First, the GP is right: most high power white LEDs are actually blue or UV LEDs with a yellow phosphor in the plastic packaging.
As for phosphors yielding a spiky mess for a spectrum: how exactly do you imagine the spectrum of an RGB LED looks? The individual primaries in such combinations are VERY narrow band, so rather than a continuous spectrum you get three distinct peaks. Phosphors are actually smoother by comparison.
Well, he does set the agenda for the meeting, so if he is serious, he could dedicate a fairly large chunk of the meeting time to this topic.
Now, theoretically the rest of them could just tell him off on this topic. But lets face it: they all need to justify being at a meeting that took a huge effort to organize with thousands of cops protecting them, tens of thousands of people demonstrating etc. In other words, they need a result. ANY result. Internet regulation is an easy topic in that respect, since the wish for some amount of control over the internet is widespread among politicians of many countries. So the meeting dynamics could very easily be stacked against common sense and individual right.
Just to add to that: NSERC specifically allows for universities to retain copyright of materials produced under NSERC grants. So yeah, if you take their money, the university can take your rights.
Even more than that, if the guy did any of his work under direct supervision of a faculty member or staff, the very fact that the supervisor was paid by the university while contributing time towards the project, allows the university to also claim at least part of the ownership.
The use of the term "free" by the FSF reminds me less and less of "free as in speech", and increasingly summons associations with "freedom fries" and "operation enduring freedom":
"We define what it means to be free, and you better agree, or else..."
That is exactly right. If any of the guys had access to code/algorithms that the old company considers trade secrets, then they are in a world of pain if they replicate that.
If it comes to a lawsuit, their old employer will most likely subpoena them to provide their new code to an independent expert for code comparison. Again, any hint of copying even small pieces (or recoding algorithms considered trade secrets from memory), and they are done for.
If they absolutely must stay in the same area they already work in, then they should hire some outside person, who performs a cleanroom reverse-engineering of the original product. Make sure you clearly document that there were no leaks of inside knowledge (keep lab notebooks etc.).
More seriously though, it is probably not worth even trying, since the GPUs used in the Tesla support a very limited model of parallelism. Shoehorning the flexibility of Erlang into that would at the very leas result in a dramatic performance loss, if it is possible at all.
You'd have to be a real intellectual lightweight to ponder this question "for decades" before realizing you can just search a dictionary. It boggles the mind.
The thing is, though, it is not just about root on machines you control. As soon as somebody has access to an ethernet port with an NFSed machine, he can unplug that machine, hook up his own notebook while configuring it to have the same MAC and/or IP address, and - voila - he can mount whatever he wants.
As for normal users wanting sudo/root, I think there is a clear trend for people to want to be able to install their own software, so this is definitely getting more widespread, especially in academic environments, but also in development groups.
Now I am not saying that NFS has no role. If you can trust everybody with access to any of the supported machines and any of the networking hardware, then fine. But you do have to be careful if you have any kind of sensitive information, such as grades on a university server where students can access the network.
...But it's simple and useful enough for a small and/or development network - and works better than SMB/CIFS when dealing with Unix to Unix and permissions.
In other words, it is useful when you trust all people that have potentially access not only to your systems, but also to your ethernet ports etc. That's fine, just don't use it for accessing a partition with my SIN or health information.
Unfortunately, NFS is not safe since it trusts clients. If users need to have root or sudo on their individual machines, they can go out and read any file on the server (well, technically partition, but who has one partition per user on their server?). NFS comes from a time of big iron servers where no end user EVER had root access. The world has changed.
CIFS/SMB may be slow, but at least it got the per-user authentication right. If you want an alternative, something like the Andrew File System (the other AFS), or OpenAFS would be better. OpenAFS exists for Macs.
Well, as I understand it (and as the summary states), it is a preliminary injunction. Decisions made in such injunctions are often questionable, since they do not include full fact-checking, which is left to the main court proceedings. I am sure similar things could happen here (Canada), or in the US, although I would certainly hope that the final ruling would reverse the decision.
Look, I was all out supporting Obama, especially considering the alternatives. But can we PLEASE stop comparing him to Kennedy and Lincoln before he has even started?
Obviously, most people would have to change their mindset before it could happen.
Well, that is exactly what the proponents of communism say. If we have learned anything from that, it is that people don't change their mindset, and a political system needs to be robust against that.
Because, just like US voters, they have no idea what's just been done.
What's just been done, is that the lying, thieving, war-mongering, and incompetent people in charge for the mess over the past 8 years have been thoroughly ousted.
I am sure people will point out that McCain!= Bush, and I will admit that McCain himself seems to be a man of integrity. However, much of the republican leadership is not. Palin serves as a perfect example. By putting unflinching loyalty over competence in her Alaskan kingdom, she shows exactly the same kind of cronyism as the Bush administration. Having McCain lead that kind of team would have been like gold-plating a turd.
