shouldn't we also have usage-based pricing for the TV they sell us? So that we pay "fairly" for for the fixed cost of establishing the network? Why would that model be different, since it's not really about congestion, as admitted in the article?
The electric bill and buffet examples in the article are terrible: when we pay for electric usage, we actually are paying for utlization/generation; use more and something (coal, natural gas, etc.) actually gets consumed more. And most buffets are all-you-can-eat; if you're paying by weight or something, the analogy is the same — you're actually consuming something. But both bandwidth and TV channels are there no matter how much they're "consumed." Bandwidth can be saturated (the congestion problem) but it can't be actually consumed.
If we're going to talk about "fairness", let's talk about:
1. Fair access to the wired networks built out, frequently, under monopoly guarantees
2. Fair levels of monetization of the network: does the telecom industry really want the equivalent of a utilities commission deciding how much they profit?
But the objection still holds: you haven't told us what consciousness is, just that it's not "something else." Fine. What about the human cognitive processes working produces whatever-it-is that we experience as introspection, and why do we experience it the way we do?
Until someone can answer that question, functionalism has written a check it can't cash. Assigning consciousness to some not-yet-understood aspect of cognitive processes is in no important respect different than simply declaring consciousness "mysterious."
There's no explanation on offer for how processes such as the ones we're currently able to describe produce phenomena like consciousness -- so either there's a "something else," or there's "something else" about the processes that we don't yet know.
You can take your mystery in whatever box you like, but the point is that consciousness is not yet anywhere near being well-understood, even in principle.
The consciousness "debate" will never be settled (at rather, widely agreed upon), because the answer just doesn't mesh intuitively with human introspection.
And, put non-technically, that's kind of a problem for functionalism.
We need either (a) an explanation of how some activity describable in functionalist terms can account for the experience of introspection or (b) an explanation of why introspection is unimportant to the relevant definition of consciousness.
Since case (b) seems unlikely, and case (a) is a blank check drawn on an unknown account, the pronouncements of experts giving us a timeframe seem fairly unreliable.
And this opens up an interesting possibility to provide an incentive to properly mark packets for QoS (on the user's end).
If Comcast added to this the following wrinkle: packets marked for "bulk" QoS automatically get assigned to the BE traffic level, but do not count against the consumption metric used to prioritize the rest of your traffic.
Presto: anyone who both torrents and watches streamed video now has an incentive to use a torrent app that marks its traffic as bulk.
Additionally, maybe streaming providers start getting sophisticated about how they deliver their streamed video to try to make part of that bulk as well, or at least do so adaptively when congestion is low. And downloads to iTunes, etc., easily go bulk.
Because we didn't know why this problem kept appearing at 40 minutes, we decided to set a timer. After 40 minutes, we would stop the car and reboot the computer to restore the performance.
So they knew that around 40 minutes, the performance would *gradually* degrade, and they fixed it with a timer that rebooted their system.
People! Of course it's a memory issue -- not a leak, as everybody else has explained, but of course the issue is that your system got memory-poor. Of course. Nobody every checked the Windows task manager to say "Gosh, no more memory. Maybe that hurts us?" You don't need a profiler to figure out what this problem was; it probably helped to figure out where the problem was.
The bigger question is, who sets a reset timer and considers their problem fixed, and, even if you don't have the timeline to find and fix the problem, who times the reset to the magic number of 40 minutes and not to the degrading performance?
The truly amazing thing is that people bright enough to build the rest of this system could be so remarkably clueless about basic debugging.
Exactly. The problem with this one isn't prior art, the problem is that this is obvious. Patent examiners need to be able to catch and weed out the obvious patents efficiently enough that it becomes pointless for them to be submitted. Otherwise we get submissions for every obvious software idea under the sun, and no one needs to be swamped by searching out prior art. We just need to be able say "obvious to a practicioner of the art."
The analogy is to the levy on blank CDs? The tax is because the downloaded music can be burned to a blank CD on which you've already paid a tax because you might burn music to it?
In related news, Canada proposed a tax on blank paper, by analogy to the sales tax which applies to books. "Someone might read what's written on the paper someday, and we won't then have the opportunity to collect the tax."
