I think the biggest problem with speech as an input for techies is that the software itself has not yet been written. While there may be recognition software that can comprehend speech at normal speed and append its dictionary as it runs, there's none that I know of that has been set up to function in a technical environment. It may be as simple as putting the pieces together, but it would probably require a lot of hacking on your own. The second biggest problem would be wearing out your voice, although that's something you can work with.
This sounds like an itch that somebody in such a situation would want to scratch. I think a lot of the basic speech recognition technology is out there. What would be really cool would be adapting it to an IDE environment. I can imagine setting it up so that it creates a dictionary of variables in use (like ctags), that you would teach it how to spell, then associate the spelling with a spoken word.
No, you don't understand. This makes is more expensive for other companies to try such a scheme, as they'd have to pay Philips royalties to implement it. And having to pay such a royalty would make those companies think things out more. Like: Hmm, if I pay $1 for this royalty, how much will I get back? Hmm, I don't think we'd get anything back; people would just walk away.
It's really just a common sense judgement, whether something is a signature or not: did the person mean it to act as a signature? That's the whole point of the decision.
Why do I get the feeling that the correct number is 1 million kg? I suspect that the original figure was 1 million kg, which then got translated to 2.2 million pounds for the non-scientific American audience. Then it got translated back to kilograms for the international audience, with some rounding errors. Even more interesting is that the number of significant digits went from 1 to 3 in the whole translation process.
The anonymous poster's link states the problem incorrectly. It's not PCs that can't be recovered, but Windows. And Microsoft has always made it difficult to quickly restore the OS. Things like GHOST exist, but are expensive. Microsoft has some build tools to automate the build process, but they take a lot more work than something like GHOST. I find this to be a serious flaw with Windows -- automated builds are too hard. (To be fair, I've not tried doing an automated build on Mac OS X. Most Linux variants make it really easy though.)
I'm having a hard time figuring out whether the Microsoft rep is trying to convince us to buy their new anti-malware product, or if they're making excuses for why their OS is so vulnerable. He is correct about social engineering being a serious problem though. Still, there are things that an OS should be doing to prevent rootkits. Like never allowing admin access without a password. And a lot of the rootkits exploit vulnerabilities that don't require any social engineering.
Definitely. VMware would be foolish not to file a lawsuit against Microsoft for leveraging its monopoly in operating systems to gain an advantage in virtualization software.
If you delete your emails after you receive a subpoena, you should go to jail. That's destroying evidence, and if you get caught, you probably WILL go to jail.
The trick a lot of companies do is to delete email BEFORE they get a subpoena. It's questionable if you know a subpoena is coming. So the new thing is to set a policy where all email must be deleted after 90 days or something. Plausible deniability. This is something that can generally be enforced if you handle your own email. If you have Google handle it, it's hard to be sure that they've really deleted it when you hit the delete button.
The articles of incorporation for most companies will include a statement that their purpose is to make a profit. The executives and board members have a fiuciary duty to ensure that they do their best to meet the stated purposes of the company.
If you have some evidence to the contrary, please present it.
In related news, web developers worldwide have given notice to Microsoft to fix Internet Explorer to actually conform to the standards it purports to support. Any web browser that is not fixed in 60 days will no longer work with many web sites.
The bigger problem is having for-profit insurance companies make health care decisions for people. Companies are required by law to maximize profits. These same companies then get to decide what level of medical care you get. If that's not unethical, I don't know what is.
Actually, I was thinking more about a media center (living room PC) than the media server (office PC) that the article was talking about. Sorry for not being clear about that.
You're right though -- the media server should normally just be doing disk and network I/O, that should not require too much CPU. But if transcoding does turn out to be necessary, you'd definitely want the second CPU on the media server.
If we're talking about a media center PC, then I'd definitely recommend a multi-core CPU. My poor little Mac mini has a hard time keeping up with decoding HDTV. Throw in another CPU-using process, and it starts to stagger pretty badly. Actually, HD access is also a big problem when there's a 2nd program doing a lot of disk reads/writes.
Another question we might ask is -- why has the Apache Foundation diversified so much? There are a lot of Apache projects out there now, many of which a lot of us are completely unfamiliar with. And some don't even have anything to do with the web. (SpamAssassin comes to mind. It's a great package, but seems to have no relation to Apache HTTPD.) What are the pros and cons of this diversity? Might it cause a lack of focus on the core HTTPD?
Universal binaries are not bloat. Sure, they may take up a bit more hard drive space. (What, 10 extra MB on your 200 GB hard drive? Who cares?) But the computer doesn't actually load any of the Intel code on a PowerPC processor.
What do you think of the newer smaller competitors? Particularly LightTPD. Would it be worthwhile for Apache to work to "slim down"? While Apache is seen by many as a slimmer alternative to more monolithic servers like IIS, it seems that it may be vulnerable to the even smaller alternatives. And this new round of competitors is nearly as featureful and extendable as Apache. What can Apache do to stay current and competitive?