It is true that the the new guys still have to prove they are any good, but we know the old guys were not, so at least there is a chance of improvement now.
Sure, but what is your point? Motion detectors and heat sensors can be fooled as well. Motion sensors by moving very slowly and deliberately. Heat sensors by wearing well insulated clothing. And that is just the simple stuff you do before you hit the high-tech workarounds that you see in any heist movie. That doesn't mean the simple home security stuff is worthless - it can help protect you against a junkie trying to score money for the next fix and other crimes of opportunity. Sure, if you have enough valuables to worry about professional criminals, you need more professional security. Most people's homes don't look like the Louvre, though.
People need to understand that security isn't a binary state, it is a continuous scale. Security professionals will never claim that anything is completely safe, they always quantify safety by the time (and other resources) it takes to circumnavigate the security system. This ought to be true for computers as well, BTW.
OpenGL ES is a subset of OpenGL. For 2.0, they removed much of the fixed function pipeline and replaced it with GLSL (OpenGL's shading language). As such, I would expect it to be relatively straightforward to write ES code that runs on a modern OpenGL desktop (one that supports the latest GLSL). At most you'll need a few #ifdefs.
WTF? Tel Aviv University is a very decent research institute that has made many important contributions to science. No, I am not and have never been affiliated with them, but the page you are referring to is obviously that of an alumni organization. And yes, they do raise money for the university, that is what alumni organizations do.
As for Yaroslavsky (the prof working on this "seeing skin" project), I know neither him nor this project (at least not more than the press release states), but his publication list shows that he regularly publishes in top journals such as Applied Optics, Optics Express, and Optics Letters. Clearly he knows a thing or two about light.
No it doesn't. It makes it an infringement to, without license, write or use software that is based on the specific analysis algorithm IBM is attempting to patent.
Slashdot Mirror
First, the GP is right: most high power white LEDs are actually blue or UV LEDs with a yellow phosphor in the plastic packaging.
As for phosphors yielding a spiky mess for a spectrum: how exactly do you imagine the spectrum of an RGB LED looks? The individual primaries in such combinations are VERY narrow band, so rather than a continuous spectrum you get three distinct peaks. Phosphors are actually smoother by comparison.
Well, he does set the agenda for the meeting, so if he is serious, he could dedicate a fairly large chunk of the meeting time to this topic.
Now, theoretically the rest of them could just tell him off on this topic. But lets face it: they all need to justify being at a meeting that took a huge effort to organize with thousands of cops protecting them, tens of thousands of people demonstrating etc. In other words, they need a result. ANY result. Internet regulation is an easy topic in that respect, since the wish for some amount of control over the internet is widespread among politicians of many countries. So the meeting dynamics could very easily be stacked against common sense and individual right.
But what else is new?
Just to add to that: NSERC specifically allows for universities to retain copyright of materials produced under NSERC grants. So yeah, if you take their money, the university can take your rights.
Even more than that, if the guy did any of his work under direct supervision of a faculty member or staff, the very fact that the supervisor was paid by the university while contributing time towards the project, allows the university to also claim at least part of the ownership.
Thank you.
The use of the term "free" by the FSF reminds me less and less of "free as in speech", and increasingly summons associations with "freedom fries" and "operation enduring freedom":
"We define what it means to be free, and you better agree, or else..."
That is exactly right. If any of the guys had access to code/algorithms that the old company considers trade secrets, then they are in a world of pain if they replicate that.
If it comes to a lawsuit, their old employer will most likely subpoena them to provide their new code to an independent expert for code comparison. Again, any hint of copying even small pieces (or recoding algorithms considered trade secrets from memory), and they are done for.
If they absolutely must stay in the same area they already work in, then they should hire some outside person, who performs a cleanroom reverse-engineering of the original product. Make sure you clearly document that there were no leaks of inside knowledge (keep lab notebooks etc.).
You can improve signal-to-noise ratio two ways: boost signal strength, or reduce noise level.
Either works for me.
By writing an Erlang-to-CUDA compiler?
More seriously though, it is probably not worth even trying, since the GPUs used in the Tesla support a very limited model of parallelism. Shoehorning the flexibility of Erlang into that would at the very leas result in a dramatic performance loss, if it is possible at all.
Yeah, it is that easy.
You'd have to be a real intellectual lightweight to ponder this question "for decades" before realizing you can just search a dictionary. It boggles the mind.
Wait, I thought not-for-profits aren't allowed to run a surplus? What am I missing?
The thing is, though, it is not just about root on machines you control. As soon as somebody has access to an ethernet port with an NFSed machine, he can unplug that machine, hook up his own notebook while configuring it to have the same MAC and/or IP address, and - voila - he can mount whatever he wants.
As for normal users wanting sudo/root, I think there is a clear trend for people to want to be able to install their own software, so this is definitely getting more widespread, especially in academic environments, but also in development groups.
