No, it's very much alive. Just before I posted this story, I sent a similar e-mail to the list. BTW, there's currently a call going out for people to work on the OpenGraphics drivers.
However, I do worry that should Intel decide to put their graphics chip on a discrete PCI card it would eat up much of our potential market...
How come PDFs don't load properly? On some computers they load properly, not on mine. It crashes Firefox. Hard. PDFs great in IE though.
Home come the newest verison of Shockwave(or Flash?) doesn't work properly in Firefox? I've never seen an installation of Firefox that properly uses the newest version of Shockwave (Flash?). Again, works great in IE.
I'm sure you're aware that neither of these issues are anything that the Firefox team can do anything about. You should be addressing your complaints and bug reports to the developers of the plugins that implement those functions. That'd be Adobe, right?
What I sometimes do is write the performance-critical parts of my program in a low-level language, usually C or C++, and bind it with Python, and have pyGTK or pyWidgets for the GUI. It gets messy handling two languages, but doing a GUI in a high-level language like Python hugely reduces GUI development time. It can also be a little boring, if not tedious, to mix C++ and Python too...
There are 12 Windows workstations available for undergraduates, but all undergraduate teaching activites that involve or require the use of computers takes place on the Linux workstations (about 100-150 of them in the Design and Project Office). (This only applies to the Engineering department here; I here the mathematicians use Windows almost exclusively).
For example, can you believe that there are no good graphical FTP clients for linux? It's true. I have been using gFTP, which most people consider to be the best one, for about a week. It crashes almost daily, isn't very good option-wise, and it is soooo slow. I want something simple, say something like WSFTP for windows, and lo and behold it just doesn't exist. Seems remarkable that a good graphical FTP client doesn't even exist.
It's called Konqueror. It has these wonderful things called 'ioslaves' which can be accessed by special 'protocols'. Not only does it support browsing the local filesystem and the web, but also ftp sites (ftp://ftp.gnu.org), secure ftp (sftp://blah), info and man page viewing (info:/libc, man:/fstab), and a whole host of other resources. And it presents a unified and self-consistent interface to all of them.
There are fantastic GUI clients for FTP on Linux, and sometimes you don't realize you're already using one on a day to day basis (assuming you're a KDE user).
In other words you can use TPM as long as you provide the means to render it useless.
Excuse me? You have the mindset that "TPM == BAD". Actually, properly used it's a really good idea. Imagine your company has one server that handles all of your credit card transactions. You install a copy of the kernel that you trust (i.e. you compiled it yourself from kernel.org source code, perhaps) and sign it and its modules with your system's TPM key. Now the server will only boot a known-good kernel, leaving you proof against some types of rootkits. Next, you do the same for your Apache and database adaptor binaries. That would a raise a very effective barrier to either remote or local compromisation of the system.
The problem is that the big players want to use TPM in a totally different way: to remove your control over your hardware by selling you the hardware but not giving you the encryption keys you require to control it (see, for instance, my Tivo example in my GP post). And I agree, TPM used in the way Microsoft and the **AA want to is a very bad thing.
Like any powerful tool, TPM has the potential for abuse. That does not make it a bad tool.
I think you misunderstand the clause's purpose
on
Linus Speaks Out On GPLv3
·
· Score: 2, Interesting
When you get past the misinformation, errors and outright lies, trusted computing is not as bad as people think it is. It is a technology for enhancing security in a variety of environments. See the TPM thread a few postings down on the slashdot main page for some commentary there.
The GPLv3 as written does not forbid running software covered by it on a TPM system. What it says is that when a TPM platform vendor distributes GPLv3 software as binaries signed to run on their platform, they must not only provide the source code as in v2, but also the keys required to get modified versions of that particular software to run on the platform.
Example case studies:
The NSA decide they wish to use some GPLv3 software to run on their TPM-enabled computers. Their IT services department deploy the software throughtout the agency, signed with their TPM key so that it will run despite the TPM lockdown. They also place the source code to the software on their public website. They are not distributing TPM binaries outside their own organization, and therefore they do not need to include their TPM keys.
