1.) Microsoft is somehow responsible for all third-party DLLs on a system. Their scanner must contain a self-sufficient, learning AI that just "knows" which DLLs to scan on any system in existence.
Yes, this makes perfect sense because the GDI detection tool and surrounding infrastructure as it stands now is so perfect that to enhance it one iota (say, by having it actually do something useful) would be to make it impossibly perfect.
Every time the darn thing runs it merely says you *may be vulnerable* and as far as I can tell it doesn't every do anything else. I've written "Hello World" applications with more pragmatic value. Not only that, but you run it on a Windows XP SP2 system, and then go to MS' website and find out that the tool can do you *no good* and should never have been downloaded because WinXP SP2 is not vulnerable to this problem!! (Or at least, not in a way fixable by this tool)
In my last WinXP SP2 full install, this was a major "head scratcher" I had getting the system up and running. Why would they ask me to download and run a tool that can't possibly do my version of Windows any good? (Only now am I beginning to realize this makes a twisted sense because the tool does my computer as much good as any other...none.) Or, perhaps there's more to the GDI exploit story. But where the heck is the more? Somebody at Microsoft really fell down on this one.
2.) Mozilla was affected by this same vulnerability, but it's okay because it's Mozilla and not Microsoft.
Fixed in Mozilla 1.4.1 In October of 2003. Not even a speedbump, just another patch in the quilt.
3.) When Mozilla's XUL bug was marked "Confidential" since 1999 only to be revealed earlier this year when exploits came out for it, that's okay too. There won't be any "open letters" to Mozilla over it, because it's Mozilla and not Microsoft.
Ya, that was a cover-up worthy of a major corporation... Not the greatest thing to do, but I don't see what this has to do with the current story. (Ie: What does keeping exploits secret have to do with really lousy exploit detection/resolution tools?)
Reading articles like this brings a queasy feeling on. Has the Linux community become complacent or even worse a little over-confident?
This is hardly the "linux community" This is slashdot... A more wretched hive of, ahh nevermind.
If you want community, try comp.os.linux.answers, debian-user@lists.debian.org, or newbie@linux-mandrake.com.
...the condescending attitude newbies often face from some Linux zealots doesn't help the situation either.
Ya, but thrice a year MSDN support phone-calls in which developers get to pay cold hard cash for the opportunity to train MS interns certainly makes the OSS model of open dialog a little more appealing. Sure, a few random individuals sneak in the door that don't belong, but if the choices are out in the cold, or in a comfy tent with an obnoxious neighbor, I'll take the tent every time.
3M actually approved this non business related use of company resources
Um... you just made the parents point. Closed source is authoritative and requires "permission" to "make cool hacks". OSS is the opposite.
This authoritarianism is an impediment to innovation. Why should anyone "need permission" to hack on a piece of software they have aquired a legal copy of?
I think the economics of this are more complicated than they seem at first glance.
Ya, it's called monopolies on broadband provision. (Which is only natural for a utility, the problem is you're stuck with the utility provider and anything they want to force you into.)
It will be done in the name of "security" and "fighting viruses".
Guess I need to bone up on my trusted computing virus writing skills then... Root chains have inherent weaknesses completely aside from software/hardware implementation. I should think having the right couple keys would make it easier, rather than harder, to exploit.
However what do you expect to do if every broadband ISP announces such terms pretty much simultaneuously?
Ya, I worry about that one a lot. Being a comcast user, I'm already stuck with an ISP I can't stand. I'd pay a significant premium to be able to use someone else instead.
If my ISP suddenly required me to have a "trusted computer" to connect to the internet, I'd drop them like a hot potato. Ditto for all the rest of these "trusted" devices, gadgets and services.
actually, considering that e is an infinite sequence and all, I suspect there's better than even odds that it does occur somewhere in there...
Especially if you buy into the whole "all data of civilization is stored in pi/e/<random possibly infinite number here>" claptrap. I mean come on. he posted it to slashdot, it better be in there if all the knowledge of the human race is in there.
That's true as far as it goes, but paypal isn't going to be able to levy a very big fine (at least against me) just for paying $10 one month for a pr0n subscription. Whether the fines stick on the other end is a different matter. (The article seemed to imply both buyers and sellers would be fined, which doesn't seem fully feasible)
Well, that was for reference--the rest of my post is the wee bit shorter version.:-) To be fair to Rowe, though, it's about as short as philosophical tracts get...just over 9 pages in my text.
