With off-the-shelf packages, perhaps. Though training does help, just as with voice recognition.
As a researcher in the computer vision field, I've seen 99.8% accuracy used for OCR on Tibetan religious texts that are in far from good condition. There are several Tibetan OCT projects on the go which aim to save religious texts from destruction.
I don't care how slow it is, or how unstable it is. I won't be using it. I want a stripped version of Windows so that I can purchase a pre-built system without shelling out $179 for Windows Foo.
If this new version of Windows sells (OEM) for $99, that's $80 in my pocket. Now that's benefit to the consumer.
It would be better if they could force Microsoft to abandon exclusionary licensing contracts, but since that won't happen, I'll be happy with this.
I can't give you an official reason, but in my opinion:
C was used because it's the default language. Not just default for Unix(tm), but the first language most people think to use for a multitude of tasks. C is fast, easy to code in, and both upwards and downwards compatible (ie, I can use GTK+ with assembly code or C++/Java/Perl)
While there's nothing that C++ does that C can't do with appropriate macros, it's harder to use a C++ library in C apps. It's easy and commonplace, however, to use a C library in C++ apps.
Also, I imagine C was the language of choice for the people who started GTK+. This has a huge impact, because despite all else that I've said, there's little functional reason to pick C, C++, or any other useful language. Personal preference and the way one approaches a problem (ie, OO, functional) are big influences.
All of this seems obvious, so why would you even ask why C was chosen as the language for GTK+? You'll get the same answer for any project: C is fast, cheap and good. This is why it's used for X, the kernel, and most other large, popular applications. Mozilla is the only app I use on a regular basis that isn't written in C, and it's dog-slow. Coincidence?
It sounds like two issues to me, or else misuse of the word 'security.' I think the word they were looking for was 'robustness,' in that severing a single line shouldn't cause discontinuities.
I know, my power gets cut off from time to time, too. But I think the 'high-voltage' (ie. non-residential) grids have redundancy in the same way that this proposed network will have better redundancy that the pedestrian Internet (where too many pidgeons on a line in Minneanapolis causes my connection to slow to a crawl).
>Third, XML parsers are no-where near as ubiquotous
>as you would like to believe. To date there are no
>(zero, zip, nada, nothing) fully compliant XML
>parsers.
That's not true!/dev/null is a fully compliant XML parser.
Actually, I've been quite impressed lately. I don't know what ad network it is, but for the last few months I've been inundated with ads from my local phone company. Telus' "Special Offer" for University of Calgary students comes up whenever I browse from school.
There's a quite popular game available now called Kohan : Immortal Sovereigns from Timegame Studios. It is available for Linux through Loki. I love this game : It has zones of supply, morale, leadership bonuses, terrain defense bonuses and other wargame concepts.
By far the most unique thing about Kohan is its 'company' concept. You commision companies with front-line and support units, and they fulfill their roles according to their own AI : No click-fests, just send them into battle and watch it unfold. This may make it sound like there's no strategy, but it really brings into play tactics such as flanking and preemptive strikes.
Clerics healing enemy infantry faster than you can kill them? Good thing you've force-marched a company of calvary behind enemy lines: those clerics won't last long under the lances. Opponent just commissioned a mighty army you have no chance of defeating? Well, strike before they're fully supplied and up to strength.
I strongly encourage people to check this game out . I had almost given up on strategy games before this came along, and the fantastic Linux port is icing on the cake.
I could be wrong, but as I understand it, "Real Stereo" as you call it splits the audio into two channels each with HALF the bitrate of the entire stream. The benefit of joint-stereo is that for sections of the file where there is no 'discernable' difference between the channels (or, for that matter, no real difference - a lot of music doesn't use stereo as much as you think it might) the stream gets the entire bitrate.
I would rather have a 256kbps center channel when there's no need to split it than two 128kbps channels with identical information.
Yes, I'll agree. This is possible, but is much more expensive than the systems I was talking about. You need two cameras in order to do this, so that you can individually track the eyes and calculate what they're looking at.
There are still some restrictions : You usually need to be a fixed distance from the monitor, although I imagine a laser range-finder (further adding to the cost) could solve that problem.
