Efficiency is only important when it's critical. Computers are so fast nowadays that saving every cycle is generally a waste of programming effort. With hardware acceleration and JVM/JIT advances the speed of a javascript codec doesn't need to be an issue - who cares if you could write a faster one in assembler if the javascript were fast enough?
The strategic advantage of a video that comes with it's own codec is the big win - you can use whatever codec you like and have it accessible to anyone with a browser without needing to win any world-domination codec wars.
I guess they could do given that it's now LGPL'd, but who knows if they'd still have a viable business model... They could still sell commercial support and training of course.
Note though that the latest KDE/Trolltech agreement there dates to 2004. It mentions the need to update it after the 2008 Nokia aquisition of Trolltech, but that's it.
Maybe the point is moot after Nokia's LGPLing of Qt, but it's odd that the KDE site doesn't mention it.
Well, Qt has bindings for many languages, and more could be added, so Qt/Silverlight would be possible, as would be creating a C++ compiler that targets the Silverlight VM. With LLVM the latter wouldn't be as difficult as it sounds - just a new code generator.
Nope - HTML5 will kill Adobe/Flash. The future is in low power smartphones and netbooks and Flash is too inefficient. The future of web video is hardware accelerated HTML5 canvas, regardless of what CODECs dominate.
I'd not be surprised to see video come with it's own Javascript based CODEC at some point in the future (thereby making the issue of a standard irrelevant) - I think we're already at the point where it's possible performance-wise.
It was news the first time someone did it, but not the 100th or whatever we're up to now. I've lost count of how many times slashdot has run "balloon takes camera to edge of space for $x" stories.
Nah - if Ford had fake customers coming into their showrooms to make fake purchases to mislead competitors who are watching, it doesn't reduce the legitimacy of watching your competitors (products, sales) to improve your own business. It might make Toyota in this analogy look stupid if they started building ridiculous cars that (wrongly) appeared to be popular at Ford, and that's all that applies here too.
There's nothing wrong with Microsoft trying to glean which of Google's "products" are most popular, so they can boost their own product range/advertising, but it does make them look a tad daft if they've been copying fake products rather than real ones.
Seeing what results a human finds relevant out of a bunch of machine predictions (aka search results) seems a rather obvious and beneficial search engine enhancement. Whether the click-throughs come from page 1 (or 42) of Google's results, Page 42 (or 1) of Bing's, or from paying college students to answer questionaires, or whatever, is entirely irrelevant.
Maybe rather than whining about Microsoft improving Bing in this way, Google should be improving their own search engine the same way, or wouldn't it be surprising (NOT!) if they were already doing this and their page 1 results include yesterdays page 2 results that got lots of click-thrus.
This is a non-story, or at least a non-outrage. It's mildly interesting at best.
That you don't know how Google search works does not excuse Microsoft's behavior.
Well, no, it wouldn't, but it seems your knowledge of Google is based on some antiquated notion of all their results being generated purely by some generic "links confer credit" page rank algorithm... That may have been true when they started, but nowadays Google's results reflect all sorts of special cases that kick-in based on query intent prediction/analysis. There's been plenty of mention of this in the press by Google engineers. For some very trivial examples look what happens when you search for an address, or for anything that looks like a conversion.
Hardly - Google and Bing are both a central algorithm plus dozens (hundreds/thousands?) of special case hacks where the algorithm doesn't cut it.
The quality of Google's (or Bing's) search results isn't due to the brilliance of page rank or any such core algorithm, but rather in deciding which hack to apply when, and in suppressing all the SEO attempts to game their results. The fact that Microsoft have apparently implemented some Google-inspired hacks has zero reflection on their technology - if anything it shows that Microsoft is just as good as Google in deciding when these particular tweaked results are appropriate (else Google would not be complaining).
BTW I hate Microsoft as well as the next guy due to the inept quality and bloated size of their software, but in this case there's really no cause for complaint. Bing is good, and they'd be stupid not to try to copy any enhancements that Google have, and vice versa. Wanna make a bet that Google isn't watching Bing just as intently and making improvements to their search results based on that?
