Re:sure it's arbitrary
on
Happy Pi Day
·
· Score: 1
Yeah, arbitrarily chosen natural phenomena. I don't know about you, but my sleeping patterns seem to relate closely to the earth's rotation around itself. Also, the rotation of the earth around the sun tends to influence my eating habits (what's in season?) and clothing habits (coat or no coat).
If indeed they're arbitrarily chosen, what are some other natural phenomena suitable for calendaring that affect my life?
If you read one of the linked articles, you'll learn that
Unlike other trials that used 10 separate 10 Gbps wavelengths to carry 100 Gbps, the Verizon test utilized a 100 Gbps signal on a single wavelength. Other interesting things from that article:
"This is another critical milestone on our way to ensuring that we have the most advanced telecom network technology at the right time, in the right place, to serve our customers," Having never used Comcast, I'll leave it to those who have to asses the meaning of "serve our customers".
"The field trial also confirms optical networking's role as a critical enabler for competitive transformation, as new services and applications reshape network requirements." I thought peer-to-peer data transfer already "reshaped network requirements" and left the ISPs struggling to keep up.
"Applications based on online video are clearly drivers for higher bandwidth [...]" Again, how about peer-to-peer transfer.
"Transmission at 100 Gbps will enable low latency and significant improvement in real-time transaction. Trading institutions and other Verizon customers using real-time communications will find the associated performance very attractive." I can transfer 100 Gbps by putting hard disks in my backpack and running a short sprint. This is orthogonal to latency (which is what real-time is about). Okay, strawman. Assume the big fiber is deployed, and everybody uses it. Then you'll run into contention issues, and your packet will sit in a queue. I'm not saying bigger pipes won't help, but I want an arguement; right now, all I have is a claim.
But no system of representing numbers could express pi's relationship to 1 exactly without an infinite amount of information. Have a look at Computable Numbers. Brief version: a number x is computable if there exists a Turing machine (basically a program, should a reader be unaware) that when given a rational epsilon > 0 computes a rational r, such that |x - r| < epsilon.
From the WP article:
The computable numbers include many of the specific real numbers which appear in practice, including all algebraic numbers, as well as e, pi, and many other transcendental numbers. If you start from ZFC you typically construct the real numbers as Cauchy sequences of rationals, each sequence representing its limit. So, let's pick the constant sequence of rational zeroes to represent the real zero. How do you represent this with a finite amount of information? Is it different then, to represent numbers by Turing machines? If so, how?
You don't define what you mean by "exactly" or "relationship", so I can't speak to that very well. How do you express 7's "relationship" to 142857 "exactly"?
Re:sure it's arbitrary
on
Happy Pi Day
·
· Score: 1
It's arbitrary insofar as [...] True, we could use a different calendar. You do get that what your parent tries to point out is that the duration of the time intervals in our current time-and-calendar system aren't arbitrary, but derived from natural phenomena, right?
Seriously, don't get the cynical mathematicians on/. going about hyperbole like "exponential rates". Assume that transistor count (not CPU speed!) grows as a^n and source code volume grows as b^n. Then each transistor will have to run (b/a)^n units of code. OMG, transistor code load grows at an exponential rate!!!multiplicativeidentity!!!fifthprime!!1!11!booleannegation
you shouldn't be running anything but compiled source on a performance cluster. you wouldn't be running anything but source compiles on a performance cluster.:p
A couple of critical observations:
So you're starting to see OEMs pre-ship Linux for the first time [...] why are they doing that? [...] Is it because Linux is more functional than it's ever been? [...] yes, it is more functional. But that functionality combined with the economics [...]. [lots of...] And so when companies like Dell or Asus or Lenovo or all these companies look at those profit margins, they say, "Why wouldn't I just create my own operating system and ship it with the device? Interesting. He argues for linux based primarily on price, and from the seller's viewpoint. Sure, some of the savings is (presumably) passed on to the consumer, but I miss a good argument for why the consumer should use it. He could have said "it does the same job for a lower price", which is a very convincing argument in my book (and to some extent also valid). It's not a very sexy argument, though, but I won't be demanding everything.
