So now apparently Wikipedia is unreliable because it: 1) defers too much to experts; and also 2) doesn't defer enough to experts.
Maybe it's my dialectical nature that wants to find a bigger picture in everything, but I don't see any real inconsistency in this. Wikipedia's slipshod handwavey approach to the matter of expertise has resulted in the problems of both extremes. The glut of sloppy amateur editing results in an automatic deference to the rare appearance of presumed expertise, verifiable or not.
Tangentially, there's a reason professors need to publish: it puts them on record. So when you have real verified experts putting their imprimatur on articles, you don't get as much runaway bullshit (it happens, but it's generally news when it does).
> If nothing else, Objectivism got a lot of traction towards the eternal problems in epistemology.
"A = A" is hardly a statement loaded with meaning. Rand was a methadrine freak who claimed that if you liked Chopin over Rachmaninov, you were "denying reality".
In sum, if there's a lesson to be learned here, it's that credentials (whether real or fabricated) really are meaningless.
If you're ever in need of medical treatment, look me up -- I only charge half what those folks with fancy letters after their names charge. I can also make filings on your behalf if you're ever in legal trouble -- strictly paperwork mind you, since those credentialist bastards won't let me argue in court.
but unless I use a particularly fancy bit of JavaScript, they almost always functionally work the same in multiple browsers. I just don't get it... are the people who are writing the web apps really that bad with their concept of standards?
Perhaps the Javascript they write is, I dunno, particularly fancy? Once you handle events, for example, in your code beyond onClick and onMouseOver, you find they need to be handled differently. And a site that's "effed up" in IE isn't actually acceptable to most web developers who want to get paid. When you start creating boxes and moving them around with JS, that box model problem is magnified.
That said, cross-platform libraries exist that work across all major browsers, so indeed I've never sweated the difference.
(Sorry for the monospace, slashdot's formatter is exceedingly lame and wanted to munge it all onto one line.)
This doesn't address the base applications, or well, anything installed by the package managers of these systems, but when I install by hand (and that certainly does happen), I tend to organize like so
Packages themselves have strict organization, having bin/ lib/ man/ subdirectories. Login scripts scan through/opt/*/bin and add it to PATH. As for lib/, in some installations I either directly fiddle with the ld.so cache config directly, set up LD_LIBRARY_PATH (probably not a good idea), or I do another indirection by having the bin/ directory consist of a symlink to a wrapper that sets up LD_LIBRARY_PATH and calls the "real" executable in the exe/ subdirectory.
This scheme is similar to what Sun uses internally, called softdist, which is basically a NFS mount point that's mirrored company-wide. softdist also has wrapper scripts for auditing, launching the right exe for multiple architectures, post-deploy scripts, and so forth, but it's probably nothing you need to care about.
It's not a bad scheme, and the indirection has helped me roll back versions very painlessly when needed. It'd be nice for a package manager to support, but it'd probably be easier to make it work with something ports/portage/pkgsrc.
> Now that we've reached the limits of the Von Neumann architecture, we're starting to see a new wave of innovation in CPU design.
They're still Von Neumann, just parallelized. Program is still data, stepped through linearly (just in more independent parallel threads), results put into storage, and so forth. It's not some kind of eigenstate weirdness or third concept apart from code and data.
> But without the 2nd, we would not now have the first,
I'm not exactly a card-carrying member of HCI myself, but it'd sure be nice to see some tangible evidence of your claim. You do realize that the American Revolution wasn't conducted under the auspices of the not-yet-existent Constitution, let alone its 2nd Amendment?
> Where "ad hominem" is a logical fallacy is when I say "The sky is green," and you say "You're an idiot."
Actually, it's not an ad hominem argument, it's just an insult. If you conclude with "therefore the sky is not green", or attempt to connect it to any other fact (other than one that necessarily follows from idiocy, and the color of the sky certainly doesn't) then it's an ad hominem argument.
That's an excellent response. I don't agree with everything in the list (some items I outright contest), but at least you took more time to respond than my off the cuff post even really deserved.
