So far everyone who has replied to you has ignored one thing. A thousand may be fine for seeing a simple "A or B" statistical difference at significant levels; with ANOVA you can even track a few different significant traits.
The number of traits they were trying to discover was unknown at the start; furthermore, they expected it to be very high. Lots of different HTML tags in the standards, but even more nonstandard tags, nonstandard attributes; they even found information about how different attributes are misspelled. Example: nobody can spell "language" on the <script> tag, and they can tell you exactly how many spell it "langauge". They found lots of data points that wouldn't have existed in a sample of a thousand. (My guess is they almost all would existed in a sample of a million, but in numbers too small for statistical significance.)
Most people would have settled for a million, I think, but if you have the resources to get a billion, there actually is useful information in there for you to use.
The same way google efficiently gives you a web page which displays only a few of thousands of hits. You think google loads up all those thousands of summaries for you, renders all those pages, and holds them all in memory for you while you go ahead and click on the first link anyway? It doesn't, for the same reason that email apps don't need to load all those thousands of messages in memory: generators. You start generating until you run out of space to display them, then you don't do anything with the rest. So if your inbox has space to show, say, 25 subject lines, then the app only needs to remember 25 messages.
(And don't give me any objections about the data backend slowing down. Those algorithms should be O(log(n)).)
FTR, I too think it's stupid to have that many messages in your inbox, but only because of the cognitive load, not the CPU load.
Reputations come from *crappy stories*. R.P. (I'm not as familiar with B.B.) posts stories that are crap. They are often unadulterated bullshit, but more likely just overblown hype. They are chosen to generate lots of heat and get people to click on the website, even if just to flame him mightily. Hey, I guess there's no such thing as bad press, right?
A story moderation scheme would allow you to identify (automatically, if you like) submitters whose submissions were frequently crappy and unnewsworthy and got through anyway. Better still would be a more thorough editorial policy, but that seems unlikely to ever happen.
There's nothing wrong with submitting a lot of *good*, *factual* stories that take place in reality. Let's reward signal, not noise.
A formal API is basically published documentation about how it works. Since Microsoft hadn't published the docs about every quirk of their implementation of their APIs, there's a lot of flexibility in how it will be implemented.
WINE takes that a step farther, though.. they're trying to implement the undocumented behaviors too. They do this mostly by running known-working windows software and seeing where it breaks in WINE. Where it breaks, this indicates a place where the WINE implementation of the API either a) doesn't conform to the documented API or b) doesn't conform to the quirks. So WINE does, as you suggested, have to get pretty close to doing exactly what Microsoft does in the implementation.
But even within this restriction there's a lot of wiggle room. No "known-working" software would rely on an exploit; to put it another way, software that made use of the exploit intentionally would not be used to prove WINE's implementation was reliable. Therefore, there's no reason for WINE to implement it that way.
All of this begs the question "where did the exploit come from?" Until I read otherwise, I'm going to assume MS and WINE made use of the same underlying library, in which case WINE is merely sharing a single buggy implementation rather than cloning a new one.
IT' needs access to do its job. We need *total* access to all systems and data or we cant be effective and might as well not goto work.
While I disagree with the whole of this statement, I disagree most vehemently with the part in bold, so I'll address that.
In world that cared about data security, NO EMPLOYEE WOULD EVER BE GIVEN ACCESS TO CUSTOMER DATA THAT WAS ONLY USED TO DRIVE THE APPLICATION. Take a look at the ideas in the book Translucent Databases (actually, even just read the summary on that page) and you'll get an idea of what can be done to minimize the risks posed by insiders. If your company deals with a lot of customer data.. let's say it has an ordering system like Amazon's.. there is NO employee in the company, not the CEO nor the CTO, who needs to know what your customer's credit card number is, or needs to be able to find out. Encrypt it so that only the customer's password can retrieve it (and that password, btw, is only in the customer's brain, because you're only storing a hash of it) and you've just eliminated the single biggest privacy threat in information systems today. The same goes for a wide variety of information about the customers which no employee ever needs to know.
Can we get over mysql already? Even sqlite is a better, and faster, database now. It's ACID, which mysql is not: http://www.sqlite.org/lockingv3.html
And Postgresql is far more robust and performs just as well.
What does mysql offer any more that the other OSS databases don't? Is it just that it's the M in LAMP? I'm so tired of hearing about Mysql, and all the Mysql drama, when it's just a shitty database that has a lot of mindshare.