Tivo decide to use some GPLv3 software on their latest set-top box, signed so that it will run on their TPM platform. They sell the box to a large number of people. In order to be in compliance, not only must they distribute the source code, but because they are distributing the signed software & TPM platform outside their own organization they must also distribute the TPM key necessary to sign the binaries in order to get them to run.
A possible workaround is for the vendor to design a special subsystem that has application-specific keys that restrict the application to only carrying the restricted subset of low-level operations that it is supposed to. As long as the binaries the vendor distributes are signed with that key, that's the only key they need to distribute.
For instance, if Tivo were using a piece of GPLv3 software to process & display TV listings, they could use a key that allows the software to run on their platform but only to access the TV listings file and a pipe to send control signals down. They could then distribute that key with the source code and be in perfect compliance.
Of course, this wouldn't be very efficient except for fairly trivial user-space programs. It certainly wouldn't work for a kernel!
I'm getting paid for working on OSS this summer. And no, it's not a SoC project.
Whether or not software is free (as in freedom) and whether or not you get paid for working on it are entirely orthogonal issues. I think a lot of people fail to realize this.
In my limited experience, research labs for technology companies (like IBM, HP and Sun) employ a very diverse group of people from multiple disciplines.
Maybe not as diverse as you'd think. When I was working in Sharp's research labs in the UK, the team I was in (about 15 people) was almost entirely engineers, and eight or nine of us were either Oxford or Cambridge graduates or starting at those universities the next year.
Having said that, the other five people came from some really suprising backgrounds (bioengineering in an electronics lab?), but all had scientific Masters or Doctorate degrees.
Sharp's various R&D laboratories still do an awful lot of pretty cool stuff. I spent my gap year working at Sharp Labs of Europe, and had lots of fun.
Sharp have a R&D lab in the USA as well.
Check it out -- I imagine many corporations still have similar dedicated and well-funded research efforts.
Please look at what you write. Which of ships, cars or airplanes got 0.7% of a nations budget? In an infantile stage, because of reasons like "The human race needs to go to the moon"?
I remember when Columbus said, "I want to cross the Atlantic Ocean," and he couldn't get funding for it because going east around the south of Africa to get to India was just fine, thank you. And gosh, didn't he discover something unexpected and enormously profitable for the sponsors he eventually found?
What ever happened to our adventurous explorers? When was it that it became verboten to look to the frontiers and want to go and see it for yourself?
We may be required to disclose user information pursuant to lawful requests, such as subpoenas or court orders, or in compliance with applicable laws. We do not reveal information until we have a good faith belief that an information request by law enforcement or private litigants meets applicable legal standards. Additionally, we may share account or other information when we believe it is necessary to comply with law, to protect our interests or property, to prevent fraud or other illegal activity perpetrated through the Facebook service or using the Facebook name, or to prevent imminent bodily harm. This may include sharing information with other companies, lawyers, agents or government agencies.
I don't think that it would be unreasonable for a court of law to rule that Facebook should comply with its published privacy policy.
As other people have commented, it sounds like they used an intern to grab the data.
As far as invoking the Patriot Act is concerned, if they invoked that to check up on a job interviewee, it's pretty piss-poor operational security to tell a someone to their face that they're a terrorism suspect -- and if the person involved is not suspected terrorist, it's misuse of the Patriot Act -- isn't that a serious offence?
I'm not sure which agency carries out Top Secret clearance checks in the US, but in the UK, all you get told is whether you passed or failed the vetting. No interview, no explanation, no appeal. They certainly don't tell you how they got the data -- that's just an invitation for unfriendly parties to 'sploit the vetting system.
The premise of my joke was that the DOD had directed its contractors to develop missile defense systems under the GPL free software license, and was then obligated as a customer of those contractors to release the code.
Uh... no. -1 Wrong, because with the GPL you are only obligated to distribute source code when you distribute binaries, and then only to the people you distributed said binaries to.