Ah. Thanks for the clarification. Turns out to be the same problem, even though a different company is behind it (Netscape instead of MS). It's still means that you're emulating an implementation instead of implementing a standard.
{VB, VBA, C++} : a superset of a subset of Fortran
Hear me out: remember from math that the null set, {}, is a subset of any set -- so if you're just supersetting the subset that is the null set, you can do whatever you want.
And it's instructive that MS thinks so.
Seriously, parent post is spot on -- MS's allegiance (and profit motive) is for the implementation, not the standard. And if they need the format to do something funky to enable the next wave of you-can't-live-without-it super-collaboration mumbo-jumbo to justify major $$ outlay for Office v. 23, they'll do it and never look back.
6,000 is a lot of pages to master, but it should be freely available for others to interpret, correct? On the other hand, since it is "essence of Microsoft", there's probably lots to misstep with and lots to nuance for interpretation letting Microsoft essentially maintain a proprietary flavor of a supposedly open standard.
The problem is, if we know anything about Microsoft, even if they're doing it with otherwise decent intentions, they're writing Office-the-software first and Office-the-standard second -- and therefore, there's a significant risk that the standard will always lag the implementation, and since their installed base is so big, the implementation will just win over the standard.
Exactly what was happening on the web for a while when IE's implementation of HTML/CSS could trump the standard to the degree that other vendors had to encode "quirks modes" into their own implementations to deal with people who wrote to the implementation rather than the standard. . .
And I would feel differently about this if it weren't for the fact that MS is bolting an XML format onto an existing product, which means that reverse-compatibility decisions are likely going to be determinative in the engineering.
So it's not the 6,000 pages -- it's the internal memos interpreting the 6,000 pages that we never get to see that are the problem.
Net loss of 1 H2O molecule in the Krebs Cycle. And plenty of other places as well, I assume.
It's impossible, one presumes, for any standard cellular organism to destroy all water in its environment, because then no biochemical processes could occur and it would be dead.
I presume the way this works is that they isolate the protein, rather than adding the organism to the water. And proteins don't self-replicate.
Well, they're mainstream -- but top 3? Or even top 10?
They're Chicago's equivalent of the NY Post or Daily News; the tabloid-ish irreverent paper, though my impression (I'm new to Chicago) is that some of their investigative stuff is stronger than the NY equivalents.
But, let's see: NYTimes, LATimes, Washington Post, Wall Street Journal, Chicago Tribune, USA Today, etc. . . . I think you have to work down the line to get to the Sun-Times.
And if you check Wikipedia or a quick Google for daily newspaper circulation, they're not even in the top 100. (Which I admit seems a bit off to me -- you'd think they'd beat out Springfield, MA or Allentown, PA, no matter what. But maybe they're some trick to what got included in the list.)
But it seems pretty definite they're not top 10, and definitively not top 3.
Not a bug - you can't register the duplicates
on
Gmail Mis.delivered?
·
· Score: 3, Informative
I don't know what this story is talking about, but the dot-ignoring delivery has been publicly known from day 1, and I just conducted the experiment of attempting to register a whole bunch of stupidly dotted variations on my username, and Gmail wouldn't let me register them.
So apparently they're doing the smart thing, and not including the dots when they do a uniqueness test on new usernames.
Maybe once upon a time in the very beginning they didn't, but I don't think that's the case now.
Actually, I think what Cory was trying to convince them of is that they're being suckered into defending a business model they're not invested in -- namely, the Hollywood/RIAA model. (And, as a side note, that given the option between defending a dying business model and developing a new one, the new one is almost always the better bet.)
His point is that Microsoft, like Sony with VCR, has no incentive to make a less capable tool.
DRM should be seen, from Microsoft's perspective, as a Linux/free software incentive program: if you build deliberately crippled tooks, you give your users reason to walk away from them.
And Microsoft has (or should have) far more interest in retaining the userbase than it does in receiving micropayments every time somebody plays a song on a DRM'd system.