InfoWorld: But Windows is still on 98, 99 per cent of PC desktops anyway, so do you think that number or that percentage will decrease? Zemlin: Yes. Yes, I think it will actually. As an extention of my previous comment, I wished InfoWorld would have asked him why he thinks that. I can say "I think I will get -1 Flamebait"; so what? It only becomes interesting when I have something to substantiate it with.
If you're a Motorola or an LG, would you rather, per device basis when you're selling tens of millions of devices, license Windows Mobile or the Symbian platform from Nokia, or would you rather have Linux, which is collaboratively designed, which supports every major architecture? He's saying "Would you rather have a, or b, or c? Note that c has properties p and q." For a meaningful comparison, he could list some properties of a and b as well. Sure, we may know some of them, but he could have emphasised what he thinks is relevant. Also, when you've picked an architecture, support for any other architecture is not of much interest.
InfoWorld: So are Microsoft's days as the dominant provider of desktop and server and maybe even handheld operating systems numbered? Zemlin: Monopolies don't last forever, so I mean, I think they've got a long way to go. It's just natural over time that people aren't going to allow a single company to dominate the market. People, as in individual consumers, always allow concentration of power; they don't care, they just want easy. Look at Microsoft; look at Google. People use their services because it's easy for them to do so. I'm conjecturing that it takes the power of the state to break up or prevent the formation of monopolies. Someone enlighten me: what happened with AT&T, Standard Oil and the Railroads? What's happening with Microsoft?
InfoWorld: Wouldn't the emergence of Linux kind of say that maybe Microsoft never really was a monopoly, that there was always room for somebody else to compete in there and that's what Linux is now doing? Zemlin: It obviously was a desktop monopoly for a period of time. Again, there's a claim but no argument. I don't want to argue against him, I just want a better argument.
Yes, the days of high-margin, vendor lock-in monopoly practices in the software business, yes, those are gone, and they're permanently gone. This is five minutes after he stated that Microsoft makes 30% profit; that doesn't like either dead or dying to me (sadly).
InfoWorld: Can Solaris compete with Linux? Zemlin: [If Solaris was FLOSS eight years ago maybe, blah blah...] Sure, that sounds plausible. What interests me is why Solaris (probably) won't be able to compete with Linux today. I'm guessing it's the network effects: with many users, it makes sense to develop for, and with much development, it makes sense to use.
(None of this is to say that I disagree with Zemlin's assertions, just that I think he could have made a better argument).
All we need is for there to be a Chinese-Scientology link and we'll have the most commented-on story ever. I for one (captcha: wisely) welcome our new Chinese DRM-loving Scientologist lawyer-wielding terrorism-inclined Vista-running meme-loving robot overlords. Netcraft says they're not eligible for being confirmed dead, as no robot was ever alive.
I didn't say anything about avoiding columns. I said they weren't a high priority. Okay, so how do you suggest I maximize the amount of text on the screen at any one time, while at the same time keeping lines fairly short?
I could switch to 640x480, but that'd be a net loss: yeah, I'd get shorter lines, but I could get the same just by wrapping lines at 66 characters at my current resolution. That's be about as efficient (in terms of screen real estate use) as limiting pages to a single $n-pixel column: not efficient at all.
Also, I think you're trying to argue against my suggestion for multi-column pages, but I don't see how your arguments work. Yes, there may be some instances (like you've pointed out) where my solution doesn't have a problem to solve. Great. What about the other instances, are there anything wrong with my idea for those?
Microsoft has absolutely no interest in package management, unless you count MSI as such. Why? Because too many software vendors are more than happy enough to kiss Microsoft's ass and produce physical media for the vast majority of its software, making package management redundant and pointless.
Package management and physical media are orthogonal.
In particular, the CD from which I installed Ubuntu has tons of packages on it. You can have physical media without packages (say, Wii Sports), packages without physical media (that's what all the mirrors in my/etc/apt/sources.list are for) and no packages on no physical media: that's what your typical windows app is (winamp, winzip, winorlose).
Making a game is expensive, so logically you want to release it to the biggest audience you can so that you can reap the most profit (or at the very least make enough to hit the break-even point). Windows, with 85% of the OS market, has the most promise of giving you the highest audience in PC gaming.
No, not true.
You say that making a game is expensive, and that's true; you also state that one wants to maximize profit, with which I can agree (you may want to maximize its first derivative, but let's assume not). You then claim that one gets the most profit by reaching the most users.