WinFS will never see daylight, but as for Monad, it's out, it's called Powershell, and Exchange 2007's backend is basically scripted entirely in it. It's still a little klunky to use until there's a good set of short command aliases built up into a standard library though. Powershell Analyzer (a third party thing, google for it) is about eight million degrees of cool, just watch the video demos. Compared to the functional programming research output of MSR, Powershell (nee Monad-- it was named for a FP device, it doesn't use it) is not much, but as a shell it has no equal at this time.
> Any given statement is not proven correct or incorrect by who said it, but by the merits of the statement itse
I don't need to be educated on the definition of appeal to authority -- especially when the argument being considered is whether this Essjay has the necessary qualifications to be a trusted authority (in both senses of the term).
I confusedly switched the terms "informal" and "formal" in my previous post, though I wouldn't claim that ad logicam arguments rise to a formal fallacy unless they actually attempt to negate the conclusion as opposed to the usual device of mudslinging to support one's own argument. The Wikipedia page (irony alert) on ad logicam illustrates the obvious formal fallacy pretty well. Like most articles on logic, the examples are contrived, but I have a hard time thinking of any of that (formal) fallacy that aren't.
Formal/informal distinctions are irrelevant -- a fallacy in an opposing argument does nothing to support the soundness of one's own argument.
The only specific item in there (/proc) did not come from Linux, it's from Solaris. Linux can take credit for what its version of/proc became in the 2.4 series. No one else wants it, not even Linux 2.6.
And if Linux was first to 64 bit support, Microsoft invented the Internet.
We think that we are inserting some ranting into a post that was never intended to be your soapbox. We are not amused. We think we should increase our medication.
If education in the US provided proper building blocks and intellectual tools, instead of rote memorization of both true and false "factoids"
Diversity in the workplace has told me one thing: primary education in other countries, particularly former British colonies, is worse when it comes to rote memorization. I guarantee you that no one in India is getting taught critical thinking, logic, or rhetorical methods in their early education. In some areas, we may well have moved too far away from rote skills: it'd be nice to see the basic rules of grammar and spelling drilled in a touch more, neh?
Appealing to authority is a logical fallacy as any properly educated person knows.
And "appealing" to formal fallacy is itself an informal fallacy -- I defer to Einstein and Hawking when it comes to matters of physics. This kid on Wikipedia isn't an authority, just a schmuck who claimed to be one.
I can see why he did it, I think you can't blame him entirely. We have a whole irrational damn-near religious awe of credentials and enormous stigma against those who do not possess this "sacred currency"
I don't. But I do have this irrational attachment to the truth.
Thing is, I still go to wikipedia to look up info, it's become a reflex, just typing a noun appended by "wikipedia" in google. But I no longer feel good about it. Nor am I particularly inclined to help edit it when I can see that my efforts would simply be sabotaged from above by malignant indifference, blundering incompetence, and (increasingly now) outright mendacity.
Technically the Korean War never really ended, there's just been a 50+ year ceasefire. We tend to not draft people when there isn't any shooting going on though.
I may be going out on a limb here, but I think they stopped drafting people for the Korean war before 1970. Perhaps you've got it confused with M*A*S*H.
Slashdot Mirror
> It's going to change the world.
No it isn't. False credentials are still an exception, and human social behavior is still the norm.
So now apparently Wikipedia is unreliable because it: 1) defers too much to experts; and also 2) doesn't defer enough to experts.
Maybe it's my dialectical nature that wants to find a bigger picture in everything, but I don't see any real inconsistency in this. Wikipedia's slipshod handwavey approach to the matter of expertise has resulted in the problems of both extremes. The glut of sloppy amateur editing results in an automatic deference to the rare appearance of presumed expertise, verifiable or not.
Tangentially, there's a reason professors need to publish: it puts them on record. So when you have real verified experts putting their imprimatur on articles, you don't get as much runaway bullshit (it happens, but it's generally news when it does).
> If nothing else, Objectivism got a lot of traction towards the eternal problems in epistemology.