My bad, I thought you were refering to the entire text.
You've got me just right, except that my argument is yet more sweeping than just addressing Anselm's definition of god. This is my semantic argument against god, and it applies to all such arguments for or against the existence of such a being.
This is all fine and dandy, and prefectly rational, but it doesn't gain you any persuasive traction, it merely moves you almost immediately to the "agree to disagree" point where meaningful discussion ends. Being more sweeping in this matter just makes you more disagreeable... (Though hopefully agreeably so. Err, i think.)
If you want me to put forward definitions of God and soul for you I can certainly try, but you're as free to deny them credibility as I am to deny realism as a basic assumption. (Also, you're the one who brought "soul" up, I never mentioned it... It is a word I prefer not to use because it is so non-specific and often abused.)
I understand your woodchuck argument. However, I must point out that the possibility to do something must exist before that something can actually be done. The existence of the form provides the possibility, while imperfect implementations of that form that exist in reality attempt to meet that full potential...but always fall short to some extent.
I guess a question would be can anything "real" ever avoid falling short of "ideal". I'm not convinced that question has been answered in the positive or negative, and I certainly am not going to accept it one way or the other just based on considering "great" as Anselm used it to be equivalent to "ideal" as Plato used it.
If it is possible that something "real" can meet "ideal standards" then it is possible that God exists (ie: the concept of God exists/is rational/what have you), and Anselm's argument holds, at least in this regard.
Note: If pressed, I'd argue that any Platonic Ideal is every bit as nebulous as God or soul, at least to my way of thinking... (Can't be nailed down, can be different to different folks etc.)
This may be true for the referent line of thought, but there are other lines of thought that posed serious problems for Anselm shortly after the publication of his argument. For instance, Gaunilo argued that Anselm's argument could be applied to all sorts of things that we know do not exist...simply replace "god" with "island," and we can prove that the island than which none greater is possible exists in reality.
This is a great argument, although in my opinion it goes too far. You see, according to Anselm, the concept "God" contains, in definition, this notion of "greatest" as a necessary component. "island" does not have such a connection with the concept of "greatest" (we can conceive of the greatest island, but there is no requirement for island to be tied to greatest). That said, the only way this argument can defeat Anselm is by attacking his definition of God...
So, you've got 3 successful attacks on Anselm's definition of God, but all more or less indirect. (meaning you can't just say: "No Anselm's got it wrong, God might exist and is this other way..." Therefore someone convinced God *might* exist and persuaded by Anselm's language is still going to be unsatisfied. As many were.)
I think this is where we begin to disagree... I will say this, though: I do think that everyone, no matter what the basis for their worldview, that set of ideas/beliefs is ultimately based on some set of fundamental, axiomatic beliefs from which their understanding and personal philosophy derives. However, everyone's set of axiomatic beliefs are not created equal...some are demonstrably better than others. One person's set of axiomatic beliefs may result in a system of thought that i
If there's nothing stopping random incoming ports, to Windows boxes in particular, you're SOL no matter what else you've got going. These days viruses can saturate a population within a couple days, much faster than virus definitions can be distributed.
900 years? The "whole thing" didn't last 6 months. Logicians and scholars almost immediately poked holes in his argument.
The point is that it is still interesting and worth debate, and has been for 900 years...
For a good argument that addresses the most popular of these points, read William L. Rowe's The Ontological Argument from Reason & Responsibility, Readings in Some Basic Problems of Philosophy, 8th Ed., by Joel Feinberg.
Um... Got anything just a wee bit shorter? Or maybe some useful excerpts?
So, that's the first problem. I would argue that Anselm's assumption that the concept of god exists is incorrect.
Okay, so you've just argued that the notion of "God", or in this case Anselm's "That than which nothing greater can be conceived" is incoherent. In doing so you've decided to disagree at the level of presupposition, and can now never have any meaningful argument with someone who considers the notion of "God" (and/or "that than which...") coherent, or even possibly so. Have fun.
I tend to go with Plato on this one, and not just in opposition to Anselm's argument...
Okay, that's fine. I happen to go with Sir Whilhelm Henry of Okham (or Occam whatever) and avoid multiplying entities unnecessarily. Of course, that has little to do with what Anselm is arguing about...