And yet, there's a valid debate to be had here on whether Linus actually influences policy of others, regardless of whether he wants to. I agree with you on these points: That Linus makes arguments on technical grounds, that his personality has not changed over time, and that he does not wish to be a demagogue or have much interest in influencing the opinion and policy of others.
Yet, he does.
For reasons that are mostly historical he has become a public figure, an August Personage among Open Source communities. He has real accomplishments and made real advances in both technology and mindshare for the open way of doing things. When he started to make these accomplishments, people started to take him seriously, and the more his accomplishments gained momentum, the more important his opinion became. It didn't even matter whether he personally was responsible for the Linux momentum; he was responsible for Linux. Because of this, Linus' words do affect policy, do influence people, and do influence technology.
Therefore I don't think it's wrong to have a debate about whether he should be out there opining on stuff that isn't the Linux kernel. I don't even come down on a different side of the debate than you; I think he should be allowed to speak his mind. It's a free country, even for celebrities. But we are at or nearing the point where his reckless leadership or reckless statements can cause harm to the greater community here, and it's ok for people to start saying to him, "Hey dude, maybe you shouldn't say that. Do you even know what you're talking about? Lots of people think you do." And if agrees, and if he thinks he is being reckless, he's a smart guy and maybe he'll listen.
But let's not dismiss the whole thing as an exercise in wankery. What he says really does matter to some people.
I was in Portland for OSCon 2003. Portland is a beautiful, elegant, well-designed city. I walked or took public transit everywhere in the city and never missed my car. Indeed, the elegance of Portland is second only to the elegance--not to mention the low cost--of its transit system.
Unfortunately that means it's not a very representative challenge for Google Transit. It's an easy target. Let me know when Google Transit can get you around LA, or even SF.
Both times I've had to engage the command line it was because a network card wasn't working properly. In one case the builtin laptop wireless card wasn't turned on in the BIOS (don't ask). In the second case the desktop's ethernet card wasn't seated in the slot all the way. (Just enough, infuriatingly, to get power to make the activity light blink. Not enough to talk to the kernel.)
No operating system in the world could have dealt sensibly with either of those problems, so, for me, Ubuntu is batting 100% on hardware success in about 9 installations.
They're giving away the servers *and* the software. I guess it's the service contract that's the razor.
Given Sun's business acumen the last decade, I expect them to start giving that away too. Not that I'd be happy about that. Competition is good, so competitors shooting themselves in the foot is bad.
Anyone can support Nessus whether they own the code or not. They can't fix bugs in it, but that's not what support is really about. Support consulting is mostly "help us set this up" or "help us customize this". Although I've never used Nessus, I suspect it's highly configurable and customizable, as are most products that have any features meaningful to "support". The company has achieved nothing by this move, and Nessus will probably become much less popular because of it, until an open source replacement for Nessus 3.x appears and the company goes under completely.
At my place of work I migrated all of our servers to Ubuntu months ago. As of Wednesday they are all running Breezy Badger, the newest. Best decision I ever made. If someone tells you Ubuntu is a "workstation OS", not a "server OS", don't believe it. It has all the chewy server goodness, and it supports all the neat server hardware we had in our fancy rackmounts. And it contains recent software.
Tell your technical staff to get with the times:-)
I, too, am all for the jury system. However, it cannot be argued that most people want to be on juries. Most people do not want to be on juries. Those who can, figure out how to get out of being on a jury.
It may be that some significant but small portion of the population feels that jury duty is so important a civic responsibility that they don't try to get out of it. If so, juries may be primarily composed of these people.
But I think this is not the case. I think people who end up on juries do so because they can't figure out how to get out of it. That suggests that these people are at least less creative/resourceful, if not exactly dumber, than the average population.
What do you think of something like this monitor filter? Maybe make it required for students who wish to use their laptops. Seems like most of the time, the laptop users would prefer this anyway; I know on the few occasions I've had a laptop in a lecture setting (conferences, not schools, but basically the same thing) the laptop wasn't distracting, but the ability of other people to read my screen made me uncomfortable, even though I was doing innocuous things.
NAT provides users with a way to make themselves invisible to incoming connections, protecting them from bugs in the operating systems of computers on the LAN. We'll still need that, we just won't call it NAT, and it'll still have most of the same problems. (The step that will be eliminated? "Now go to whatsmyip.org.")
PNG is lossless, so it looks good. JPEG is lossy, so it's small. You can get much better compression on photographic (lots of colors, few uniform color areas, few if any geometric features--this is only my crude layman's impression) images using lossy compression with JPEG, and furthermore, it's adjustable so if you need less quality, you can recover more bandwidth. PNG can only compress such images so far, and no farther.