"A = A" is hardly a statement loaded with meaning. Rand was a methadrine freak who claimed that if you liked Chopin over Rachmaninov, you were "denying reality".
So, if Anne Frank is in your attic, and the SS is at your door, you'd have no problem being truthful about Anne's whereabouts?
I have Mike Godwin holding on line 1.
In sum, if there's a lesson to be learned here, it's that credentials (whether real or fabricated) really are meaningless.
If you're ever in need of medical treatment, look me up -- I only charge half what those folks with fancy letters after their names charge. I can also make filings on your behalf if you're ever in legal trouble -- strictly paperwork mind you, since those credentialist bastards won't let me argue in court.
but unless I use a particularly fancy bit of JavaScript, they almost always functionally work the same in multiple browsers. I just don't get it... are the people who are writing the web apps really that bad with their concept of standards?
Perhaps the Javascript they write is, I dunno, particularly fancy? Once you handle events, for example, in your code beyond onClick and onMouseOver, you find they need to be handled differently. And a site that's "effed up" in IE isn't actually acceptable to most web developers who want to get paid. When you start creating boxes and moving them around with JS, that box model problem is magnified.
That said, cross-platform libraries exist that work across all major browsers, so indeed I've never sweated the difference.
Done. http://www.caucho.com/resin-3.1/doc/quercus.xtp
> Anyone ever see how many symlinks they can nest.
Quoth POSIX: eight. That's the value of SYMLOOP_MAX.
(Sorry for the monospace, slashdot's formatter is exceedingly lame and wanted to munge it all onto one line.)
0 609082 -> /u01/oracle
../pkg/foo-1.5.7 ../pkg/bar-20060908 ../pkg/baz-1.5abeta7
/pkg
/opt/*/bin and add it to PATH. As for lib/, in some installations I either directly fiddle with the ld.so cache config directly, set up LD_LIBRARY_PATH (probably not a good idea), or I do another indirection by having the bin/ directory consist of a symlink to a wrapper that sets up LD_LIBRARY_PATH and calls the "real" executable in the exe/ subdirectory.
This doesn't address the base applications, or well, anything installed by the package managers of these systems, but when I install by hand (and that certainly does happen), I tend to organize like so
/pkg/foo-1.2.3
/pkg/foo-1.5.7
/pkg/bar-20
/pkg/baz-1.5abeta7
/pkg/oracle-9i-2006091
you get the idea, name and whatever version convention the package chooses to use. Big things might symlink elsewhere.
/opt/foo ->
/opt/bar ->
/opt/baz ->
that is, symlinks into
Packages themselves have strict organization, having bin/ lib/ man/ subdirectories. Login scripts scan through
This scheme is similar to what Sun uses internally, called softdist, which is basically a NFS mount point that's mirrored company-wide. softdist also has wrapper scripts for auditing, launching the right exe for multiple architectures, post-deploy scripts, and so forth, but it's probably nothing you need to care about.
It's not a bad scheme, and the indirection has helped me roll back versions very painlessly when needed. It'd be nice for a package manager to support, but it'd probably be easier to make it work with something ports/portage/pkgsrc.
Look for the column of vodka hanging in mid-air.
> I've seen cats pull stunts that I'm pretty sure your average MySpace poster couldn't pull off
I'm not entirely certain that the average MySpace poster could pass the Turing Test either.
> Now that we've reached the limits of the Von Neumann architecture, we're starting to see a new wave of innovation in CPU design.
They're still Von Neumann, just parallelized. Program is still data, stepped through linearly (just in more independent parallel threads), results put into storage, and so forth. It's not some kind of eigenstate weirdness or third concept apart from code and data.
> But without the 2nd, we would not now have the first,
I'm not exactly a card-carrying member of HCI myself, but it'd sure be nice to see some tangible evidence of your claim. You do realize that the American Revolution wasn't conducted under the auspices of the not-yet-existent Constitution, let alone its 2nd Amendment?
Whatever happened to Chip Salzenberg? He seems to have pretty much vanished since mid-2006.