The problem in this case is that God is essentially a giant "do gooder" and a "do gooder" that doesn't "do good" isn't much "good". How much good can a do gooder do if a do gooder doesn't exist? You might argue that "good" is also an incoherent concept. That's fine, but you'd be missing my point. Try "woodchuck" instead of "do gooder" if that helps. Note: The woodchuck that exists is more woodchuck than any ideal, the woodchuck that doesn't exist is more ideal than any woodchuck, but a woodchuck that doesn't exist is only an ideal, not really a woodchuck at all.
Anselm argues that, no, there is one yet greater thing possible--that the concept of god itself exists. But this assumes that such an existence is possible...perhaps existence in our understanding is the greatest possible thing because to exist in reality would be not be possible.
I'm not sure I follow your whole argument in #3... Perhaps you're merely agreeing with what I said originally? To properly conceive of Anselm's notion of "God" one must conceive of him/her as existing. This is true WHETHER OR NOT "God exists" "in reality".
It wasn't until Kant that this line of thought was well-established... a bit more than 6 months.
Of course, many would also argue that leaps of faith are required for normal day to day life, active science, and the like.
Yes, we all make tiny leaps of faith every day for practical reasons, just to get things done. ... You are arguing here, it seems, that since we often make leaps of faith, some correct, some not, that making leaps of faith is generally an acceptable practice in determining what to believe.
I'm just pointing out that leaps of faith are not always bad. In fact they are necessary as a basis for knowledge and understanding.
Here's an article apologetic to the use of methodological naturalism in science. The author makes the following statement: "[Methodological Naturalism] is what science employs, the belief that natural events have natural causes and that the physical world is logical and understandable." He goes on to make another statement: "Science itself, which uses methodological and not metaphysical naturalism, assumes that all events it can observe and study are natural in origin."
This seems a rather large and singular leap in a world of Heisenberg's uncertainty, chaos theory and quantum fluctuation (not to mention free-will).
If a leap of faith in this case is justified for some other reason, I am at a loss to see it. Can you tell me?
Objective knowledge requires leaps of faith, is it so strange to think subjective knowledge (or spirituality whatever your preference) also requires leaps of faith?
Re:The problem I have with essays....
on
The Age of the Essay
·
· Score: 2, Funny
No. Go to your room.
Now if you'd put spelling/grammer, you might've gotten away wit it.
It would, indeed, be a good world where everybody would read everything regardless of style or content...but that's not going to happen.
Shyah! Why should people waste their time working and spending time with friends and family when they could be reading every single hair-brained post on slashdot... and every poorly written weblog... and every badly written book... etc...
Anselm's argument for the existence of god, in the philosophy course I took (and every one I've ever heard of), is immediately followed by several arguments to the contrary that poke one giant hole after another in it.
I was aware that Anselm's argument was quite often side-stepped and berated but I wasn't aware there were any holes in it. (Hence the whole lasting for 900 years thing.)
To me, Anselm winds up showing that to properly conceive of God you have to conceive of God as existing, because otherwise you're leaving out a very major and obvious piece of the puzzle before you even start. Whether this "conception" and reality have anything to do with each other or if "God" is even a coherent notion is entirely another matter. Nevertheless, Anselm's argument seems to lend perspective to a lot of theist vs atheist debate. (disagreements at the level of presupposition are pretty hard to work out...)
If you really believe that there can be a logical argument that proves the existence of god, then you are in opposition to nearly every great philosopher, professor, rabbi, priest, theological scholar, etc...they all say that a leap of faith is necessary for the big 3 religions (Islam, Judaism, and Christianity).
Yes. Of course, many would also argue that leaps of faith are required for normal day to day life, active science, and the like. The disagreements are about which leaps make sense and whose are smaller.
By the tone of your post I'm not sure if you don't understand or are bitter.
Perhaps a bit of both... (or maybe just jaded) I'd inferred you were including GUI development in reference to your comments, since that's what the article was about and I don't see how testing can be simplified that far for GUIs.
inputs are finite and outputs are finite...
By that logic all encryption is "breakable" because it requires finite computations to break it. In a sense this is true, but that sense isn't very practical.
It is quite trivial to create a system in which the number of possible inputs is much too large to feasibly test. Theoretically, 128 bits of meaningful data is plenty to ensure no exhaustive proof is feasible in the near term. More practically GUI inputs including timing, mouse position, button clicking and general mayhem provides much larger sets of possible input.
