There are plenty of other languages which read left to right and are supported like Arabic. There are harder asian languages supported: Korean, Japanese, and the worst Chinese.
Mac OSX supports Hebrew. The real issue is cost, not ease. Working for a speech/language company, it is the total cost of a product, not how hard it is to develop that kills most projects like this.
We dropped Japanese, not because it was hard (the product was complete and japanese had been done in previous versions). It was dropped because the salary for QA, support, management, OEM sales chain, advertising, and maintanance were just too high. There was very little reuse of staff due to the language, a QA engineer who does not know Japanese (Hebrew) isn't going to be any help. One more language means one more product in the release schedual, which extends the time it takes to make releases and move on to developing the next new killer feature.
What incentive does MS really have? Some small % of the 4% of their sales in a country (This is Mac specific, not all %4 is Mac). It's a big drop in the bucket, but its not enough to pay for all those people and the potential for derailment of other projects. What is the potential to sell this 'feature' to recoup the cost? HEbrew on Mac Office? Very little to none I'd guess.
No, its not because its 'hard' (and I doubt its that). It's cost verses potential profit. When looking for a reason, look to money first.
So as prior art did they list the PC?
I'm sure I've managed to rip CDs to the hard drive as the same time I'm playing music. Sure it's audio vs video, but it amounts to the same thing don't it?
Were you ripping your audio CD to hard drive while listening to that same CD 30 seconds earlier on that same track?
And there is a huge difference between dealing with a readable medium (a CD) and a broadcast stream (live TV/Radio). You can go forward and backward on the CD. It is part of the data delivery, namely physical recorded media.
The Tivo Patent specifically talks about realtime streamed data.
Compairing CD ripping/listening to the Tivo patent is like compairing the match to the lightbulb. They both get hot! They both give light!!! There is plenty of prior art!
The origional post has bee updated in a vague way notting that this is a dispute about money and not about being monitored without explaining.
And no-one has posted the explination, and it seems that no one in the newengland area who watches the news posts to/. (and neither do plow operators is seems).
At issue is the pay drivers recieve will be tied to the GPS system. The gps system being used was shown to have issues tracking durring really heavy storms and when more than one device is really close together (one may drop off from the monitors standpoint).
Plows work the hardest in heavy storms, and also do alot of tandum work. One plow within 2 car lengths of the other.
The dispute is about their pay. Their pay would automatically be assigned from the GPS monotoring system. There is no proceedure for plow operators to dispute the GPS data. If the GPS system goes down, the older means of recieving pay would NOT go into effect, the plow operators would go UNPAID.
That was the dispute.
As for the public safty issues, there were plenty of other private plow operators who were happy to sign a temporary contract to plow for this weekends storm (With the GPS). The public safty issue was used by current contractors to pull for public support and to pressure the administration.
Everytime Maxtor releases a new 'largest' hard drive to the market, they always release the 5400, 2mb cache version first. I dont know if the reason if due to manufacturing or not. They have more plant lines producing the slower speed/less cache drives for OEM VAR's. Still I like to think its smart marketing, so all the techies will run out to buy the 'largest' drive at the highest price now.
When the 120's came out I waited a painfull 6 weeks (yes I'm a geek.. it was painful for me) for the 7200rpm version to be released. With the 160's I faithfully waited almost two months, and got a rebate to boot.
In the end its not if this 5400rpm drive will perform close to smaller 5400rpm drives, but if the latency and seek on a 300gb 7200rpm is small enough to warrent buying only one drive when ATA cards are cheap and two drives could get you better performance for more storage.
If you really need to cram multiple 300gb drives into any enclosure, speed is not what you are going after, its density, so its a moot point on that scale.
Voice Recognition is essential in hospitals and courts all over the US.
It is a multi-million dollar a year buisness.
As for the future of speech recognition, people often make the mistake of seeing it as a replacement for all other input interfaces. This is just absurd, as you point out.
More and more products (like cell phones, car GPS, ATM's, LARAN ) use speech recognition in conjunction with other more traditional means of input/output to get the job done.
It is NOT the fastest or the most reliable means of interfacing, but it is the most NATURAL.