It also bears pointing out, of course, that there is a version of events in which DRM is a winner for Microsoft -- it's the version where we posit strict legal enforcement of restrictions on the right to create new digital technology and innovation is never allowed to outstrip DRM. Setting aside for the moment the moral arguments against that, Cory points out that history suggests that betting on the 1984 vision of DRM and computers is pretty long odds.
I'm not really sure what your final point is, but on the "software patents are evil" bandwagon, the badness here is that once we have the transparent xterm, I think making transparency a function of time and user input counts as "obvious" to a practicioner of the art, hence un-patentable.
I could possibly see a defense for patents related to alpha-transparency implementations, either in firmware or at the OS level -- possibly.
But the use of an extant feature in designing UIs I think should generally count as obvious.
And that's the real problem with SW patents -- we don't have a currently useful standard for the obviousness of software innovation, which leads to abominations like the 1-click patents. SW patents, with a correct obviousness standard, might not necessarily be evil -- though they would of course be a hell of a lot more rare.
We have begun to create, to learn to manipulate and to use as building blocks of new formal symbolic systems (languages), the graphic equivalents of phonemes and morphemes, words, syntax and grammar. If we can build civilizations on sounds, then we can build extensions to civilizations and things as yet undreamed with graphic symbols.
Text . . . the graphic equivalent of phonemes and morphemes, words, syntax, and grammar. Maybe I should try to sell somebody on this "text" revolution.
Right, but the way in which you want to use and reuse those components becomes the problem.
I'm sure that there still some way to go in creating more and more powerful reuseable components, but I think that we've already got a very considerable component set -- think, for instance, of the Java and.NET API libraries.
A lot of what's amenable to be componentized has been, in the APIs. A lot of what real programmers actually do consists of gluing those components -- network and GUI components, especially -- into efficiently usable forms. But a visual designer isn't a whole lot of help there, since all the interesting work is almost by definition novel and unique.
I can't remember ever recoding a text box or a socket from scratch -- but the code that decides how to display the text box or what data to shoot through the socket is almost always novel, so I can't just "wire" one of the components to the other one (or the cost in component-abstraction to do so is unreasonably high).
I think this analysis is in most ways spot-on. I would add just two things:
The space shuttle is a more difficult engineering problem than software (even within equivalent "part counts") not just because our "parts" are 100% reliable, but also because they can be fully and reliably modelled. Real-world, physical parts can only be modelled to certain degrees of tolerance, and have to interact with systems that have the same limitations. Also, while we have a vastly larger total number of "parts" in software, we have a much smaller number of distinct parts -- ultimately, only the primitive instructions of the Turing machine.
Visual design systems suffer from exactly the problem you've described: what I call the death of a thousand attributes. This is also where a lot of markup editors and interface design systems fall down -- you've got a visual representation of a thing (mostly) but in order to manipulate the behavior, you have to tweak tens or more of attributes that control the widget, and those tweaks can't be made or represented in the visual depiction. Often times, that translates to a net decrease in efficiency, since it forces you to use a visual metaphor to accomplish what is fundamentally a coding task.
Right, but hand your credit card to a waiter in a restaurant, he disappears into the back, pulls out a Chameleon and sucks down your card information into his Chameleon . . .
Slashdot Mirror
shouldn't we also have usage-based pricing for the TV they sell us? So that we pay "fairly" for for the fixed cost of establishing the network? Why would that model be different, since it's not really about congestion, as admitted in the article?
The electric bill and buffet examples in the article are terrible: when we pay for electric usage, we actually are paying for utlization/generation; use more and something (coal, natural gas, etc.) actually gets consumed more. And most buffets are all-you-can-eat; if you're paying by weight or something, the analogy is the same — you're actually consuming something. But both bandwidth and TV channels are there no matter how much they're "consumed." Bandwidth can be saturated (the congestion problem) but it can't be actually consumed.
If we're going to talk about "fairness", let's talk about:
But the objection still holds: you haven't told us what consciousness is, just that it's not "something else." Fine. What about the human cognitive processes working produces whatever-it-is that we experience as introspection, and why do we experience it the way we do?