Your point... "The only way you can "prove" your program is correct is by mangling the specification one way or another" is absolutely false...
That depends on your definition of specification. I was thinking along the lines of what is requested/desired by a customer, in the form they choose to request it in. You seem to have already limited specification to mean only "a precise and testable set of inputs".
I suppose I was thinking mostly of custom software , since that is the area I work in, but even a specification drawn up based on case studies, user surveys and the like can contain elements that don't break down well into small testable pieces.
The simple point is that certain types of specification can be validated to one degree or another (statistically if not exhaustively), but fulfilling the user's desires and intents, not meeting a specification drawn up by consultants and/or developers, is the primary goal.
Using proofs based on specifications is a bit like using test-driven development. The development process can become more controlled and predictable, but when you step back and look at overall project success rate and customer satisfaction, the numbers aren't much better than any of the other methodologies. This is because the rift between end-user and satisfaction exists just the same whether it is between user requests and developer understanding of the problem or between developer understanding and implementation. Decreasing the rift between developer understanding and implementation can be very helpful, but tends to increase overall cost and limit developer flexibility (ie: we can't do X because it isn't easily testable) therefore decreasing overall customer satisfaction.
(Note: All I really know about test driven development is what I've read. In trying it, I've read through a few examples, but they always leave a bad taste in my mouth... High level acceptance testing seems to have much more aesthetic appeal. )
As a counter-example to proof/test driven design I'll hold up the tiniest known C implementation of the CSS descrambler at the gallery of CSS descramblers. This is, in my opinion, an awesome piece of code. I don't see how anything close (in terms of size and speed) could be written using TDD or some other proof driven form of development. (obviously some level of testing is possible on this piece of code, but there's certainly no easy way I can think of to verify it as "100% correct" or "contains no bugs") Note: if you're going to read the code, you'll want to use a de-obfuscating tool of some sort to fix the lack of white space...
Slashdot Mirror
1.) Microsoft is somehow responsible for all third-party DLLs on a system. Their scanner must contain a self-sufficient, learning AI that just "knows" which DLLs to scan on any system in existence.
Yes, this makes perfect sense because the GDI detection tool and surrounding infrastructure as it stands now is so perfect that to enhance it one iota (say, by having it actually do something useful) would be to make it impossibly perfect.
Every time the darn thing runs it merely says you *may be vulnerable* and as far as I can tell it doesn't every do anything else. I've written "Hello World" applications with more pragmatic value. Not only that, but you run it on a Windows XP SP2 system, and then go to MS' website and find out that the tool can do you *no good* and should never have been downloaded because WinXP SP2 is not vulnerable to this problem!! (Or at least, not in a way fixable by this tool)
In my last WinXP SP2 full install, this was a major "head scratcher" I had getting the system up and running. Why would they ask me to download and run a tool that can't possibly do my version of Windows any good? (Only now am I beginning to realize this makes a twisted sense because the tool does my computer as much good as any other...none.) Or, perhaps there's more to the GDI exploit story. But where the heck is the more? Somebody at Microsoft really fell down on this one.
2.) Mozilla was affected by this same vulnerability, but it's okay because it's Mozilla and not Microsoft.
Fixed in Mozilla 1.4.1 In October of 2003. Not even a speedbump, just another patch in the quilt.
3.) When Mozilla's XUL bug was marked "Confidential" since 1999 only to be revealed earlier this year when exploits came out for it, that's okay too. There won't be any "open letters" to Mozilla over it, because it's Mozilla and not Microsoft.
Ya, that was a cover-up worthy of a major corporation... Not the greatest thing to do, but I don't see what this has to do with the current story. (Ie: What does keeping exploits secret have to do with really lousy exploit detection/resolution tools?)
Reading articles like this brings a queasy feeling on. Has the Linux community become complacent or even worse a little over-confident?
...the condescending attitude newbies often face from some Linux zealots doesn't help the situation either.
This is hardly the "linux community" This is slashdot... A more wretched hive of, ahh nevermind.
If you want community, try comp.os.linux.answers, debian-user@lists.debian.org, or newbie@linux-mandrake.com.
Ya, but thrice a year MSDN support phone-calls in which developers get to pay cold hard cash for the opportunity to train MS interns certainly makes the OSS model of open dialog a little more appealing. Sure, a few random individuals sneak in the door that don't belong, but if the choices are out in the cold, or in a comfy tent with an obnoxious neighbor, I'll take the tent every time.