How about shelling out $1 supporting Imaginary OS which has one user, and that user won't pay any money for your game? I'd say that's a net loss. Maximizing profit requires a cost-benefit analysis, not just a benefit analysis.
Dude, you have no idea what kind of dangerous weapons of mass destruction you can disguise as breast milk, hand cream and Starbucks coffee. Terrorists are using them to kidnap and molest your children. Won't somebody think of the children?? With this new scanner, we can protect and--God forbid we will have to--save your children. Why do you hate America's children, you crazy hippie?
(Committee for Aviation Transportation Security says "All your privacy are belong to us")
It took THREE VERSIONS to come up with a layout idea that's been used in newspapers for books for literally centuries?!
Web pages have infinite vertical space. Newspapers and books don't. Horizontal space is at a premium for web pages. It's not as important for newspapers and books. Unsurprisingly, a layout strategy that trades horizontal space for vertical space isn't a high priority for a technology primarily aimed at web pages. I wouldn't say that web standards that actually prioritise the web are nothing but "idiocy", I'd say that's entirely sensible.
Let's see.
Yes, I agree: vertical space is infinite and horizontal space is limited. Going from there straight to "no columns" is pulling a fast one, though: lines of text are generally easier to read--in particular the \r\n is easier, ISTR--when lines are roughly 66 characters long [1]. It's generally desirable to have as much text on the screen as possible. The only way of doing both 66 chars/line and packing the screen is by having several lines be adjacent; that is, by doing columns.
One problem of columns is of course that unless you specify Everything (TM), you don't know how tall it's going to be; as it's very desirable not to have to scroll upwards, you're somewhat in a pickle.
I'm toying with an idea for what I think I want: an option for a text block (p or div, no?) that means "show the text as follows: create as many columns of 66-char lines that will fit horizontally, and make them so tall that they take up almost all the screen. Repeat this for any remaining text". That should give the best of all worlds, but there might be issues.
"Microsoft's interest in avoiding unnecessary litigation costs, preserving the time of its employees, insulating OEMs, wholesalers, and retailers from discovery into confidential pricing policies, and maintaining its goodwill far outweighs the interest of class members in relief they never expected before filing this action," Microsoft said.
Let's analyse this.
On one hand, there's Microsoft keeping money saved on lawsuits and salaries, preventing anyone besides themselves (and probably few of themselves at that) from knowing just how much money they extract from you and trying to seem like a Good Corporate Citizen (TM).
On the other hand, there's your interest in saving the money that Microsoft has only been able to demand because they've been able to keep their pricing scheme secret from you.
Microsoft says that money in their pockets is more important than money in your pockets. Colour me unsurprised.
The facts and theories are the result of the scientific method and would be pretty much unthinkable without it.
Yeah.
Kinda. Depends on what you define the scientific method to be. [I'm reiterating a point made previously (http://science.slashdot.org/comments.pl?sid=466370&cid=22553660)]. New scientific knowledge (and let's not go out on a cartesian limb here) is the product of two major components, one of them being being the design and execution of experiments combined with peer review.
The other is insights, ideas, and imaginations. There is no well-known algorithm, no "scientific method", by which we can generate ideas and infer connections in the data we observe. All we can do (as far as I know) is to take the ideas we generate and apply the first component, rigorous testing, to discard the ones of non-maximal usefulness.
This is elegantly summed up in the phrase "It is not the purpose of science to produce infinite wisdom, but to guard against infinite folly" [citation needed] (linked post says Brecht).
As an anecdotal example of what I'm talking about, it one day dawned on me that there must be a reason why code colouring is such a deal-breaking feature of editors (in the mind of most coders, at least), and that the reason is that colours are in some way perceived more readily (easier or faster). [That's the insight component]. Later, I listened to the an OCW intropsych lecture about visual perception, and indeed the lecturer made a few experiments (without the rigour part) with the class showing that:
1. Finding a red circle in a blob of green ones is fast 2. Finding a tee in a blob of ells is slow. 3. Finding a tee in a blob of ells is fast if there's a red tee, no red ells and a bunch of green letters (my assumption about the details).
Presumably, he could apply a bit of rigour and publish a peer-reviewed article containing "my" idea (only it's probably not that new by now).