Speaking as someone who works on a product with LNP grammars, we have them already. (blatent plug, burning karma: http://www.scansoft.com/ )
Products like Voice Xpress, Dragon Naturally Speaking, and even IBM's Via Voice have had very powerfull NLP processing with full speech recognition for quite some time.
Other PIM products in the works have NLP processing on the server that is just amazing. Where these systems fall apart is with the actual speech recognition. Even the server processing systems at best get only 99.4% accuracy for adapted speech data.
NLP has come along way from even these humble beginnings as well. New technologies being developed for multimodal are starting to take shape. Currently this technology is only being used in the medical and transcription domains, but this is changing. On-star, many in-car GPS systems, and services that read you your e-mail and schedual from your cell phone are becomming more popular. The demand for good NLP grammar generation is high, and companies that have the engineers that know how to make a GOOD nlp grammar are limited.
In the end the gating issue is not the nlp grammar but the quality of speech audio data and the need for realtime recognition. Noisy car audio with a celular telephony feature set is very difficult to recognize with any accuracy, especially when the user wants an imediate responce. NLP inherently means more words are spoken. (i.e. 'where is the closest starbucks' verses 'find starbucks'.) More recongition errors + the flexability of an NLP grammar (verses a FSG) means things that are realy realy wrong will appear correct to the device.
Recognition is currently the bottleneck, not the NLP.
Heat will definatly be an issue, but much less power will be required. The majority of the power required by chips is used to push data on and off the chip. It takes alot of poser to move a signal from a 25 micron PCB path.
This technology (if it pans out) will mostlikly enter teh private sector in cell phones, DVD players and other small consumer electronics that have a very large number of units produced.
Silicon wafer production has always had one major problem. Impurities. The ability to use more of the waffer to produce smaller chips that can later be 'put back together' in arrays that may not be any larger than the origional single chip solution has the potential to be much cheaper to manufacture in mass quantity.
Granted this is part of the theory behind 6 Sigma, which does not always work out.
Reading the links is the Rosetta for this story.
on
Telstar 4 is Down
·
· Score: 5, Informative
Gah, I read this one 5 times and still didnt understand. After reading all the links and doing some googling here it is again in a lower form of techno-geek (or would it be higher? not sure).
T4 is a broadcast Satalite used to transmit the raw station feeds from the central offices (read networks) to local broadcasters (cable, local ABC affiliate etc).
What Data T4 is responcible for: Robust broadcast and syndication neighborhood anchored by ABC and CBS; host to SNG, data and distance learning applications. (Also hosted is Spice Channel and two other adult viewing stations which are not mentioned on the main site)
Where it effects: The central portion of North America
The origional poster mentions the Carrier frequencies that the data is normally transfered on. The poster also mentions that there is no data being broadcast, just the main freq.
I have no way of confirming this myself, and dont see anything about it on any of the satalite pages. All they report is that a satalite is up and running (i.e. the Carrier frequency is present, but nothing on the data being transmitted)
As for the information on replacing the old T4 with T8, well T8 is currently handling south america, and the information I could find on google about the sale of the out dated satalites it vague at best.
Who does this effect: Possably people out in the midwest are not getting any TV on their cable boxes. mostlikly its effecting the cable companies and local affiliates who need to reroute to another satalite, and the central offices which also need to switch the feeds to other sat feed providers.
In the end nothing is confirmed.
Re:What's with all of the bellyaching about speed?
on
Does C# Measure Up?
·
· Score: 1
The performance argument is a red herring.
I would not go that far.
As many posters have pointed out, its not really about which single language is better overall, but which languages are best suited for which applications.
There is a reason why perl and python have a published C/API for extending and imbeding, there is a reason for JNI, blah, blah, blah.
As with many things in life, the intended purpose of the article is not its best use.
After reading the article it has changed my opinion on when and where to draw the line between the different languages. Many times I have planned from the start to do my network programming in C over C++ or Java due to the belief that they were slower or contained undue overhead.
Which language has the best performance may be what the author intended to write an article about, in the end it is more a helpfull guide to help decide which side of the fence to write that piece of code.