C'mon. What do you need a GeForce 6800 for on Linux? Admit it... You've got Doom3 on a windows partition don't ya!
3M actually approved this non business related use of company resources
Um... you just made the parents point. Closed source is authoritative and requires "permission" to "make cool hacks". OSS is the opposite.
This authoritarianism is an impediment to innovation. Why should anyone "need permission" to hack on a piece of software they have aquired a legal copy of?
I just wish there were someone else I could share the joke with...
Considering the length of the article, "I speak therefore I am" does seem like a good working title.
doublethink? Isn't understanding that term grounds for imprisonment under the DMCA or something?
I think the economics of this are more complicated
than they seem at first glance.
Ya, it's called monopolies on broadband provision. (Which is only natural for a utility, the problem is you're stuck with the utility provider and anything they want to force you into.)
It will be done in the name of "security" and "fighting viruses".
Guess I need to bone up on my trusted computing virus writing skills then... Root chains have inherent weaknesses completely aside from software/hardware implementation. I should think having the right couple keys would make it easier, rather than harder, to exploit.
However what do you expect to do if every broadband ISP announces such terms pretty much simultaneuously?
Ya, I worry about that one a lot. Being a comcast user, I'm already stuck with an ISP I can't stand. I'd pay a significant premium to be able to use someone else instead.
I thought I got AOL to do it when they bought Winamp...
We'll see how this stuff holds up in the market.
If my ISP suddenly required me to have a "trusted computer" to connect to the internet, I'd drop them like a hot potato. Ditto for all the rest of these "trusted" devices, gadgets and services.
What did he time it with? The cheater!
actually, considering that e is an infinite sequence and all, I suspect there's better than even odds that it does occur somewhere in there...
Especially if you buy into the whole "all data of civilization is stored in pi/e/<random possibly infinite number here>" claptrap. I mean come on. he posted it to slashdot, it better be in there if all the knowledge of the human race is in there.
That's true as far as it goes, but paypal isn't going to be able to levy a very big fine (at least against me) just for paying $10 one month for a pr0n subscription. Whether the fines stick on the other end is a different matter. (The article seemed to imply both buyers and sellers would be fined, which doesn't seem fully feasible)
I'm seriously considering canceling my account.
...most often due to husbands claiming the charge is fraudulent when the wife discovers it...
Bah! Moronic husbands. They should never share their paypal account details with their wives in the first place. *sigh*
Paypal doesn't have any of my money in their accounts, it's all in my bank and credit card accounts until I actually order something...
Well, that was for reference--the rest of my post is the wee bit shorter version. :-) To be fair to Rowe, though, it's about as short as philosophical tracts get...just over 9 pages in my text.
My bad, I thought you were refering to the entire text.
You've got me just right, except that my argument is yet more sweeping than just addressing Anselm's definition of god. This is my semantic argument against god, and it applies to all such arguments for or against the existence of such a being.
This is all fine and dandy, and prefectly rational, but it doesn't gain you any persuasive traction, it merely moves you almost immediately to the "agree to disagree" point where meaningful discussion ends. Being more sweeping in this matter just makes you more disagreeable... (Though hopefully agreeably so. Err, i think.)
If you want me to put forward definitions of God and soul for you I can certainly try, but you're as free to deny them credibility as I am to deny realism as a basic assumption. (Also, you're the one who brought "soul" up, I never mentioned it... It is a word I prefer not to use because it is so non-specific and often abused.)
I understand your woodchuck argument. However, I must point out that the possibility to do something must exist before that something can actually be done. The existence of the form provides the possibility, while imperfect implementations of that form that exist in reality attempt to meet that full potential...but always fall short to some extent.
I guess a question would be can anything "real" ever avoid falling short of "ideal". I'm not convinced that question has been answered in the positive or negative, and I certainly am not going to accept it one way or the other just based on considering "great" as Anselm used it to be equivalent to "ideal" as Plato used it.
If it is possible that something "real" can meet "ideal standards" then it is possible that God exists (ie: the concept of God exists/is rational/what have you), and Anselm's argument holds, at least in this regard.
Note: If pressed, I'd argue that any Platonic Ideal is every bit as nebulous as God or soul, at least to my way of thinking... (Can't be nailed down, can be different to different folks etc.)