By contrast, PNG can represent alpha transparency, so (if your browser supports it.. hope IE7 is out soon) you can get neat effects with PNG. PNG is also great at presenting images, created with vector graphics, to software that doesn't do vector rendering. That means logos on web pages, line drawings of all sorts such as scientific plots, etc. But the reason it's good for such things is that such images benefit far more from lossless rendering (than photographs or 2d art or the like).
PNG is great technology, but you can't simply put aside JPEG yet. Like PNG, JPEG is good at what it does, what it's intended for.
Southern California may not have road salt, but it has just about every other unfriendly driving condition you can think of. Heat, crazy traffic, crazy traffic enforcement, etc. I think they could have done a lot worse, if "harsh driving environment" was one of the driving forces.
Another commenter gave a laundry list of software that he routinely installs and uses in his Windows environment. I was pleasantly amused to see how closely it dovetails with my own.
I write software using Python, which is available on most every useful platform. When it's infrastructure-type-stuff, I make sure that all but the very outermost layers are crossplatform. Platform-specific stuff is limited to/etc/init.d or its Windows equivalent the ntservice architecture or the darwin equivalent,/Library/StartupItems.
If it's gui-type-stuff, I use a cross-platform toolkit, such as GTK or wxpython. (I'm very much looking forward to pyqt4, which is supposed to look great on all three of my usual, aforementioned development targets, but it isn't out yet.)
Making these choices up front saves porting costs, and keeps the zealots at bay because your software already does exist on a fully OSS stack.
Someone has already mentioned Needful Things. Flagg's character shows up in a lot of books, often as Randall, or the Walkin' Dude. The Stand also involves Flagg. I can't think of any more off the top of my head, but he has a presence.
Slashdot Mirror
So far everyone who has replied to you has ignored one thing. A thousand may be fine for seeing a simple "A or B" statistical difference at significant levels; with ANOVA you can even track a few different significant traits.
The number of traits they were trying to discover was unknown at the start; furthermore, they expected it to be very high. Lots of different HTML tags in the standards, but even more nonstandard tags, nonstandard attributes; they even found information about how different attributes are misspelled. Example: nobody can spell "language" on the <script> tag, and they can tell you exactly how many spell it "langauge". They found lots of data points that wouldn't have existed in a sample of a thousand. (My guess is they almost all would existed in a sample of a million, but in numbers too small for statistical significance.)
Most people would have settled for a million, I think, but if you have the resources to get a billion, there actually is useful information in there for you to use.
Gabriel's primary concern.
The same way google efficiently gives you a web page which displays only a few of thousands of hits. You think google loads up all those thousands of summaries for you, renders all those pages, and holds them all in memory for you while you go ahead and click on the first link anyway? It doesn't, for the same reason that email apps don't need to load all those thousands of messages in memory: generators. You start generating until you run out of space to display them, then you don't do anything with the rest. So if your inbox has space to show, say, 25 subject lines, then the app only needs to remember 25 messages.
(And don't give me any objections about the data backend slowing down. Those algorithms should be O(log(n)).)
FTR, I too think it's stupid to have that many messages in your inbox, but only because of the cognitive load, not the CPU load.
Reputations come from *crappy stories*. R.P. (I'm not as familiar with B.B.) posts stories that are crap. They are often unadulterated bullshit, but more likely just overblown hype. They are chosen to generate lots of heat and get people to click on the website, even if just to flame him mightily. Hey, I guess there's no such thing as bad press, right?
A story moderation scheme would allow you to identify (automatically, if you like) submitters whose submissions were frequently crappy and unnewsworthy and got through anyway. Better still would be a more thorough editorial policy, but that seems unlikely to ever happen.
There's nothing wrong with submitting a lot of *good*, *factual* stories that take place in reality. Let's reward signal, not noise.
A formal API is basically published documentation about how it works. Since Microsoft hadn't published the docs about every quirk of their implementation of their APIs, there's a lot of flexibility in how it will be implemented.
WINE takes that a step farther, though.. they're trying to implement the undocumented behaviors too. They do this mostly by running known-working windows software and seeing where it breaks in WINE. Where it breaks, this indicates a place where the WINE implementation of the API either a) doesn't conform to the documented API or b) doesn't conform to the quirks. So WINE does, as you suggested, have to get pretty close to doing exactly what Microsoft does in the implementation.