Until someone can answer that question, functionalism has written a check it can't cash. Assigning consciousness to some not-yet-understood aspect of cognitive processes is in no important respect different than simply declaring consciousness "mysterious."
There's no explanation on offer for how processes such as the ones we're currently able to describe produce phenomena like consciousness -- so either there's a "something else," or there's "something else" about the processes that we don't yet know.
You can take your mystery in whatever box you like, but the point is that consciousness is not yet anywhere near being well-understood, even in principle.
The consciousness "debate" will never be settled (at rather, widely agreed upon), because the answer just doesn't mesh intuitively with human introspection.
And, put non-technically, that's kind of a problem for functionalism.
We need either (a) an explanation of how some activity describable in functionalist terms can account for the experience of introspection or (b) an explanation of why introspection is unimportant to the relevant definition of consciousness.
Since case (b) seems unlikely, and case (a) is a blank check drawn on an unknown account, the pronouncements of experts giving us a timeframe seem fairly unreliable.
If Comcast added to this the following wrinkle: packets marked for "bulk" QoS automatically get assigned to the BE traffic level, but do not count against the consumption metric used to prioritize the rest of your traffic.
Presto: anyone who both torrents and watches streamed video now has an incentive to use a torrent app that marks its traffic as bulk.
Additionally, maybe streaming providers start getting sophisticated about how they deliver their streamed video to try to make part of that bulk as well, or at least do so adaptively when congestion is low. And downloads to iTunes, etc., easily go bulk.
People! Of course it's a memory issue -- not a leak, as everybody else has explained, but of course the issue is that your system got memory-poor. Of course. Nobody every checked the Windows task manager to say "Gosh, no more memory. Maybe that hurts us?" You don't need a profiler to figure out what this problem was; it probably helped to figure out where the problem was.
The bigger question is, who sets a reset timer and considers their problem fixed, and, even if you don't have the timeline to find and fix the problem, who times the reset to the magic number of 40 minutes and not to the degrading performance?
The truly amazing thing is that people bright enough to build the rest of this system could be so remarkably clueless about basic debugging.
Exactly. The problem with this one isn't prior art, the problem is that this is obvious. Patent examiners need to be able to catch and weed out the obvious patents efficiently enough that it becomes pointless for them to be submitted. Otherwise we get submissions for every obvious software idea under the sun, and no one needs to be swamped by searching out prior art. We just need to be able say "obvious to a practicioner of the art."
In related news, Canada proposed a tax on blank paper, by analogy to the sales tax which applies to books. "Someone might read what's written on the paper someday, and we won't then have the opportunity to collect the tax."
see subject
Ah. Thanks for the clarification. Turns out to be the same problem, even though a different company is behind it (Netscape instead of MS). It's still means that you're emulating an implementation instead of implementing a standard.
Hear me out: remember from math that the null set, {}, is a subset of any set -- so if you're just supersetting the subset that is the null set, you can do whatever you want.
And it's instructive that MS thinks so.
Seriously, parent post is spot on -- MS's allegiance (and profit motive) is for the implementation, not the standard. And if they need the format to do something funky to enable the next wave of you-can't-live-without-it super-collaboration mumbo-jumbo to justify major $$ outlay for Office v. 23, they'll do it and never look back.
The problem is, if we know anything about Microsoft, even if they're doing it with otherwise decent intentions, they're writing Office-the-software first and Office-the-standard second -- and therefore, there's a significant risk that the standard will always lag the implementation, and since their installed base is so big, the implementation will just win over the standard.
Exactly what was happening on the web for a while when IE's implementation of HTML/CSS could trump the standard to the degree that other vendors had to encode "quirks modes" into their own implementations to deal with people who wrote to the implementation rather than the standard. . .
And I would feel differently about this if it weren't for the fact that MS is bolting an XML format onto an existing product, which means that reverse-compatibility decisions are likely going to be determinative in the engineering.
So it's not the 6,000 pages -- it's the internal memos interpreting the 6,000 pages that we never get to see that are the problem.
Net loss of 1 H2O molecule in the Krebs Cycle. And plenty of other places as well, I assume.