This may be true for the referent line of thought, but there are other lines of thought that posed serious problems for Anselm shortly after the publication of his argument. For instance, Gaunilo argued that Anselm's argument could be applied to all sorts of things that we know do not exist...simply replace "god" with "island," and we can prove that the island than which none greater is possible exists in reality.
This is a great argument, although in my opinion it goes too far. You see, according to Anselm, the concept "God" contains, in definition, this notion of "greatest" as a necessary component. "island" does not have such a connection with the concept of "greatest" (we can conceive of the greatest island, but there is no requirement for island to be tied to greatest). That said, the only way this argument can defeat Anselm is by attacking his definition of God...
So, you've got 3 successful attacks on Anselm's definition of God, but all more or less indirect. (meaning you can't just say: "No Anselm's got it wrong, God might exist and is this other way..." Therefore someone convinced God *might* exist and persuaded by Anselm's language is still going to be unsatisfied. As many were.)
I think this is where we begin to disagree... I will say this, though: I do think that everyone, no matter what the basis for their worldview, that set of ideas/beliefs is ultimately based on some set of fundamental, axiomatic beliefs from which their understanding and personal philosophy derives. However, everyone's set of axiomatic beliefs are not created equal...some are demonstrably better than others. One person's set of axiomatic beliefs may result in a system of thought that i
...a firewall of some-sort.
If there's nothing stopping random incoming ports, to Windows boxes in particular, you're SOL no matter what else you've got going. These days viruses can saturate a population within a couple days, much faster than virus definitions can be distributed.
The point is that it is still interesting and worth debate, and has been for 900 years...
For a good argument that addresses the most popular of these points, read William L. Rowe's The Ontological Argument from Reason & Responsibility, Readings in Some Basic Problems of Philosophy, 8th Ed., by Joel Feinberg.
Um... Got anything just a wee bit shorter? Or maybe some useful excerpts?
So, that's the first problem. I would argue that Anselm's assumption that the concept of god exists is incorrect.
Okay, so you've just argued that the notion of "God", or in this case Anselm's "That than which nothing greater can be conceived" is incoherent. In doing so you've decided to disagree at the level of presupposition, and can now never have any meaningful argument with someone who considers the notion of "God" (and/or "that than which...") coherent, or even possibly so. Have fun.
I tend to go with Plato on this one, and not just in opposition to Anselm's argument...
Okay, that's fine. I happen to go with Sir Whilhelm Henry of Okham (or Occam whatever) and avoid multiplying entities unnecessarily. Of course, that has little to do with what Anselm is arguing about...
The problem in this case is that God is essentially a giant "do gooder" and a "do gooder" that doesn't "do good" isn't much "good". How much good can a do gooder do if a do gooder doesn't exist? You might argue that "good" is also an incoherent concept. That's fine, but you'd be missing my point. Try "woodchuck" instead of "do gooder" if that helps. Note: The woodchuck that exists is more woodchuck than any ideal, the woodchuck that doesn't exist is more ideal than any woodchuck, but a woodchuck that doesn't exist is only an ideal, not really a woodchuck at all.
Anselm argues that, no, there is one yet greater thing possible--that the concept of god itself exists. But this assumes that such an existence is possible...perhaps existence in our understanding is the greatest possible thing because to exist in reality would be not be possible.
I'm not sure I follow your whole argument in #3... Perhaps you're merely agreeing with what I said originally? To properly conceive of Anselm's notion of "God" one must conceive of him/her as existing. This is true WHETHER OR NOT "God exists" "in reality".
It wasn't until Kant that this line of thought was well-established... a bit more than 6 months.
Yes, we all make tiny leaps of faith every day for practical reasons, just to get things done.
...
You are arguing here, it seems, that since we often make leaps of faith, some correct, some not, that making leaps of faith is generally an acceptable practice in determining what to believe.
I'm just pointing out that leaps of faith are not always bad. In fact they are necessary as a basis for knowledge and understanding.
Here's an article apologetic to the use of methodological naturalism in science. The author makes the following statement: "[Methodological Naturalism] is what science employs, the belief that natural events have natural causes and that the physical world is logical and understandable." He goes on to make another statement: "Science itself, which uses methodological and not metaphysical naturalism, assumes that all events it can observe and study are natural in origin."
This seems a rather large and singular leap in a world of Heisenberg's uncertainty, chaos theory and quantum fluctuation (not to mention free-will).