Microsoft would do themselves a HUGE advantage if they diverged the corporate and home editions of Windows much more. (Yes, I'm aware that they recently merged the two with 2000 and XP. I'm talking release schedules more than core technologies.) Corporations want to move at a more conservative pace. Consumers want the latest and greatest new thing.
I think Microsoft should release a home version of Windows on a schedule of 12-18 months. Sort of use them as a beta cycle, similar to Red Hat's Fedora. (Sure, lots of people would still stick with Windows 98, but profits from the bleeding edge users would increase.) Then they can see what works well technology-wise, and include those portions in the corporate edition, which would have a 4-5 year cycle, closer to how companies amortize their computers.
I don't expect Microsoft to be able to do this though. They're still struggling to get into a modular mentality. And a good portion of their income comes from keeping corporations on the upgrade treadmill. But the way things have been working out, new versions of Windows have been taking 4-5 years anyway. Forcing themselves onto an 18-month home-user schedule would help them make better corporate editions.
If you've got more work scheduled than a single core can keep up with, then you DO NEED a second core. Or a faster CPU that can keep up with the workload. Also keep in mind that if you're playing back video and audio, those need to be done in real-time, or they will stutter. A second CPU core allows the main decoding task to be handled by one CPU, and other tasks can be run on the 2nd, preventing them from interfering with the primary task.
Re:VM Layer or legacy-free installs?
on
Why Windows is Slow
·
· Score: 2, Informative
GNU/Linux and UNIX systems solve this problem mostly with versioned libraries. (So does.NET, by the way.) When an API changes in a way that is not backwards-compatible, the library version is incremented. Programs that need the old API can keep using the old library, while new apps can use the new version. Granted, it takes up a little more memory if you've got 2 versions of the library loaded, but UNIX libraries tend to be broken into reasonably small pieces.
Kernel-level APIs can't be handled this way, but most programs don't directly access kernel calls. The standard C library (libc) handles most kernel-level access in a nice thin abstraction layer, which can handle kernel API changes.
I'm not sure exactly how Win32 works, but it seems from my point of view to be way too monolithic. So they don't have the option to upgrade a small portion of the API without creating a set of all new functions.
Thanks for the links. I read through the Wikipedia article. Very impressive. Sounds like it's pretty consistent with both General Relativity and Quantum Mechanics. Too bad he didn't have any of his stuff peer-reviewed. On the other hand, this probably freed him to concentrate on his theories more.
And I agree that this might be some experimental evidence of his theories.
Slashdot Mirror
No, you don't understand. This makes is more expensive for other companies to try such a scheme, as they'd have to pay Philips royalties to implement it. And having to pay such a royalty would make those companies think things out more. Like: Hmm, if I pay $1 for this royalty, how much will I get back? Hmm, I don't think we'd get anything back; people would just walk away.
It's really just a common sense judgement, whether something is a signature or not: did the person mean it to act as a signature? That's the whole point of the decision.
I didn't realize that there was a big demand for games based on soap operas. (I guess that's why the game was renamed in the US.)
Why do I get the feeling that the correct number is 1 million kg? I suspect that the original figure was 1 million kg, which then got translated to 2.2 million pounds for the non-scientific American audience. Then it got translated back to kilograms for the international audience, with some rounding errors. Even more interesting is that the number of significant digits went from 1 to 3 in the whole translation process.
I'm glad I'm not the only one who read it that way. On second thought, I'm kind of glad my mind is more interested in tall girls than call girls.
LibraryThing is pretty cool. It's a member-built database, but it links back to Amazon to get a lot of the details.
$23 per seat is not cheap, considering that I'm not even going to use it on a majority of the systems.
The anonymous poster's link states the problem incorrectly. It's not PCs that can't be recovered, but Windows. And Microsoft has always made it difficult to quickly restore the OS. Things like GHOST exist, but are expensive. Microsoft has some build tools to automate the build process, but they take a lot more work than something like GHOST. I find this to be a serious flaw with Windows -- automated builds are too hard. (To be fair, I've not tried doing an automated build on Mac OS X. Most Linux variants make it really easy though.)
I'm having a hard time figuring out whether the Microsoft rep is trying to convince us to buy their new anti-malware product, or if they're making excuses for why their OS is so vulnerable. He is correct about social engineering being a serious problem though. Still, there are things that an OS should be doing to prevent rootkits. Like never allowing admin access without a password. And a lot of the rootkits exploit vulnerabilities that don't require any social engineering.
Definitely. VMware would be foolish not to file a lawsuit against Microsoft for leveraging its monopoly in operating systems to gain an advantage in virtualization software.
If you delete your emails after you receive a subpoena, you should go to jail. That's destroying evidence, and if you get caught, you probably WILL go to jail.
The trick a lot of companies do is to delete email BEFORE they get a subpoena. It's questionable if you know a subpoena is coming. So the new thing is to set a policy where all email must be deleted after 90 days or something. Plausible deniability. This is something that can generally be enforced if you handle your own email. If you have Google handle it, it's hard to be sure that they've really deleted it when you hit the delete button.