Sounds like someone's getting a bit over-excited here.
Since when is improving a product cheating, and does Google really need a "sting" to figure out that Microsoft is trying to (shock! horror!) keeping an eye on Google using the results to improve Bing?
It seems that Google's complaint (not "complain" you illiterate mofos) is that Bing is just tweaking results instead of investing time/money making those results come out of an algorithm, but I'd be very surprised if Google themselves don't have a zillion special rules in addition to their magical page rank.
getting ready to go over the major ins and outs of the Linux terminal and GUI
WTF? An introduction to the Linux command line/shell (pipes, "find", etc) might be useful, but teaching a GUI to CS students is just insulting and a waste of time.
...to connect the dots and understand how it relates to what they already know about computers
Has it occurred to you that the problem might be you and your choice of topic rather than the students?
For a "linux terminal" class with some meat to it, how about a Perl class instead?
Growth stocks trade at P/E premiums to the market based on their earnings growth rate, but you can't figure growth from a single datapoint!... not enough information
That said, with annual sales of $1.6B and earnings of $0.473B, a market cap of $50B would give a(VERY high) P/E ratio of 105.7 and P/S of 31.25
Compare these with Google, with a market cap of ~$200B, P/E of 24.75 and P/S of 7.12
So, Facebook certainly looks pretty rich at that valuation, but it's early days for them in terms of monetizing their user-base, so there may well be rapid earnings growth to come for a number of years, possibly enough to justify that P/E ratio. OTOH a market cap of $50B isn't small, so there's only so much they can grow... They've got the momentum going for them now though, so they should be able to post good numbers at least for a few years.
The trouble with investing in tech companies though is that the tech future is hard to predict. It's not just about valuation but rather how the competetive landscape changes, and it can change very rapidly. Google could announce a Facebook killer tommorow. Who knows?
Surely the level of detail in the describing the experiment should match the level of specificicity of the claim/theory it aims to prove/disprove.
If your claim is that is a mixture of 50% dog shit and 50% ketchup will spontaneously ignite, then it shouldn't matter what type of container you used to prove the result. If you say the container type matters, then go ahead any specify what you used.
How you fabricated your test equipment would seem to be entirely irrelevant unless your theory was that it matters.
This is basically about poorly designed clinical drug trials without sufficient controls. Sloppy work, even if it seemed rigorous enough at the time.
The sensationalistic "scientific method in question" stuff is pure BS, but after all this is New Yorker magazine we're talking about, so one wouldn't expect too much scientific literacy. It was the scientific method of "predict and test" that caught these erroneous results, so the method itself is fine. The "scientist" who designed a sloppy experiment is too blame, not the method.
However, I'm not sure that psychiatric drug trials even deserve to be called science in the first place. The principle of GIGO (Garbage In - Garbage Out) applies. This is touchy-feely soft science at best. How do you feel today on a scale of 1-10? Do the green pills make you happy?
Online, in general, means "currently connected". There's the sense of being logged into (or connected to) an online service of some sort, but its also used to differentiate secondary storage that's currently accessible (disks and mounted tapes would both count) from that which isn't (e.g. an unmounted disk pack, or tape, or a CD/DVD not in a drive for that matter).
Random access has nothing to do with online/offline. Random access storage refers to storage where any (random) component of it can be accessed in aproximately equal time, and includes hard drives notwithstanding the differencees in access times due to head movement. The other major class is sequential access storage, such as tape drives, where data can only be accessed in a certain sequential order and therefore not all data can be accessed in equal time (nearby data is faster to access).
A random access storage device may currently be either online or offline, as may a sequential access one.
The use of online/offline to refer to storage is becoming less common as storage capacities increase and there's less need for offline storage, but offline backup/archive media is still common.
Punched cards don't belong in the "the evolution of the hard drive"... they weren't used for online storage but rather for a combination of data entry and for data transportation (in which, latter, role they might be considered a precursor to floppys and nowadays USB drives which fill that role).