But even within this restriction there's a lot of wiggle room. No "known-working" software would rely on an exploit; to put it another way, software that made use of the exploit intentionally would not be used to prove WINE's implementation was reliable. Therefore, there's no reason for WINE to implement it that way.
All of this begs the question "where did the exploit come from?" Until I read otherwise, I'm going to assume MS and WINE made use of the same underlying library, in which case WINE is merely sharing a single buggy implementation rather than cloning a new one.
Is that so hard a concept to understand?
IT' needs access to do its job. We need *total* access to all systems and data or we cant be effective and might as well not goto work.
While I disagree with the whole of this statement, I disagree most vehemently with the part in bold, so I'll address that.
In world that cared about data security, NO EMPLOYEE WOULD EVER BE GIVEN ACCESS TO CUSTOMER DATA THAT WAS ONLY USED TO DRIVE THE APPLICATION. Take a look at the ideas in the book Translucent Databases (actually, even just read the summary on that page) and you'll get an idea of what can be done to minimize the risks posed by insiders. If your company deals with a lot of customer data.. let's say it has an ordering system like Amazon's.. there is NO employee in the company, not the CEO nor the CTO, who needs to know what your customer's credit card number is, or needs to be able to find out. Encrypt it so that only the customer's password can retrieve it (and that password, btw, is only in the customer's brain, because you're only storing a hash of it) and you've just eliminated the single biggest privacy threat in information systems today. The same goes for a wide variety of information about the customers which no employee ever needs to know.
Your post had nothing to do with the parent. Stay on topic with the thread, don't freeload.
Can we get over mysql already? Even sqlite is a better, and faster, database now. It's ACID, which mysql is not: http://www.sqlite.org/lockingv3.html
And Postgresql is far more robust and performs just as well.
What does mysql offer any more that the other OSS databases don't? Is it just that it's the M in LAMP? I'm so tired of hearing about Mysql, and all the Mysql drama, when it's just a shitty database that has a lot of mindshare.
Try knowledge instead: http://www.bartleby.com/61/80/X0008000.html
Under a libertarian scenario, the cops never show up because the citizen Timmy is threatening hasn't sent his monthly police services check.
Under the current system, however, the victim can shoot him as long as he's feeling threatened. That's called self-defense.
And yet, there's a valid debate to be had here on whether Linus actually influences policy of others, regardless of whether he wants to. I agree with you on these points: That Linus makes arguments on technical grounds, that his personality has not changed over time, and that he does not wish to be a demagogue or have much interest in influencing the opinion and policy of others.
Yet, he does.
For reasons that are mostly historical he has become a public figure, an August Personage among Open Source communities. He has real accomplishments and made real advances in both technology and mindshare for the open way of doing things. When he started to make these accomplishments, people started to take him seriously, and the more his accomplishments gained momentum, the more important his opinion became. It didn't even matter whether he personally was responsible for the Linux momentum; he was responsible for Linux. Because of this, Linus' words do affect policy, do influence people, and do influence technology.
Therefore I don't think it's wrong to have a debate about whether he should be out there opining on stuff that isn't the Linux kernel. I don't even come down on a different side of the debate than you; I think he should be allowed to speak his mind. It's a free country, even for celebrities. But we are at or nearing the point where his reckless leadership or reckless statements can cause harm to the greater community here, and it's ok for people to start saying to him, "Hey dude, maybe you shouldn't say that. Do you even know what you're talking about? Lots of people think you do." And if agrees, and if he thinks he is being reckless, he's a smart guy and maybe he'll listen.
But let's not dismiss the whole thing as an exercise in wankery. What he says really does matter to some people.
I was in Portland for OSCon 2003. Portland is a beautiful, elegant, well-designed city. I walked or took public transit everywhere in the city and never missed my car. Indeed, the elegance of Portland is second only to the elegance--not to mention the low cost--of its transit system.
Unfortunately that means it's not a very representative challenge for Google Transit. It's an easy target. Let me know when Google Transit can get you around LA, or even SF.
Both times I've had to engage the command line it was because a network card wasn't working properly. In one case the builtin laptop wireless card wasn't turned on in the BIOS (don't ask). In the second case the desktop's ethernet card wasn't seated in the slot all the way. (Just enough, infuriatingly, to get power to make the activity light blink. Not enough to talk to the kernel.)
No operating system in the world could have dealt sensibly with either of those problems, so, for me, Ubuntu is batting 100% on hardware success in about 9 installations.
They're giving away the servers *and* the software. I guess it's the service contract that's the razor.