It's not that hard, really, but you need a lot of high-quality equipment.
This can be done, and is done for physically disabled people who cannot use a mouse. The problem is that a sub-$100 'webcam' doesn't cut it. You need high-quality video in reasonably real-time in order for this to work. The cameras are more like $400-$500 (CDN, so that's $250-$350 USD).
The main problem with the systems right now is that they cannot track only eye movement. You need to use your whole head for large-distance traversals. You think RSI sucks in your wrist, wait until you start getting neck cramps from your favorite RTS game.
This might be an urban legend, but my Doctoral advisor told me this story about a graduate project that went like this :
An HCI (Human Computer Interface) student was looking for a new way to get input into a computer. Partnering with a biological science student, the two created a device which measured simple brain activity and transformed it into keystrokes. After training himself for a few weeks, he was able to type 25-30 words per minute by thinking about each letter.
The response from his defence committe : "So? I can _type_ faster than that!"
If the United States Government wants to restrict civil liberties for the purposes of keeping the populace safe, then they are admitting that a free people cannot be a secure people.
The Freedom Experiment has failed, my friends. Let us all run back to the warmth and security of the police state.
"The Pattern on the Stone : The Simple Ideas That Make Computers Work" by Daniel Hillis
This book is a great primer on computing for anyone, and an enjoyable read to boot. It covers basic hardware concepts, starting from logic gates, etc, before moving up into high-level concepts such as programming. The last bit of the book contains a look a AI, perceptrons and other "frontiers" of computing.
This short book is a must-buy for anyone interested in computers, and is the first book I recommend to anyone starting out.
When you get one of those nice spams that ask you to reply that that address with the word remove, it's awfully satisfying to spoof a message from postmaster@.com with a subject of remove.
The first two lines of the white paper have the phrase 'paradigm shift' and the word 'digerati'. If we were playing buzzword bingo, the first paragraph would end the game.
I havn't seen this much I'm-trying-to-be-hip-everyone-look-at-me since Wired was launched.
Maybe there are honest-to-goodness people inside those Armanis, but I think they need to spend a few weeks reading -dev lists themselves before anything else. Developers, especially Open Source developers, tend to be a pragmatic lot. While 'paradigm shifts' may be arousing for the marketing deparment and Ted in Sales, I don't think they'll have the intended response with the Open Source community.
Wow - you actually have one of the recent nightlies that registers clicks (bug <A HREF="http://bugzilla.mozilla.org/show_bug.cgi?id= 33952">33952</A>) or doesn't segfault on any action whatsoever (bug <A HREF="http://bugzilla.mozilla.org/show_bug.cgi?id= 34528"34528</A>)??? That's impressive.
My X fonts look great - better than Windows, in my opinion. I guess it depends on your font server, but if you'd looked at the screenshots I'd linked to, you'd have seen how pretty Netscape shows things.
This reply goes to everyone else who suggested that it was a problem with my font server/config.
All right, I'll give you a specific example : the Slashdot story previous to this one. The l0phtCrack story has ~350 comments. In Mozilla M14, this took 116 seconds to render - That's almost two full minutes!!
In Netscape 4.72, the same page rendered in about six seconds.
I'm writing this in mozilla M14 right now, and I've one thing to say about it : the fonts are still AWFUL! See Mozilla Bug # 29726 for my report and screenshots - does everyone else see this problem? Any ideas what to do?
This would be a great thing for some people (although fragmenting Windows even more would be a pain in the ass), but think about this : A great deal of Windows code doesn't belong to Microsoft.
MS has NDAs with hundreds of companies, ranging from Apple (Sorenson codec, anyone?) to Intel to Unisys. Hell, I think I've heard there's some Berkely code in Windows that's legally undisclosable.
If they disclosed only what they had written, you'd have a pretty incomplete picture, which may or may not be useful.
Slashdot Mirror
K12, or more usually K-12 means Kindergarten through twelfth grade.
With off-the-shelf packages, perhaps. Though training does help, just as with voice recognition.
As a researcher in the computer vision field, I've seen 99.8% accuracy used for OCR on Tibetan religious texts that are in far from good condition. There are several Tibetan OCT projects on the go which aim to save religious texts from destruction.