Punched cards belong to the era of batch computing (submit job, come back later and collect results), before being "online" (initially on a mainframe/minicomputer terminal) became common/possible. Rather than sitting at a computer terminal typing your program in an editor, you'd instead sit at a card punch machine typing your program onto punched cards (one line per card); each keystroke caused that character pattern to be punched onto the card, and, since you can't "unpunch" a card, there was no backspace key - if you made a mistake youd have to feed in a new blank card and could hold down the "copy" key to copy the old card up to the point of your mistake (this rapid copying/punching made a very loud noise like a machine gun).
Once you'd punched your cards you'd put a rubber band around them to keep them in order (if you dropped them, there were sorting machines that could resort them based on numbers punched into the cards), then submit them to the computer operator who, when your time came (no multitasking), would put the cards into a card reader where they'd be read into computer memory for execution. Your printed output (maybe a syntax error, or core dump, or your results if you're program was working), together with your card deck, would be returned to you later when it was available. If you wanted to change your program you could now insert/remove punched cards from your deck, and resubmit the job. Core dumps (printed on fanfold paper, which you'd stretch out across the floor) originated from this batch era, since without the ability to debug your program online (as it runs), this was one way (other than print statements) you could debug them between batch runs.
***
Other than removing puched cards from this "evolution", they should really have stared it with reel-reel mag tape which was the original online storage media, and should really have put removable disk packs in there someplace (disk packs were common with PDP 11/23, etc minicomputers in the early Unix days, and consisted on your disk platters on a spindle in hard plastic housing with a handle on it - the platters were seperate from the drive itself into which you inserted the disk pack. Since disk packs had to have an opening for the disk heads, you were able to smell head crashes where the disk head had crashed (due to a dust particle or whatever) into the surface of your platter and ground it up:-(
***
I was waxing nostalgic over computer storage myselkf the other day. My first home computer c.1978 used a 300 baud (10 bits/char => 30 char/sec) audio cassette for storage, and I well remember the first 5MB personal hard disks (an external unit about the size of a shoe-box) that appeared in the early 80's. It makes me appreciate the 8GB of RAM ($100) I just popped into my latest PC, not to mention the 1TB hard drive.
It sounds more like it's a sensor/noise problem that can (as always) be somewhat improved in software by (time or space) averaging to get rid of the noise.
i.e. while there's a software fix, it wasn't a software problem.
Maybe the folks who wrote the communications app didn't want to wait for a fix (workaround) to be made in the driver, so just put it in the app themselves.
The word becomes meaningless if you use it like that.
Amazon has no authority over whether you publish or sell a book, so they're not in a position to censor you.
The fact that Amazon themselves won't sell your book (or your favorite brand of processed cheese) isn't infringing on your rights in any way whatsoever.
I'm amazed this story didn't get a bad (or rather just plain WRONG) summary tag yet.
You're not likely in the slowest line, but rather you're likely not in the fastest one (well, duh - with N lines you've only got a 1/N probability of being in the fastest one).
FWIW, your probability of being in the single slowest line is just the same, 1/N, IOW you're most likely *NOT* in the slowest line.
A quick Google for "arsenic bacteria" shows that these bacteria from lake Mono were already known to be special since they have an arsenic-based metabolism, so presumably they're already being studied by plenty of scientists.
The news here is that there maybe a link between chemicals used in antibacterial soaps, etc, and immume disfunction (over activity - allergies/etc).
This is NOT at all the same as the trite observation that your immune system (mostly) needs to be exposed to stuff to protect you from it. Lack of protection isn't the same as disfunction, and this isn't about NOT being exposed to anything - it's about BEING exposed to something (certain harmful chemicals).
Of course, correlation isn't causation, and it's not necessarily the chemicals cited that are causing the disfucntion, so (as the authors conclude) this only incidates the need for further study.
Slashdot Mirror
Efficiency is only important when it's critical. Computers are so fast nowadays that saving every cycle is generally a waste of programming effort. With hardware acceleration and JVM/JIT advances the speed of a javascript codec doesn't need to be an issue - who cares if you could write a faster one in assembler if the javascript were fast enough?