Now I am not saying that NFS has no role. If you can trust everybody with access to any of the supported machines and any of the networking hardware, then fine. But you do have to be careful if you have any kind of sensitive information, such as grades on a university server where students can access the network.
Ever hear of the NFS options to squash root access? It'll map to the user 'nobody' if you do it right. Presto, instant client root limitations.
See my reply to the AC: http://slashdot.org/comments.pl?sid=1032335&cid=25793973.
...But it's simple and useful enough for a small and/or development network - and works better than SMB/CIFS when dealing with Unix to Unix and permissions.
In other words, it is useful when you trust all people that have potentially access not only to your systems, but also to your ethernet ports etc. That's fine, just don't use it for accessing a partition with my SIN or health information.
True, but this always works:
> su # to become root
> su otheruser # to become otheruser (no password required, since you are root)
> cd ~otheruser # access otheruser's files
Unfortunately, NFS is not safe since it trusts clients. If users need to have root or sudo on their individual machines, they can go out and read any file on the server (well, technically partition, but who has one partition per user on their server?). NFS comes from a time of big iron servers where no end user EVER had root access. The world has changed.
CIFS/SMB may be slow, but at least it got the per-user authentication right. If you want an alternative, something like the Andrew File System (the other AFS), or OpenAFS would be better. OpenAFS exists for Macs.
Actually, a police raid sounds loads more interesting than the crap they have been showing in theaters lately ;-)
Well, as I understand it (and as the summary states), it is a preliminary injunction. Decisions made in such injunctions are often questionable, since they do not include full fact-checking, which is left to the main court proceedings. I am sure similar things could happen here (Canada), or in the US, although I would certainly hope that the final ruling would reverse the decision.
Look, I was all out supporting Obama, especially considering the alternatives. But can we PLEASE stop comparing him to Kennedy and Lincoln before he has even started?
Paper isn't exactly ideal for keeping things sterile. Are you typing in your handwritten notes later? Are you replacing the printouts regularly?
Why not just use ceran wrap on the keyboard (change frequently)?
Obviously, most people would have to change their mindset before it could happen.
Well, that is exactly what the proponents of communism say. If we have learned anything from that, it is that people don't change their mindset, and a political system needs to be robust against that.
Whoa. Thanks for that!
Because, just like US voters, they have no idea what's just been done.
What's just been done, is that the lying, thieving, war-mongering, and incompetent people in charge for the mess over the past 8 years have been thoroughly ousted.
I am sure people will point out that McCain!= Bush, and I will admit that McCain himself seems to be a man of integrity. However, much of the republican leadership is not. Palin serves as a perfect example. By putting unflinching loyalty over competence in her Alaskan kingdom, she shows exactly the same kind of cronyism as the Bush administration. Having McCain lead that kind of team would have been like gold-plating a turd.
It is true that the the new guys still have to prove they are any good, but we know the old guys were not, so at least there is a chance of improvement now.
Sure, but what is your point? Motion detectors and heat sensors can be fooled as well. Motion sensors by moving very slowly and deliberately. Heat sensors by wearing well insulated clothing. And that is just the simple stuff you do before you hit the high-tech workarounds that you see in any heist movie. That doesn't mean the simple home security stuff is worthless - it can help protect you against a junkie trying to score money for the next fix and other crimes of opportunity. Sure, if you have enough valuables to worry about professional criminals, you need more professional security. Most people's homes don't look like the Louvre, though.
People need to understand that security isn't a binary state, it is a continuous scale. Security professionals will never claim that anything is completely safe, they always quantify safety by the time (and other resources) it takes to circumnavigate the security system. This ought to be true for computers as well, BTW.
I wonder how much of that is just because she was frowning in the original vs. smiling in the result.
OpenGL ES is a subset of OpenGL. For 2.0, they removed much of the fixed function pipeline and replaced it with GLSL (OpenGL's shading language). As such, I would expect it to be relatively straightforward to write ES code that runs on a modern OpenGL desktop (one that supports the latest GLSL). At most you'll need a few #ifdefs.
WTF? Tel Aviv University is a very decent research institute that has made many important contributions to science. No, I am not and have never been affiliated with them, but the page you are referring to is obviously that of an alumni organization. And yes, they do raise money for the university, that is what alumni organizations do.
As for Yaroslavsky (the prof working on this "seeing skin" project), I know neither him nor this project (at least not more than the press release states), but his publication list shows that he regularly publishes in top journals such as Applied Optics, Optics Express, and Optics Letters. Clearly he knows a thing or two about light.
http://www.eng.tau.ac.il/~yaro/RecentPublications/index.html
No it doesn't. It makes it an infringement to, without license, write or use software that is based on the specific analysis algorithm IBM is attempting to patent.
Nice FUD, though.