Slashdot Mirror
Anyone know if Dreamworks has a good inhouse distrobution channel, or if they use someone else?
Dreamworks has the only other animation house out there worth its salt. Imagine if these two joined up?
Members of Congress (both senate and house) are exempt from the DMCA. It does not apply to them.
There are plenty of other languages which read left to right and are supported like Arabic. There are harder asian languages supported: Korean, Japanese, and the worst Chinese.
Mac OSX supports Hebrew. The real issue is cost, not ease. Working for a speech/language company, it is the total cost of a product, not how hard it is to develop that kills most projects like this.
We dropped Japanese, not because it was hard (the product was complete and japanese had been done in previous versions). It was dropped because the salary for QA, support, management, OEM sales chain, advertising, and maintanance were just too high. There was very little reuse of staff due to the language, a QA engineer who does not know Japanese (Hebrew) isn't going to be any help. One more language means one more product in the release schedual, which extends the time it takes to make releases and move on to developing the next new killer feature.
What incentive does MS really have? Some small % of the 4% of their sales in a country (This is Mac specific, not all %4 is Mac). It's a big drop in the bucket, but its not enough to pay for all those people and the potential for derailment of other projects. What is the potential to sell this 'feature' to recoup the cost? HEbrew on Mac Office? Very little to none I'd guess.
No, its not because its 'hard' (and I doubt its that). It's cost verses potential profit. When looking for a reason, look to money first.
So as prior art did they list the PC? I'm sure I've managed to rip CDs to the hard drive as the same time I'm playing music. Sure it's audio vs video, but it amounts to the same thing don't it? Were you ripping your audio CD to hard drive while listening to that same CD 30 seconds earlier on that same track? And there is a huge difference between dealing with a readable medium (a CD) and a broadcast stream (live TV/Radio). You can go forward and backward on the CD. It is part of the data delivery, namely physical recorded media. The Tivo Patent specifically talks about realtime streamed data. Compairing CD ripping/listening to the Tivo patent is like compairing the match to the lightbulb. They both get hot! They both give light!!! There is plenty of prior art!
The origional post has bee updated in a vague way notting that this is a dispute about money and not about being monitored without explaining.
/. (and neither do plow operators is seems).
And no-one has posted the explination, and it seems that no one in the newengland area who watches the news posts to
At issue is the pay drivers recieve will be tied to the GPS system. The gps system being used was shown to have issues tracking durring really heavy storms and when more than one device is really close together (one may drop off from the monitors standpoint).
Plows work the hardest in heavy storms, and also do alot of tandum work. One plow within 2 car lengths of the other.
The dispute is about their pay. Their pay would automatically be assigned from the GPS monotoring system. There is no proceedure for plow operators to dispute the GPS data. If the GPS system goes down, the older means of recieving pay would NOT go into effect, the plow operators would go UNPAID.
That was the dispute.
As for the public safty issues, there were plenty of other private plow operators who were happy to sign a temporary contract to plow for this weekends storm (With the GPS). The public safty issue was used by current contractors to pull for public support and to pressure the administration.
Everytime Maxtor releases a new 'largest' hard drive to the market, they always release the 5400, 2mb cache version first. I dont know if the reason if due to manufacturing or not. They have more plant lines producing the slower speed/less cache drives for OEM VAR's. Still I like to think its smart marketing, so all the techies will run out to buy the 'largest' drive at the highest price now.
When the 120's came out I waited a painfull 6 weeks (yes I'm a geek.. it was painful for me) for the 7200rpm version to be released. With the 160's I faithfully waited almost two months, and got a rebate to boot.
In the end its not if this 5400rpm drive will perform close to smaller 5400rpm drives, but if the latency and seek on a 300gb 7200rpm is small enough to warrent buying only one drive when ATA cards are cheap and two drives could get you better performance for more storage.
If you really need to cram multiple 300gb drives into any enclosure, speed is not what you are going after, its density, so its a moot point on that scale.
Voice Recognition is essential in hospitals and courts all over the US.
It is a multi-million dollar a year buisness.
As for the future of speech recognition, people often make the mistake of seeing it as a replacement for all other input interfaces. This is just absurd, as you point out.