So in your hypothetical scenario, the contractors would be obligated to send a copy of the source code to whichever agency is responsible for loading the software binaries into the missile systems. Oh, wait... that's the DoD, isn't it?
About 60% of these are un-numbered symlinks to libraries with version numbers appended to the filename. Almost everything is (redundantly) prefixed with "lib".
These are automatically generated by a wonderful thing called 'libtool' which allows you to have multiple versions of the same library installed at the same time, while your applications magically use the one they need. The 'lib' prefix is added by libtool, IIRC.
The contents of/lib/,/usr/lib/ and/usr/local/lib aren't meant to be usable by humans. If you want to know what libraries & versions you have installed, use ldconfig -v. Nowadays, you don't often need to know unless you're building software, in which case there are lots of tools that build configuration scripts can use to work out which versions of libraries you have and tell you if you need newer ones: autoconf, pkgconfig, etc.
I think libtool is a remarkably clever and flexible system. FIf you can think of a better way of handling it, feel free to disagree with me. Don't forget, Windows doesn't even try: every application brings along its own copies of the libraries it uses...
Well, the USRP contains an Altera Cyclone 12 FPGA, which contains 12000 LEs and 240 kbits of RAM.
OGD1 has two FPGAs, a small Lattice XP6 that implements the PCI bridge, and a Lattice ECP2-50 for the actual board function. The latter contains 48000 LEs and 387 kbits of RAM, as well as seventy-two 18-bit multipliers. OGD1 is a lot more powerful than the USRP motherboard, and it's on the PCI bus rather than USB, so it has a lot lower CPU latency as well as a much higher bandwidth.
On the other hand, it's going to be about twice the price.
The term native would imply that it's a complete port to the X Window System. This clearly is not. Whether or not an individual cares about the API details isn't relevant to this thread. The point the original poster was making was simply this is not a "native" port.
Which is daft. It's written using Qt and OpenGL: I really fail to understand what makes a native Linux port so difficult!
What would end the argument, Bruce is open-source hardware.
We're working on it. The OpenGraphics project is working on an open-architecture GPU which will have BSD-licensed drivers, and GPL'd board schematics and artwork.
There's nothing stopping another group of hackers setting up similar projects: OpenWireless, OpenSATARaid,...
Especially since the OpenGraphics project will be bringing out an PCI card with a big FPGA on it soon (OGD1). Although it'll be primarily aimed at development of the OGA graphics pipeline, it's got a big header on it, so there's no reason it couldn't be used for something else. Accelerating POVRay, perhaps?
What's to guarantee that the person a company finds on Myspace or Livejournal - I don't know much about Facebook - is the same person they're actually considering employing?
I don't mind, if people Google for me they come to the conclusion that I own this company.
As one of the results of our K3M multimedia meeting in the Netherlands, your favourite media player is now called "Amarok". [some reasons] 4) Amarok is intended to be a software for all desktops, not just KDE. The capital K suggests that it's a KDE only application. [emphasis added]
Also, I like the way you show your appreciation for their work on a media player you obviously like by calling them "stupid". I'm sure it would make them feel really good about sharing their work with you.
...why, when something goes wrong in an organization, does the head of organization get called on to resign, when 90% of the time the incident didn't have anything to do with negligence or error on their part?
Personally, I think there should be three distributions: GNOME, KDE, and one for power-users to tinker around with. It doesn't make sense to have five distributions that use the GNOME "window environment" and all happen to look and behave alike because they all use GNOME. Why have 5?
Simple: so that you have a choice. Beneath the desktop environment there are many, many different ways of doing things. runit vs init, yum vs apt-get, etc, etc.
Not to mention that a distribution targetting ARM-based handheld systems is going to be totally different to one targetting tablets, which is in turn going to be very different to a distribution optimised for high-stability 99.99% uptime clustered servers.
Trying to make something that works for everyone will inevitably end up as something that isn't good enough for anyone.