If a leap of faith in this case is justified for some other reason, I am at a loss to see it. Can you tell me?
Objective knowledge requires leaps of faith, is it so strange to think subjective knowledge (or spirituality whatever your preference) also requires leaps of faith?
No. Go to your room.
Now if you'd put spelling/grammer, you might've gotten away wit it.
It would, indeed, be a good world where everybody would read everything regardless of style or content...but that's not going to happen.
Shyah! Why should people waste their time working and spending time with friends and family when they could be reading every single hair-brained post on slashdot... and every poorly written weblog... and every badly written book... etc...
...and the question won't earn you credit either.
Here's Google's thoughts on the meaning.
Anselm's argument for the existence of god, in the philosophy course I took (and every one I've ever heard of), is immediately followed by several arguments to the contrary that poke one giant hole after another in it.
I was aware that Anselm's argument was quite often side-stepped and berated but I wasn't aware there were any holes in it. (Hence the whole lasting for 900 years thing.)
To me, Anselm winds up showing that to properly conceive of God you have to conceive of God as existing, because otherwise you're leaving out a very major and obvious piece of the puzzle before you even start. Whether this "conception" and reality have anything to do with each other or if "God" is even a coherent notion is entirely another matter. Nevertheless, Anselm's argument seems to lend perspective to a lot of theist vs atheist debate. (disagreements at the level of presupposition are pretty hard to work out...)
If you really believe that there can be a logical argument that proves the existence of god, then you are in opposition to nearly every great philosopher, professor, rabbi, priest, theological scholar, etc...they all say that a leap of faith is necessary for the big 3 religions (Islam, Judaism, and Christianity).
Yes. Of course, many would also argue that leaps of faith are required for normal day to day life, active science, and the like. The disagreements are about which leaps make sense and whose are smaller.
Who'da thunk all the servers actually have all the articles...
By the tone of your post I'm not sure if you don't understand or are bitter.
Perhaps a bit of both... (or maybe just jaded) I'd inferred you were including GUI development in reference to your comments, since that's what the article was about and I don't see how testing can be simplified that far for GUIs.
inputs are finite and outputs are finite...
By that logic all encryption is "breakable" because it requires finite computations to break it. In a sense this is true, but that sense isn't very practical.
It is quite trivial to create a system in which the number of possible inputs is much too large to feasibly test. Theoretically, 128 bits of meaningful data is plenty to ensure no exhaustive proof is feasible in the near term. More practically GUI inputs including timing, mouse position, button clicking and general mayhem provides much larger sets of possible input.
Your point... "The only way you can "prove" your program is correct is by mangling the specification one way or another" is absolutely false...
That depends on your definition of specification. I was thinking along the lines of what is requested/desired by a customer, in the form they choose to request it in. You seem to have already limited specification to mean only "a precise and testable set of inputs".
I suppose I was thinking mostly of custom software , since that is the area I work in, but even a specification drawn up based on case studies, user surveys and the like can contain elements that don't break down well into small testable pieces.
The simple point is that certain types of specification can be validated to one degree or another (statistically if not exhaustively), but fulfilling the user's desires and intents, not meeting a specification drawn up by consultants and/or developers, is the primary goal.
Using proofs based on specifications is a bit like using test-driven development. The development process can become more controlled and predictable, but when you step back and look at overall project success rate and customer satisfaction, the numbers aren't much better than any of the other methodologies. This is because the rift between end-user and satisfaction exists just the same whether it is between user requests and developer understanding of the problem or between developer understanding and implementation. Decreasing the rift between developer understanding and implementation can be very helpful, but tends to increase overall cost and limit developer flexibility (ie: we can't do X because it isn't easily testable) therefore decreasing overall customer satisfaction.
(Note: All I really know about test driven development is what I've read. In trying it, I've read through a few examples, but they always leave a bad taste in my mouth... High level acceptance testing seems to have much more aesthetic appeal. )
As a counter-example to proof/test driven design I'll hold up the tiniest known C implementation of the CSS descrambler at the gallery of CSS descramblers. This is, in my opinion, an awesome piece of code. I don't see how anything close (in terms of size and speed) could be written using TDD or some other proof driven form of development. (obviously some level of testing is possible on this piece of code, but there's certainly no easy way I can think of to verify it as "100% correct" or "contains no bugs") Note: if you're going to read the code, you'll want to use a de-obfuscating tool of some sort to fix the lack of white space...