More and more products (like cell phones, car GPS, ATM's, LARAN ) use speech recognition in conjunction with other more traditional means of input/output to get the job done.
It is NOT the fastest or the most reliable means of interfacing, but it is the most NATURAL.
Speaking as someone who works on a product with LNP grammars, we have them already.
(blatent plug, burning karma: http://www.scansoft.com/ )
Products like Voice Xpress, Dragon Naturally Speaking, and even IBM's Via Voice have had very powerfull NLP processing with full speech recognition for quite some time.
Other PIM products in the works have NLP processing on the server that is just amazing. Where these systems fall apart is with the actual speech recognition. Even the server processing systems at best get only 99.4% accuracy for adapted speech data.
NLP has come along way from even these humble beginnings as well. New technologies being developed for multimodal are starting to take shape. Currently this technology is only being used in the medical and transcription domains, but this is changing. On-star, many in-car GPS systems, and services that read you your e-mail and schedual from your cell phone are becomming more popular. The demand for good NLP grammar generation is high, and companies that have the engineers that know how to make a GOOD nlp grammar are limited.
In the end the gating issue is not the nlp grammar but the quality of speech audio data and the need for realtime recognition. Noisy car audio with a celular telephony feature set is very difficult to recognize with any accuracy, especially when the user wants an imediate responce. NLP inherently means more words are spoken. (i.e. 'where is the closest starbucks' verses 'find starbucks'.) More recongition errors + the flexability of an NLP grammar (verses a FSG) means things that are realy realy wrong will appear correct to the device.
Recognition is currently the bottleneck, not the NLP.
Heat will definatly be an issue, but much less power will be required. The majority of the power required by chips is used to push data on and off the chip. It takes alot of poser to move a signal from a 25 micron PCB path.
This technology (if it pans out) will mostlikly enter teh private sector in cell phones, DVD players and other small consumer electronics that have a very large number of units produced.
Silicon wafer production has always had one major problem. Impurities. The ability to use more of the waffer to produce smaller chips that can later be 'put back together' in arrays that may not be any larger than the origional single chip solution has the potential to be much cheaper to manufacture in mass quantity.
Granted this is part of the theory behind 6 Sigma, which does not always work out.
Gah, I read this one 5 times and still didnt understand. After reading all the links and doing some googling here it is again in a lower form of techno-geek (or would it be higher? not sure).
T4 is a broadcast Satalite used to transmit the raw station feeds from the central offices (read networks) to local broadcasters (cable, local ABC affiliate etc).
What Data T4 is responcible for:
Robust broadcast and syndication neighborhood anchored by ABC and CBS; host to SNG, data and distance learning applications. (Also hosted is Spice Channel and two other adult viewing stations which are not mentioned on the main site)
Where it effects:
The central portion of North America
The origional poster mentions the Carrier frequencies that the data is normally transfered on. The poster also mentions that there is no data being broadcast, just the main freq.
I have no way of confirming this myself, and dont see anything about it on any of the satalite pages. All they report is that a satalite is up and running (i.e. the Carrier frequency is present, but nothing on the data being transmitted)
As for the information on replacing the old T4 with T8, well T8 is currently handling south america, and the information I could find on google about the sale of the out dated satalites it vague at best.
Who does this effect:
Possably people out in the midwest are not getting any TV on their cable boxes.
mostlikly its effecting the cable companies and local affiliates who need to reroute to another satalite, and the central offices which also need to switch the feeds to other sat feed providers.
In the end nothing is confirmed.
The performance argument is a red herring. I would not go that far. As many posters have pointed out, its not really about which single language is better overall, but which languages are best suited for which applications. There is a reason why perl and python have a published C/API for extending and imbeding, there is a reason for JNI, blah, blah, blah. As with many things in life, the intended purpose of the article is not its best use. After reading the article it has changed my opinion on when and where to draw the line between the different languages. Many times I have planned from the start to do my network programming in C over C++ or Java due to the belief that they were slower or contained undue overhead. Which language has the best performance may be what the author intended to write an article about, in the end it is more a helpfull guide to help decide which side of the fence to write that piece of code.