The important point is: how did I come up with the idea? It just sorta' popped into my head, and no methodology could reliably have produced that idea.
Slashdot Mirror
Wasn't that the plot of a batman movie? ;)
If indeed they're arbitrarily chosen, what are some other natural phenomena suitable for calendaring that affect my life?
From the WP article: The computable numbers include many of the specific real numbers which appear in practice, including all algebraic numbers, as well as e, pi, and many other transcendental numbers. If you start from ZFC you typically construct the real numbers as Cauchy sequences of rationals, each sequence representing its limit. So, let's pick the constant sequence of rational zeroes to represent the real zero. How do you represent this with a finite amount of information? Is it different then, to represent numbers by Turing machines? If so, how?
You don't define what you mean by "exactly" or "relationship", so I can't speak to that very well. How do you express 7's "relationship" to 142857 "exactly"?
(just teasing)
Cars? Just wait, he's thinking about which 'science fiction' technologies may really fly some day. ;)
Yes, whoosh. Or maybe I should learn to make my irony more explicit.
Zemlin: Yes. Yes, I think it will actually. As an extention of my previous comment, I wished InfoWorld would have asked him why he thinks that. I can say "I think I will get -1 Flamebait"; so what? It only becomes interesting when I have something to substantiate it with. If you're a Motorola or an LG, would you rather, per device basis when you're selling tens of millions of devices, license Windows Mobile or the Symbian platform from Nokia, or would you rather have Linux, which is collaboratively designed, which supports every major architecture? He's saying "Would you rather have a, or b, or c? Note that c has properties p and q." For a meaningful comparison, he could list some properties of a and b as well. Sure, we may know some of them, but he could have emphasised what he thinks is relevant. Also, when you've picked an architecture, support for any other architecture is not of much interest. InfoWorld: So are Microsoft's days as the dominant provider of desktop and server and maybe even handheld operating systems numbered?
Zemlin: Monopolies don't last forever, so I mean, I think they've got a long way to go. It's just natural over time that people aren't going to allow a single company to dominate the market. People, as in individual consumers, always allow concentration of power; they don't care, they just want easy. Look at Microsoft; look at Google. People use their services because it's easy for them to do so. I'm conjecturing that it takes the power of the state to break up or prevent the formation of monopolies. Someone enlighten me: what happened with AT&T, Standard Oil and the Railroads? What's happening with Microsoft? InfoWorld: Wouldn't the emergence of Linux kind of say that maybe Microsoft never really was a monopoly, that there was always room for somebody else to compete in there and that's what Linux is now doing?
Zemlin: It obviously was a desktop monopoly for a period of time. Again, there's a claim but no argument. I don't want to argue against him, I just want a better argument. Yes, the days of high-margin, vendor lock-in monopoly practices in the software business, yes, those are gone, and they're permanently gone. This is five minutes after he stated that Microsoft makes 30% profit; that doesn't like either dead or dying to me (sadly). InfoWorld: Can Solaris compete with Linux?
Zemlin: [If Solaris was FLOSS eight years ago maybe, blah blah
(None of this is to say that I disagree with Zemlin's assertions, just that I think he could have made a better argument).
This is a sign that emacs is evil: xenu is an acronym of the precursor to emacs, the POSIX-compatible enux.
(Interestingly, enux was a microkernel OS. That didn't fit will with RMS's idealistic agenda of monolithic kernels, hence emacs.)
Now that's what I call a good story.
I could switch to 640x480, but that'd be a net loss: yeah, I'd get shorter lines, but I could get the same just by wrapping lines at 66 characters at my current resolution. That's be about as efficient (in terms of screen real estate use) as limiting pages to a single $n-pixel column: not efficient at all.
Also, I think you're trying to argue against my suggestion for multi-column pages, but I don't see how your arguments work. Yes, there may be some instances (like you've pointed out) where my solution doesn't have a problem to solve. Great. What about the other instances, are there anything wrong with my idea for those?
I've always wondered about how this scenario is handled:
Stick a bunch of Stuff That Goes Bang up your $CAVITY.
Get onto plane.
Make Stuff That Goes Bang go bang.
Anyone got any ideas? If we want to be protected against every possible scenario, why not start with this?