> Where "ad hominem" is a logical fallacy is when I say "The sky is green," and you say "You're an idiot."
Actually, it's not an ad hominem argument, it's just an insult. If you conclude with "therefore the sky is not green", or attempt to connect it to any other fact (other than one that necessarily follows from idiocy, and the color of the sky certainly doesn't) then it's an ad hominem argument.
That's an excellent response. I don't agree with everything in the list (some items I outright contest), but at least you took more time to respond than my off the cuff post even really deserved.
WinFS will never see daylight, but as for Monad, it's out, it's called Powershell, and Exchange 2007's backend is basically scripted entirely in it. It's still a little klunky to use until there's a good set of short command aliases built up into a standard library though. Powershell Analyzer (a third party thing, google for it) is about eight million degrees of cool, just watch the video demos. Compared to the functional programming research output of MSR, Powershell (nee Monad-- it was named for a FP device, it doesn't use it) is not much, but as a shell it has no equal at this time.
> Any given statement is not proven correct or incorrect by who said it, but by the merits of the statement itse
I don't need to be educated on the definition of appeal to authority -- especially when the argument being considered is whether this Essjay has the necessary qualifications to be a trusted authority (in both senses of the term).
I confusedly switched the terms "informal" and "formal" in my previous post, though I wouldn't claim that ad logicam arguments rise to a formal fallacy unless they actually attempt to negate the conclusion as opposed to the usual device of mudslinging to support one's own argument. The Wikipedia page (irony alert) on ad logicam illustrates the obvious formal fallacy pretty well. Like most articles on logic, the examples are contrived, but I have a hard time thinking of any of that (formal) fallacy that aren't.
Formal/informal distinctions are irrelevant -- a fallacy in an opposing argument does nothing to support the soundness of one's own argument.
The only specific item in there (/proc) did not come from Linux, it's from Solaris. Linux can take credit for what its version of /proc became in the 2.4 series. No one else wants it, not even Linux 2.6.
And if Linux was first to 64 bit support, Microsoft invented the Internet.
> Microsoft hasn't been innovating for years
Microsoft Research innovates like crazy. It's just rare that anything ever escapes alive and in recognizable form from MSR.
Hell, what has Linux innovated lately? Desktops on spinning cubes?
> Any scandal that erupts out of this is largely Ad Hominem.
Did you get that from the Wikipedia article on logic? ad hominem is not Latin for "wrong".
We think that we are inserting some ranting into a post that was never intended to be your soapbox. We are not amused. We think we should increase our medication.
If education in the US provided proper building blocks and intellectual tools, instead of rote memorization of both true and false "factoids"
Diversity in the workplace has told me one thing: primary education in other countries, particularly former British colonies, is worse when it comes to rote memorization. I guarantee you that no one in India is getting taught critical thinking, logic, or rhetorical methods in their early education. In some areas, we may well have moved too far away from rote skills: it'd be nice to see the basic rules of grammar and spelling drilled in a touch more, neh?
Appealing to authority is a logical fallacy as any properly educated person knows.
And "appealing" to formal fallacy is itself an informal fallacy -- I defer to Einstein and Hawking when it comes to matters of physics. This kid on Wikipedia isn't an authority, just a schmuck who claimed to be one.
I can see why he did it, I think you can't blame him entirely. We have a whole irrational damn-near religious awe of credentials and enormous stigma against those who do not possess this "sacred currency"
I don't. But I do have this irrational attachment to the truth.
Thing is, I still go to wikipedia to look up info, it's become a reflex, just typing a noun appended by "wikipedia" in google. But I no longer feel good about it. Nor am I particularly inclined to help edit it when I can see that my efforts would simply be sabotaged from above by malignant indifference, blundering incompetence, and (increasingly now) outright mendacity.
Technically the Korean War never really ended, there's just been a 50+ year ceasefire. We tend to not draft people when there isn't any shooting going on though.
I may be going out on a limb here, but I think they stopped drafting people for the Korean war before 1970. Perhaps you've got it confused with M*A*S*H.