The strategic advantage of a video that comes with it's own codec is the big win - you can use whatever codec you like and have it accessible to anyone with a browser without needing to win any world-domination codec wars.
I guess they could do given that it's now LGPL'd, but who knows if they'd still have a viable business model... They could still sell commercial support and training of course.
Note though that the latest KDE/Trolltech agreement there dates to 2004. It mentions the need to update it after the 2008 Nokia aquisition of Trolltech, but that's it.
Maybe the point is moot after Nokia's LGPLing of Qt, but it's odd that the KDE site doesn't mention it.
For anyone looking for it, this is the (no Qt for WP7) link:
http://www.engadget.com/2011/02/11/nokia-notifies-developers-that-qt-is-out-for-windows-phone-devel/
Well, Qt has bindings for many languages, and more could be added, so Qt/Silverlight would be possible, as would be creating a C++ compiler that targets the Silverlight VM. With LLVM the latter wouldn't be as difficult as it sounds - just a new code generator.
Nope - HTML5 will kill Adobe/Flash. The future is in low power smartphones and netbooks and Flash is too inefficient. The future of web video is hardware accelerated HTML5 canvas, regardless of what CODECs dominate.
I'd not be surprised to see video come with it's own Javascript based CODEC at some point in the future (thereby making the issue of a standard irrelevant) - I think we're already at the point where it's possible performance-wise.
It was news the first time someone did it, but not the 100th or whatever we're up to now. I've lost count of how many times slashdot has run "balloon takes camera to edge of space for $x" stories.
Nah - if Ford had fake customers coming into their showrooms to make fake purchases to mislead competitors who are watching, it doesn't reduce the legitimacy of watching your competitors (products, sales) to improve your own business. It might make Toyota in this analogy look stupid if they started building ridiculous cars that (wrongly) appeared to be popular at Ford, and that's all that applies here too.
There's nothing wrong with Microsoft trying to glean which of Google's "products" are most popular, so they can boost their own product range/advertising, but it does make them look a tad daft if they've been copying fake products rather than real ones.
Seeing what results a human finds relevant out of a bunch of machine predictions (aka search results) seems a rather obvious and beneficial search engine enhancement. Whether the click-throughs come from page 1 (or 42) of Google's results, Page 42 (or 1) of Bing's, or from paying college students to answer questionaires, or whatever, is entirely irrelevant.
Maybe rather than whining about Microsoft improving Bing in this way, Google should be improving their own search engine the same way, or wouldn't it be surprising (NOT!) if they were already doing this and their page 1 results include yesterdays page 2 results that got lots of click-thrus.
This is a non-story, or at least a non-outrage. It's mildly interesting at best.
That you don't know how Google search works does not excuse Microsoft's behavior.
Well, no, it wouldn't, but it seems your knowledge of Google is based on some antiquated notion of all their results being generated purely by some generic "links confer credit" page rank algorithm... That may have been true when they started, but nowadays Google's results reflect all sorts of special cases that kick-in based on query intent prediction/analysis. There's been plenty of mention of this in the press by Google engineers. For some very trivial examples look what happens when you search for an address, or for anything that looks like a conversion.
Hardly - Google and Bing are both a central algorithm plus dozens (hundreds/thousands?) of special case hacks where the algorithm doesn't cut it.
The quality of Google's (or Bing's) search results isn't due to the brilliance of page rank or any such core algorithm, but rather in deciding which hack to apply when, and in suppressing all the SEO attempts to game their results. The fact that Microsoft have apparently implemented some Google-inspired hacks has zero reflection on their technology - if anything it shows that Microsoft is just as good as Google in deciding when these particular tweaked results are appropriate (else Google would not be complaining).
BTW I hate Microsoft as well as the next guy due to the inept quality and bloated size of their software, but in this case there's really no cause for complaint. Bing is good, and they'd be stupid not to try to copy any enhancements that Google have, and vice versa. Wanna make a bet that Google isn't watching Bing just as intently and making improvements to their search results based on that?
Sounds like someone's getting a bit over-excited here.