The articles of incorporation for most companies will include a statement that their purpose is to make a profit. The executives and board members have a fiuciary duty to ensure that they do their best to meet the stated purposes of the company.
If you have some evidence to the contrary, please present it.
In related news, web developers worldwide have given notice to Microsoft to fix Internet Explorer to actually conform to the standards it purports to support. Any web browser that is not fixed in 60 days will no longer work with many web sites.
The bigger problem is having for-profit insurance companies make health care decisions for people. Companies are required by law to maximize profits. These same companies then get to decide what level of medical care you get. If that's not unethical, I don't know what is.
Decoding hi-def video will definitely peg your CPU. My single-PowerPC Mac mini can't even keep up with a full 1080i MPEG2 stream.
Actually, I was thinking more about a media center (living room PC) than the media server (office PC) that the article was talking about. Sorry for not being clear about that.
You're right though -- the media server should normally just be doing disk and network I/O, that should not require too much CPU. But if transcoding does turn out to be necessary, you'd definitely want the second CPU on the media server.
If we're talking about a media center PC, then I'd definitely recommend a multi-core CPU. My poor little Mac mini has a hard time keeping up with decoding HDTV. Throw in another CPU-using process, and it starts to stagger pretty badly. Actually, HD access is also a big problem when there's a 2nd program doing a lot of disk reads/writes.
Another question we might ask is -- why has the Apache Foundation diversified so much? There are a lot of Apache projects out there now, many of which a lot of us are completely unfamiliar with. And some don't even have anything to do with the web. (SpamAssassin comes to mind. It's a great package, but seems to have no relation to Apache HTTPD.) What are the pros and cons of this diversity? Might it cause a lack of focus on the core HTTPD?
Universal binaries are not bloat. Sure, they may take up a bit more hard drive space. (What, 10 extra MB on your 200 GB hard drive? Who cares?) But the computer doesn't actually load any of the Intel code on a PowerPC processor.
What do you think of the newer smaller competitors? Particularly LightTPD. Would it be worthwhile for Apache to work to "slim down"? While Apache is seen by many as a slimmer alternative to more monolithic servers like IIS, it seems that it may be vulnerable to the even smaller alternatives. And this new round of competitors is nearly as featureful and extendable as Apache. What can Apache do to stay current and competitive?
Microsoft would do themselves a HUGE advantage if they diverged the corporate and home editions of Windows much more. (Yes, I'm aware that they recently merged the two with 2000 and XP. I'm talking release schedules more than core technologies.) Corporations want to move at a more conservative pace. Consumers want the latest and greatest new thing.
I think Microsoft should release a home version of Windows on a schedule of 12-18 months. Sort of use them as a beta cycle, similar to Red Hat's Fedora. (Sure, lots of people would still stick with Windows 98, but profits from the bleeding edge users would increase.) Then they can see what works well technology-wise, and include those portions in the corporate edition, which would have a 4-5 year cycle, closer to how companies amortize their computers.
I don't expect Microsoft to be able to do this though. They're still struggling to get into a modular mentality. And a good portion of their income comes from keeping corporations on the upgrade treadmill. But the way things have been working out, new versions of Windows have been taking 4-5 years anyway. Forcing themselves onto an 18-month home-user schedule would help them make better corporate editions.
If you've got more work scheduled than a single core can keep up with, then you DO NEED a second core. Or a faster CPU that can keep up with the workload. Also keep in mind that if you're playing back video and audio, those need to be done in real-time, or they will stutter. A second CPU core allows the main decoding task to be handled by one CPU, and other tasks can be run on the 2nd, preventing them from interfering with the primary task.
GNU/Linux and UNIX systems solve this problem mostly with versioned libraries. (So does .NET, by the way.) When an API changes in a way that is not backwards-compatible, the library version is incremented. Programs that need the old API can keep using the old library, while new apps can use the new version. Granted, it takes up a little more memory if you've got 2 versions of the library loaded, but UNIX libraries tend to be broken into reasonably small pieces.
Kernel-level APIs can't be handled this way, but most programs don't directly access kernel calls. The standard C library (libc) handles most kernel-level access in a nice thin abstraction layer, which can handle kernel API changes.
I'm not sure exactly how Win32 works, but it seems from my point of view to be way too monolithic. So they don't have the option to upgrade a small portion of the API without creating a set of all new functions.
My favorite part of the email exchange is when the CentOS tech told Mr. Taylor:
If you will not let me help you, or at least talk to someone who knows
what Linux is, then you will look like an idiot.
Talk about predicting the future.
No, I'm Spartacus!
Thanks for the links. I read through the Wikipedia article. Very impressive. Sounds like it's pretty consistent with both General Relativity and Quantum Mechanics. Too bad he didn't have any of his stuff peer-reviewed. On the other hand, this probably freed him to concentrate on his theories more.
And I agree that this might be some experimental evidence of his theories.