Slashdot Mirror
No, it's very much alive. Just before I posted this story, I sent a similar e-mail to the list. BTW, there's currently a call going out for people to work on the OpenGraphics drivers.
However, I do worry that should Intel decide to put their graphics chip on a discrete PCI card it would eat up much of our potential market...
I'm sure you're aware that neither of these issues are anything that the Firefox team can do anything about. You should be addressing your complaints and bug reports to the developers of the plugins that implement those functions. That'd be Adobe, right?
Check out PyQT.
There are 12 Windows workstations available for undergraduates, but all undergraduate teaching activites that involve or require the use of computers takes place on the Linux workstations (about 100-150 of them in the Design and Project Office). (This only applies to the Engineering department here; I here the mathematicians use Windows almost exclusively).
Gosh, where I am must be really unusual, then.
Here, the undergraduate teaching is all done on Linux boxes running a customized version of Knoppix that includes all the important apps:
And every undergraduate gets a copy.
Not to mention that out of eight people in my research group, five use Linux out of choice.
It's called Konqueror. It has these wonderful things called 'ioslaves' which can be accessed by special 'protocols'. Not only does it support browsing the local filesystem and the web, but also ftp sites (ftp://ftp.gnu.org), secure ftp (sftp://blah), info and man page viewing (info:/libc, man:/fstab), and a whole host of other resources. And it presents a unified and self-consistent interface to all of them.
There are fantastic GUI clients for FTP on Linux, and sometimes you don't realize you're already using one on a day to day basis (assuming you're a KDE user).
...but in the UK we use highly precise and effective vote collection machinery, consisting of:
And it works perfectly well, and is completely anonymous and auditable.
Excuse me? You have the mindset that "TPM == BAD". Actually, properly used it's a really good idea. Imagine your company has one server that handles all of your credit card transactions. You install a copy of the kernel that you trust (i.e. you compiled it yourself from kernel.org source code, perhaps) and sign it and its modules with your system's TPM key. Now the server will only boot a known-good kernel, leaving you proof against some types of rootkits. Next, you do the same for your Apache and database adaptor binaries. That would a raise a very effective barrier to either remote or local compromisation of the system.
The problem is that the big players want to use TPM in a totally different way: to remove your control over your hardware by selling you the hardware but not giving you the encryption keys you require to control it (see, for instance, my Tivo example in my GP post). And I agree, TPM used in the way Microsoft and the **AA want to is a very bad thing.
Like any powerful tool, TPM has the potential for abuse. That does not make it a bad tool.
The GPLv3 as written does not forbid running software covered by it on a TPM system. What it says is that when a TPM platform vendor distributes GPLv3 software as binaries signed to run on their platform, they must not only provide the source code as in v2, but also the keys required to get modified versions of that particular software to run on the platform.
Example case studies:
A possible workaround is for the vendor to design a special subsystem that has application-specific keys that restrict the application to only carrying the restricted subset of low-level operations that it is supposed to. As long as the binaries the vendor distributes are signed with that key, that's the only key they need to distribute.
For instance, if Tivo were using a piece of GPLv3 software to process & display TV listings, they could use a key that allows the software to run on their platform but only to access the TV listings file and a pipe to send control signals down. They could then distribute that key with the source code and be in perfect compliance.
Of course, this wouldn't be very efficient except for fairly trivial user-space programs. It certainly wouldn't work for a kernel!
I'm getting paid for working on OSS this summer. And no, it's not a SoC project.
Whether or not software is free (as in freedom) and whether or not you get paid for working on it are entirely orthogonal issues. I think a lot of people fail to realize this.
Maybe not as diverse as you'd think. When I was working in Sharp's research labs in the UK, the team I was in (about 15 people) was almost entirely engineers, and eight or nine of us were either Oxford or Cambridge graduates or starting at those universities the next year.
Having said that, the other five people came from some really suprising backgrounds (bioengineering in an electronics lab?), but all had scientific Masters or Doctorate degrees.