Since when is improving a product cheating, and does Google really need a "sting" to figure out that Microsoft is trying to (shock! horror!) keeping an eye on Google using the results to improve Bing?
It seems that Google's complaint (not "complain" you illiterate mofos) is that Bing is just tweaking results instead of investing time/money making those results come out of an algorithm, but I'd be very surprised if Google themselves don't have a zillion special rules in addition to their magical page rank.
WTF? An introduction to the Linux command line/shell (pipes, "find", etc) might be useful, but teaching a GUI to CS students is just insulting and a waste of time.
Has it occurred to you that the problem might be you and your choice of topic rather than the students?
For a "linux terminal" class with some meat to it, how about a Perl class instead?
Is it really social networking if most people are anonymous (either literally or using a handle). How may people make their e-mail addresses public?
This is just a giant chat group.
Growth stocks trade at P/E premiums to the market based on their earnings growth rate, but you can't figure growth from a single datapoint! ... not enough information
That said, with annual sales of $1.6B and earnings of $0.473B, a market cap of $50B would give a(VERY high) P/E ratio of 105.7 and P/S of 31.25
Compare these with Google, with a market cap of ~$200B, P/E of 24.75 and P/S of 7.12
So, Facebook certainly looks pretty rich at that valuation, but it's early days for them in terms of monetizing their user-base, so there may well be rapid earnings growth to come for a number of years, possibly enough to justify that P/E ratio. OTOH a market cap of $50B isn't small, so there's only so much they can grow... They've got the momentum going for them now though, so they should be able to post good numbers at least for a few years.
The trouble with investing in tech companies though is that the tech future is hard to predict. It's not just about valuation but rather how the competetive landscape changes, and it can change very rapidly. Google could announce a Facebook killer tommorow. Who knows?
Surely the level of detail in the describing the experiment should match the level of specificicity of the claim/theory it aims to prove/disprove.
If your claim is that is a mixture of 50% dog shit and 50% ketchup will spontaneously ignite, then it shouldn't matter what type of container you used to prove the result. If you say the container type matters, then go ahead any specify what you used.
How you fabricated your test equipment would seem to be entirely irrelevant unless your theory was that it matters.
Did you even read the article?
This is basically about poorly designed clinical drug trials without sufficient controls. Sloppy work, even if it seemed rigorous enough at the time.
The sensationalistic "scientific method in question" stuff is pure BS, but after all this is New Yorker magazine we're talking about, so one wouldn't expect too much scientific literacy. It was the scientific method of "predict and test" that caught these erroneous results, so the method itself is fine. The "scientist" who designed a sloppy experiment is too blame, not the method.
However, I'm not sure that psychiatric drug trials even deserve to be called science in the first place. The principle of GIGO (Garbage In - Garbage Out) applies. This is touchy-feely soft science at best. How do you feel today on a scale of 1-10? Do the green pills make you happy?
Online, in general, means "currently connected". There's the sense of being logged into (or connected to) an online service of some sort, but its also used to differentiate secondary storage that's currently accessible (disks and mounted tapes would both count) from that which isn't (e.g. an unmounted disk pack, or tape, or a CD/DVD not in a drive for that matter).
Random access has nothing to do with online/offline. Random access storage refers to storage where any (random) component of it can be accessed in aproximately equal time, and includes hard drives notwithstanding the differencees in access times due to head movement. The other major class is sequential access storage, such as tape drives, where data can only be accessed in a certain sequential order and therefore not all data can be accessed in equal time (nearby data is faster to access).
A random access storage device may currently be either online or offline, as may a sequential access one.
The use of online/offline to refer to storage is becoming less common as storage capacities increase and there's less need for offline storage, but offline backup/archive media is still common.
Punched cards don't belong in the "the evolution of the hard drive"... they weren't used for online storage but rather for a combination of data entry and for data transportation (in which, latter, role they might be considered a precursor to floppys and nowadays USB drives which fill that role).