(wait, did I just advocate mandatory cavity searches?)
In particular, the CD from which I installed Ubuntu has tons of packages on it. You can have physical media without packages (say, Wii Sports), packages without physical media (that's what all the mirrors in my
Slashdot editor. Just think of all the firehose entries you have to post^H^H^H^Hedit.
I, for one, welcome our assembled-in-space armed-with-11-feet-weapons robotic overlords.
You say that making a game is expensive, and that's true; you also state that one wants to maximize profit, with which I can agree (you may want to maximize its first derivative, but let's assume not). You then claim that one gets the most profit by reaching the most users.
How about shelling out $1 supporting Imaginary OS which has one user, and that user won't pay any money for your game? I'd say that's a net loss. Maximizing profit requires a cost-benefit analysis, not just a benefit analysis.
According to the story, having a ceiling cat works fine too. ;)
Dude, you have no idea what kind of dangerous weapons of mass destruction you can disguise as breast milk, hand cream and Starbucks coffee. Terrorists are using them to kidnap and molest your children. Won't somebody think of the children?? With this new scanner, we can protect and--God forbid we will have to--save your children. Why do you hate America's children, you crazy hippie?
(Committee for Aviation Transportation Security says "All your privacy are belong to us")
Yes, I agree: vertical space is infinite and horizontal space is limited. Going from there straight to "no columns" is pulling a fast one, though: lines of text are generally easier to read--in particular the \r\n is easier, ISTR--when lines are roughly 66 characters long [1]. It's generally desirable to have as much text on the screen as possible. The only way of doing both 66 chars/line and packing the screen is by having several lines be adjacent; that is, by doing columns.
One problem of columns is of course that unless you specify Everything (TM), you don't know how tall it's going to be; as it's very desirable not to have to scroll upwards, you're somewhat in a pickle.
I'm toying with an idea for what I think I want: an option for a text block (p or div, no?) that means "show the text as follows: create as many columns of 66-char lines that will fit horizontally, and make them so tall that they take up almost all the screen. Repeat this for any remaining text". That should give the best of all worlds, but there might be issues.
[1] http://www.ctan.org/tex-archive/info/lshort/english/lshort.pdf, search for "longer than 66 characters".
On one hand, there's Microsoft keeping money saved on lawsuits and salaries, preventing anyone besides themselves (and probably few of themselves at that) from knowing just how much money they extract from you and trying to seem like a Good Corporate Citizen (TM).
On the other hand, there's your interest in saving the money that Microsoft has only been able to demand because they've been able to keep their pricing scheme secret from you.
Microsoft says that money in their pockets is more important than money in your pockets. Colour me unsurprised.
Yeah, something like that would be horrible.
Kinda. Depends on what you define the scientific method to be. [I'm reiterating a point made previously (http://science.slashdot.org/comments.pl?sid=466370&cid=22553660)]. New scientific knowledge (and let's not go out on a cartesian limb here) is the product of two major components, one of them being being the design and execution of experiments combined with peer review.
The other is insights, ideas, and imaginations. There is no well-known algorithm, no "scientific method", by which we can generate ideas and infer connections in the data we observe. All we can do (as far as I know) is to take the ideas we generate and apply the first component, rigorous testing, to discard the ones of non-maximal usefulness.
This is elegantly summed up in the phrase "It is not the purpose of science to produce infinite wisdom, but to guard against infinite folly" [citation needed] (linked post says Brecht).
As an anecdotal example of what I'm talking about, it one day dawned on me that there must be a reason why code colouring is such a deal-breaking feature of editors (in the mind of most coders, at least), and that the reason is that colours are in some way perceived more readily (easier or faster). [That's the insight component]. Later, I listened to the an OCW intropsych lecture about visual perception, and indeed the lecturer made a few experiments (without the rigour part) with the class showing that:
1. Finding a red circle in a blob of green ones is fast
2. Finding a tee in a blob of ells is slow.
3. Finding a tee in a blob of ells is fast if there's a red tee, no red ells and a bunch of green letters (my assumption about the details).
Presumably, he could apply a bit of rigour and publish a peer-reviewed article containing "my" idea (only it's probably not that new by now).
The important point is: how did I come up with the idea? It just sorta' popped into my head, and no methodology could reliably have produced that idea.