Given Sun's business acumen the last decade, I expect them to start giving that away too. Not that I'd be happy about that. Competition is good, so competitors shooting themselves in the foot is bad.
My ADHD can't handle one more flashing device telling me something has happened on the internets.
Please, give me a mouse that emits white noise, and laserbeams anyone who is trying to approach and distract me.
Anyone can support Nessus whether they own the code or not. They can't fix bugs in it, but that's not what support is really about. Support consulting is mostly "help us set this up" or "help us customize this". Although I've never used Nessus, I suspect it's highly configurable and customizable, as are most products that have any features meaningful to "support". The company has achieved nothing by this move, and Nessus will probably become much less popular because of it, until an open source replacement for Nessus 3.x appears and the company goes under completely.
At my place of work I migrated all of our servers to Ubuntu months ago. As of Wednesday they are all running Breezy Badger, the newest. Best decision I ever made. If someone tells you Ubuntu is a "workstation OS", not a "server OS", don't believe it. It has all the chewy server goodness, and it supports all the neat server hardware we had in our fancy rackmounts. And it contains recent software.
:-)
Tell your technical staff to get with the times
I, too, am all for the jury system. However, it cannot be argued that most people want to be on juries. Most people do not want to be on juries. Those who can, figure out how to get out of being on a jury.
It may be that some significant but small portion of the population feels that jury duty is so important a civic responsibility that they don't try to get out of it. If so, juries may be primarily composed of these people.
But I think this is not the case. I think people who end up on juries do so because they can't figure out how to get out of it. That suggests that these people are at least less creative/resourceful, if not exactly dumber, than the average population.
What do you think of something like this monitor filter? Maybe make it required for students who wish to use their laptops. Seems like most of the time, the laptop users would prefer this anyway; I know on the few occasions I've had a laptop in a lecture setting (conferences, not schools, but basically the same thing) the laptop wasn't distracting, but the ability of other people to read my screen made me uncomfortable, even though I was doing innocuous things.
NAT provides users with a way to make themselves invisible to incoming connections, protecting them from bugs in the operating systems of computers on the LAN. We'll still need that, we just won't call it NAT, and it'll still have most of the same problems. (The step that will be eliminated? "Now go to whatsmyip.org.")
PNG is lossless, so it looks good. JPEG is lossy, so it's small. You can get much better compression on photographic (lots of colors, few uniform color areas, few if any geometric features--this is only my crude layman's impression) images using lossy compression with JPEG, and furthermore, it's adjustable so if you need less quality, you can recover more bandwidth. PNG can only compress such images so far, and no farther.
By contrast, PNG can represent alpha transparency, so (if your browser supports it.. hope IE7 is out soon) you can get neat effects with PNG. PNG is also great at presenting images, created with vector graphics, to software that doesn't do vector rendering. That means logos on web pages, line drawings of all sorts such as scientific plots, etc. But the reason it's good for such things is that such images benefit far more from lossless rendering (than photographs or 2d art or the like).
PNG is great technology, but you can't simply put aside JPEG yet. Like PNG, JPEG is good at what it does, what it's intended for.
Southern California may not have road salt, but it has just about every other unfriendly driving condition you can think of. Heat, crazy traffic, crazy traffic enforcement, etc. I think they could have done a lot worse, if "harsh driving environment" was one of the driving forces.
Another commenter gave a laundry list of software that he routinely installs and uses in his Windows environment. I was pleasantly amused to see how closely it dovetails with my own.
/etc/init.d or its Windows equivalent the ntservice architecture or the darwin equivalent, /Library/StartupItems.
I write software using Python, which is available on most every useful platform. When it's infrastructure-type-stuff, I make sure that all but the very outermost layers are crossplatform. Platform-specific stuff is limited to
If it's gui-type-stuff, I use a cross-platform toolkit, such as GTK or wxpython. (I'm very much looking forward to pyqt4, which is supposed to look great on all three of my usual, aforementioned development targets, but it isn't out yet.)
Making these choices up front saves porting costs, and keeps the zealots at bay because your software already does exist on a fully OSS stack.
As a person who uses vim for everything up to but not including sex with my wife, I gotta say you're right about the yellow blocks.
:set nohls
Try this --
(that stands for highlightsearch)
Or add it to your ~/.vimrc (sans the colon).
Someone has already mentioned Needful Things. Flagg's character shows up in a lot of books, often as Randall, or the Walkin' Dude. The Stand also involves Flagg. I can't think of any more off the top of my head, but he has a presence.