OCR is very popular with Asian dialects.
I don't care how slow it is, or how unstable it is. I won't be using it. I want a stripped version of Windows so that I can purchase a pre-built system without shelling out $179 for Windows Foo.
If this new version of Windows sells (OEM) for $99, that's $80 in my pocket. Now that's benefit to the consumer.
It would be better if they could force Microsoft to abandon exclusionary licensing contracts, but since that won't happen, I'll be happy with this.
I can't give you an official reason, but in my opinion :
C was used because it's the default language. Not just default for Unix(tm), but the first language most people think to use for a multitude of tasks. C is fast, easy to code in, and both upwards and downwards compatible (ie, I can use GTK+ with assembly code or C++/Java/Perl)
While there's nothing that C++ does that C can't do with appropriate macros, it's harder to use a C++ library in C apps. It's easy and commonplace, however, to use a C library in C++ apps.
Also, I imagine C was the language of choice for the people who started GTK+. This has a huge impact, because despite all else that I've said, there's little functional reason to pick C, C++, or any other useful language. Personal preference and the way one approaches a problem (ie, OO, functional) are big influences.
All of this seems obvious, so why would you even ask why C was chosen as the language for GTK+? You'll get the same answer for any project: C is fast, cheap and good. This is why it's used for X, the kernel, and most other large, popular applications. Mozilla is the only app I use on a regular basis that isn't written in C, and it's dog-slow. Coincidence?
It sounds like two issues to me, or else misuse of the word 'security.' I think the word they were looking for was 'robustness,' in that severing a single line shouldn't cause discontinuities.
I know, my power gets cut off from time to time, too. But I think the 'high-voltage' (ie. non-residential) grids have redundancy in the same way that this proposed network will have better redundancy that the pedestrian Internet (where too many pidgeons on a line in Minneanapolis causes my connection to slow to a crawl).
>Third, XML parsers are no-where near as ubiquotous
/dev/null is a fully compliant XML parser.
>as you would like to believe. To date there are no
>(zero, zip, nada, nothing) fully compliant XML
>parsers.
That's not true!
Actually, I've been quite impressed lately. I don't know what ad network it is, but for the last few months I've been inundated with ads from my local phone company. Telus' "Special Offer" for University of Calgary students comes up whenever I browse from school.
There's a quite popular game available now called Kohan : Immortal Sovereigns from Timegame Studios. It is available for Linux through Loki. I love this game : It has zones of supply, morale, leadership bonuses, terrain defense bonuses and other wargame concepts.
By far the most unique thing about Kohan is its 'company' concept. You commision companies with front-line and support units, and they fulfill their roles according to their own AI : No click-fests, just send them into battle and watch it unfold. This may make it sound like there's no strategy, but it really brings into play tactics such as flanking and preemptive strikes.
Clerics healing enemy infantry faster than you can kill them? Good thing you've force-marched a company of calvary behind enemy lines: those clerics won't last long under the lances. Opponent just commissioned a mighty army you have no chance of defeating? Well, strike before they're fully supplied and up to strength.
I strongly encourage people to check this game out . I had almost given up on strategy games before this came along, and the fantastic Linux port is icing on the cake.
I could be wrong, but as I understand it, "Real Stereo" as you call it splits the audio into two channels each with HALF the bitrate of the entire stream. The benefit of joint-stereo is that for sections of the file where there is no 'discernable' difference between the channels (or, for that matter, no real difference - a lot of music doesn't use stereo as much as you think it might) the stream gets the entire bitrate.
I would rather have a 256kbps center channel when there's no need to split it than two 128kbps channels with identical information.
Yes, I'll agree. This is possible, but is much more expensive than the systems I was talking about. You need two cameras in order to do this, so that you can individually track the eyes and calculate what they're looking at.
There are still some restrictions : You usually need to be a fixed distance from the monitor, although I imagine a laser range-finder (further adding to the cost) could solve that problem.
It's not that hard, really, but you need a lot of high-quality equipment.