It's impossible, one presumes, for any standard cellular organism to destroy all water in its environment, because then no biochemical processes could occur and it would be dead.
I presume the way this works is that they isolate the protein, rather than adding the organism to the water. And proteins don't self-replicate.
They're Chicago's equivalent of the NY Post or Daily News; the tabloid-ish irreverent paper, though my impression (I'm new to Chicago) is that some of their investigative stuff is stronger than the NY equivalents.
But, let's see: NYTimes, LATimes, Washington Post, Wall Street Journal, Chicago Tribune, USA Today, etc. . . . I think you have to work down the line to get to the Sun-Times.
And if you check Wikipedia or a quick Google for daily newspaper circulation, they're not even in the top 100. (Which I admit seems a bit off to me -- you'd think they'd beat out Springfield, MA or Allentown, PA, no matter what. But maybe they're some trick to what got included in the list.)
But it seems pretty definite they're not top 10, and definitively not top 3.
I don't know what this story is talking about, but the dot-ignoring delivery has been publicly known from day 1, and I just conducted the experiment of attempting to register a whole bunch of stupidly dotted variations on my username, and Gmail wouldn't let me register them.
So apparently they're doing the smart thing, and not including the dots when they do a uniqueness test on new usernames.
Maybe once upon a time in the very beginning they didn't, but I don't think that's the case now.
traditional telephony companies need to adapt or die ... in order to remain viable.
Dying to remain viable. Gotta buy me some of that stock.
I was waiting for someone to catch that typo. My apologies to Pippin.
Actually, I think what Cory was trying to convince them of is that they're being suckered into defending a business model they're not invested in -- namely, the Hollywood/RIAA model. (And, as a side note, that given the option between defending a dying business model and developing a new one, the new one is almost always the better bet.)
His point is that Microsoft, like Sony with VCR, has no incentive to make a less capable tool.
DRM should be seen, from Microsoft's perspective, as a Linux/free software incentive program: if you build deliberately crippled tooks, you give your users reason to walk away from them.
And Microsoft has (or should have) far more interest in retaining the userbase than it does in
receiving micropayments every time somebody plays a song on a DRM'd system.
It also bears pointing out, of course, that there is a version of events in which DRM is a winner for Microsoft -- it's the version where we posit strict legal enforcement of restrictions on the right to create new digital technology and innovation is never allowed to outstrip DRM. Setting aside for the moment the moral arguments against that, Cory points out that history suggests that betting on the 1984 vision of DRM and computers is pretty long odds.
There's a way of looking at the evidence that says that Turing machine has all the "features" that could ever be patented in software.
I could possibly see a defense for patents related to alpha-transparency implementations, either in firmware or at the OS level -- possibly.
But the use of an extant feature in designing UIs I think should generally count as obvious.
And that's the real problem with SW patents -- we don't have a currently useful standard for the obviousness of software innovation, which leads to abominations like the 1-click patents. SW patents, with a correct obviousness standard, might not necessarily be evil -- though they would of course be a hell of a lot more rare.
That is, before it melted.
Text . . . the graphic equivalent of phonemes and morphemes, words, syntax, and grammar. Maybe I should try to sell somebody on this "text" revolution.
See subject.
I'm sure that there still some way to go in creating more and more powerful reuseable components, but I think that we've already got a very considerable component set -- think, for instance, of the Java and .NET API libraries.
A lot of what's amenable to be componentized has been, in the APIs. A lot of what real programmers actually do consists of gluing those components -- network and GUI components, especially -- into efficiently usable forms. But a visual designer isn't a whole lot of help there, since all the interesting work is almost by definition novel and unique.
I can't remember ever recoding a text box or a socket from scratch -- but the code that decides how to display the text box or what data to shoot through the socket is almost always novel, so I can't just "wire" one of the components to the other one (or the cost in component-abstraction to do so is unreasonably high).
I think this analysis is in most ways spot-on. I would add just two things:
Right, but hand your credit card to a waiter in a restaurant, he disappears into the back, pulls out a Chameleon and sucks down your card information into his Chameleon . . .