examine whether this transaction makes sense given the history of the customer. And for non-personal accounts, they don't allow username/password web access for making transactions. Billions are actualy *not* at stake, because they go way beyond simple username and password for checking out transactions.
With wikipedia or any other bulletin board system, username/password is *all they have*.
"I'd suggest you drop the GPG aspect, because it just confuses and obfuscates the explanation of what's fundamentally a good idea."
I would argue that what is a good idea is reputation. A GPG system creates a reputation system that *persists* across the web. This is the fundamental good idea. If we could rank information *anywhere*, not just on Wikipedia, that would be great.
Read this post for my repsonse to someone else asking about the value of GPG.
Boils down to this:
. GPG authentication is *simpler* than the 1000 different crappy authentication-over-HTTP-sessions schemes going around.
. Wikipedia isn't responsible for maintaining the authentication credentials for the entire community. It's good not to have all your eggs in one basket.
. In the long run, GPG could replace any crappy authentication system in any bulliten board system anywhere. It would *simplify* the web.
There are two reasons to use GPG technology:
1. It's harder to steal someone's GPG identity.
2. You're not putting all your eggs in one basket like you do with logins. If wikipedia had a catastrophic server failure, they might lose all the authentication data. Goodbye wikipedia community. With GPG keys, there isn't such a large risk.
Here's a feature you may be overlooking: GPG keys are *universal* username/password credentials. Any bulliten board system could use GPG signed messages. That would do away with everybody re-inventing this authentication system and site security.
I would argue that GPG authentication is actually simpler than a username/password over HTTP security system. If that's the case, how can you call it overengineering, especially if any other bulliten board can drop their lousy HTTP authentication mechanism and use this one? That reduces complexity for site admins all over the world.
About a year ago, I posted a discussion to some part of Wikipedia advocating digitally signing articles with GPG keys.
The plan was that each author, editor, and reader signs off for or against part or the whole of an article. The fallout should be that some articles get nearly universal positive sign offs, some get nearly universal against votes, and some are recorded as controversial. With GPG keys, we can also start ranking authors and editors -- are they generally agreed with, are they controversial, are they trolls. This is a codification of the skepticism that proponents of Wikipedia claim that any internet user should employ.
Something else I thought would be good would be to have branching articles. For instance, the entry for Hitler would have the main entry, which is the most agreed upon, a white-supremacist/neo-nazi version which stirs a lot of controversy, and maybe a David Icke version, which, while against Hitler, involves space reptiles and is therefore also controversial. Using the ranking and reputation system, a casual user can see how agreeable or controversial an article is.