This can be done, and is done for physically disabled people who cannot use a mouse. The problem is that a sub-$100 'webcam' doesn't cut it. You need high-quality video in reasonably real-time in order for this to work. The cameras are more like $400-$500 (CDN, so that's $250-$350 USD).
The main problem with the systems right now is that they cannot track only eye movement. You need to use your whole head for large-distance traversals. You think RSI sucks in your wrist, wait until you start getting neck cramps from your favorite RTS game.
This might be an urban legend, but my Doctoral advisor told me this story about a graduate project that went like this :
An HCI (Human Computer Interface) student was looking for a new way to get input into a computer. Partnering with a biological science student, the two created a device which measured simple brain activity and transformed it into keystrokes. After training himself for a few weeks, he was able to type 25-30 words per minute by thinking about each letter.
The response from his defence committe : "So? I can _type_ faster than that!"
If the United States Government wants to restrict civil liberties for the purposes of keeping the populace safe, then they are admitting that a free people cannot be a secure people.
The Freedom Experiment has failed, my friends. Let us all run back to the warmth and security of the police state.
Nah. I'd rather die free.
"The Pattern on the Stone : The Simple Ideas That Make Computers Work" by Daniel Hillis
This book is a great primer on computing for anyone, and an enjoyable read to boot. It covers basic hardware concepts, starting from logic gates, etc, before moving up into high-level concepts such as programming. The last bit of the book contains a look a AI, perceptrons and other "frontiers" of computing.
This short book is a must-buy for anyone interested in computers, and is the first book I recommend to anyone starting out.
When you get one of those nice spams that ask you to reply that that address with the word remove, it's awfully satisfying to spoof a message from postmaster@.com with a subject of remove.
The first two lines of the white paper have the phrase 'paradigm shift' and the word 'digerati'. If we were playing buzzword bingo, the first paragraph would end the game.
I havn't seen this much I'm-trying-to-be-hip-everyone-look-at-me since Wired was launched.
Maybe there are honest-to-goodness people inside those Armanis, but I think they need to spend a few weeks reading -dev lists themselves before anything else. Developers, especially Open Source developers, tend to be a pragmatic lot. While 'paradigm shifts' may be arousing for the marketing deparment and Ted in Sales, I don't think they'll have the intended response with the Open Source community.
http://www.stupidonline.com/products/COOL-0011.htm l
Shipping April 13.
> (Posted with 2000032909)
= 33952">33952</A>) or doesn't segfault on any action whatsoever (bug <A HREF="http://bugzilla.mozilla.org/show_bug.cgi?id= 34528"34528</A>)??? That's impressive.
Wow - you actually have one of the recent nightlies that registers clicks (bug <A HREF="http://bugzilla.mozilla.org/show_bug.cgi?id
Thanks for some confirmation that I'm not on crack.
If you'd like to see Slashdot look pretty, please vote for my bug at Mozilla bug 29726.
My X fonts look great - better than Windows, in my opinion. I guess it depends on your font server, but if you'd looked at the screenshots I'd linked to, you'd have seen how pretty Netscape shows things.
This reply goes to everyone else who suggested that it was a problem with my font server/config.
All right, I'll give you a specific example : the Slashdot story previous to this one. The l0phtCrack story has ~350 comments. In Mozilla M14, this took 116 seconds to render - That's almost two full minutes!!
In Netscape 4.72, the same page rendered in about six seconds.
Yeah, my two-day old nightly release did the same thing. Eventually gave up and used command line ftp.
I'm writing this in mozilla M14 right now, and I've one thing to say about it : the fonts are still AWFUL! See Mozilla Bug # 29726 for my report and screenshots - does everyone else see this problem? Any ideas what to do?
It's time to inject some journalistic integrity into the New York Times. And maybe they'll rid of the registration, too.
This would be a great thing for some people (although fragmenting Windows even more would be a pain in the ass), but think about this : A great deal of Windows code doesn't belong to Microsoft.
MS has NDAs with hundreds of companies, ranging from Apple (Sorenson codec, anyone?) to Intel to Unisys. Hell, I think I've heard there's some Berkely code in Windows that's legally undisclosable.
If they disclosed only what they had written, you'd have a pretty incomplete picture, which may or may not be useful.
Just something to think about.