Punched cards belong to the era of batch computing (submit job, come back later and collect results), before being "online" (initially on a mainframe/minicomputer terminal) became common/possible. Rather than sitting at a computer terminal typing your program in an editor, you'd instead sit at a card punch machine typing your program onto punched cards (one line per card); each keystroke caused that character pattern to be punched onto the card, and, since you can't "unpunch" a card, there was no backspace key - if you made a mistake youd have to feed in a new blank card and could hold down the "copy" key to copy the old card up to the point of your mistake (this rapid copying/punching made a very loud noise like a machine gun).
Once you'd punched your cards you'd put a rubber band around them to keep them in order (if you dropped them, there were sorting machines that could resort them based on numbers punched into the cards), then submit them to the computer operator who, when your time came (no multitasking), would put the cards into a card reader where they'd be read into computer memory for execution. Your printed output (maybe a syntax error, or core dump, or your results if you're program was working), together with your card deck, would be returned to you later when it was available. If you wanted to change your program you could now insert/remove punched cards from your deck, and resubmit the job. Core dumps (printed on fanfold paper, which you'd stretch out across the floor) originated from this batch era, since without the ability to debug your program online (as it runs), this was one way (other than print statements) you could debug them between batch runs.
***
Other than removing puched cards from this "evolution", they should really have stared it with reel-reel mag tape which was the original online storage media, and should really have put removable disk packs in there someplace (disk packs were common with PDP 11/23, etc minicomputers in the early Unix days, and consisted on your disk platters on a spindle in hard plastic housing with a handle on it - the platters were seperate from the drive itself into which you inserted the disk pack. Since disk packs had to have an opening for the disk heads, you were able to smell head crashes where the disk head had crashed (due to a dust particle or whatever) into the surface of your platter and ground it up :-(
***
I was waxing nostalgic over computer storage myselkf the other day. My first home computer c.1978 used a 300 baud (10 bits/char => 30 char/sec) audio cassette for storage, and I well remember the first 5MB personal hard disks (an external unit about the size of a shoe-box) that appeared in the early 80's. It makes me appreciate the 8GB of RAM ($100) I just popped into my latest PC, not to mention the 1TB hard drive.
It sounds more like it's a sensor/noise problem that can (as always) be somewhat improved in software by (time or space) averaging to get rid of the noise.
i.e. while there's a software fix, it wasn't a software problem.
Maybe the folks who wrote the communications app didn't want to wait for a fix (workaround) to be made in the driver, so just put it in the app themselves.
The word becomes meaningless if you use it like that.
Amazon has no authority over whether you publish or sell a book, so they're not in a position to censor you.
The fact that Amazon themselves won't sell your book (or your favorite brand of processed cheese) isn't infringing on your rights in any way whatsoever.
They're trying to build a simulator for Steve Jobs ego before he kicks the bucket.
I'm amazed this story didn't get a bad (or rather just plain WRONG) summary tag yet.
You're not likely in the slowest line, but rather you're likely not in the fastest one (well, duh - with N lines you've only got a 1/N probability of being in the fastest one).
FWIW, your probability of being in the single slowest line is just the same, 1/N, IOW you're most likely *NOT* in the slowest line.
A quick Google for "arsenic bacteria" shows that these bacteria from lake Mono were already known to be special since they have an arsenic-based metabolism, so presumably they're already being studied by plenty of scientists.
http://www.newscientist.com/article/dn14537-arseniceating-bacteria-rewrite-evolutionary-history.html
Of course having "DNA" (I guess it needs a new name since it's a new compond) based on arsenic raises the interest level a lot, assuming its true!
The news here is that there maybe a link between chemicals used in antibacterial soaps, etc, and immume disfunction (over activity - allergies/etc).
This is NOT at all the same as the trite observation that your immune system (mostly) needs to be exposed to stuff to protect you from it. Lack of protection isn't the same as disfunction, and this isn't about NOT being exposed to anything - it's about BEING exposed to something (certain harmful chemicals).
Of course, correlation isn't causation, and it's not necessarily the chemicals cited that are causing the disfucntion, so (as the authors conclude) this only incidates the need for further study.