Sharp's various R&D laboratories still do an awful lot of pretty cool stuff. I spent my gap year working at Sharp Labs of Europe, and had lots of fun. Sharp have a R&D lab in the USA as well. Check it out -- I imagine many corporations still have similar dedicated and well-funded research efforts.
I am an Engineering student, and I resent your anonymous, cowardly and entirely unsubstantiated remark.
I'm sure it's only you who find optimistic and naive people annoying!
I remember when Columbus said, "I want to cross the Atlantic Ocean," and he couldn't get funding for it because going east around the south of Africa to get to India was just fine, thank you. And gosh, didn't he discover something unexpected and enormously profitable for the sponsors he eventually found?
What ever happened to our adventurous explorers? When was it that it became verboten to look to the frontiers and want to go and see it for yourself?
The relevant part Facebook's privacy policy:
I don't think that it would be unreasonable for a court of law to rule that Facebook should comply with its published privacy policy.
As other people have commented, it sounds like they used an intern to grab the data.
As far as invoking the Patriot Act is concerned, if they invoked that to check up on a job interviewee, it's pretty piss-poor operational security to tell a someone to their face that they're a terrorism suspect -- and if the person involved is not suspected terrorist, it's misuse of the Patriot Act -- isn't that a serious offence?
I'm not sure which agency carries out Top Secret clearance checks in the US, but in the UK, all you get told is whether you passed or failed the vetting. No interview, no explanation, no appeal. They certainly don't tell you how they got the data -- that's just an invitation for unfriendly parties to 'sploit the vetting system.
Uh... no. -1 Wrong, because with the GPL you are only obligated to distribute source code when you distribute binaries, and then only to the people you distributed said binaries to.
So in your hypothetical scenario, the contractors would be obligated to send a copy of the source code to whichever agency is responsible for loading the software binaries into the missile systems. Oh, wait... that's the DoD, isn't it?
These are automatically generated by a wonderful thing called 'libtool' which allows you to have multiple versions of the same library installed at the same time, while your applications magically use the one they need. The 'lib' prefix is added by libtool, IIRC.
The contents of /lib/, /usr/lib/ and /usr/local/lib aren't meant to be usable by humans. If you want to know what libraries & versions you have installed, use ldconfig -v. Nowadays, you don't often need to know unless you're building software, in which case there are lots of tools that build configuration scripts can use to work out which versions of libraries you have and tell you if you need newer ones: autoconf, pkgconfig, etc.
I think libtool is a remarkably clever and flexible system. FIf you can think of a better way of handling it, feel free to disagree with me. Don't forget, Windows doesn't even try: every application brings along its own copies of the libraries it uses...
Well, the USRP contains an Altera Cyclone 12 FPGA, which contains 12000 LEs and 240 kbits of RAM.
OGD1 has two FPGAs, a small Lattice XP6 that implements the PCI bridge, and a Lattice ECP2-50 for the actual board function. The latter contains 48000 LEs and 387 kbits of RAM, as well as seventy-two 18-bit multipliers. OGD1 is a lot more powerful than the USRP motherboard, and it's on the PCI bus rather than USB, so it has a lot lower CPU latency as well as a much higher bandwidth.
On the other hand, it's going to be about twice the price.
Which is daft. It's written using Qt and OpenGL: I really fail to understand what makes a native Linux port so difficult!
We're working on it. The OpenGraphics project is working on an open-architecture GPU which will have BSD-licensed drivers, and GPL'd board schematics and artwork.
There's nothing stopping another group of hackers setting up similar projects: OpenWireless, OpenSATARaid, ...
Especially since the OpenGraphics project will be bringing out an PCI card with a big FPGA on it soon (OGD1). Although it'll be primarily aimed at development of the OGA graphics pipeline, it's got a big header on it, so there's no reason it couldn't be used for something else. Accelerating POVRay, perhaps?
Having used both K3B and Nero, I am firmly of the opinion that K3B is better.
I'll bite. My Fedora Core 4 x86-64 box works absolutely perfectly for me, and identically to the FC4 i386 box that sits next to it.
More lies, surely? Mark Kretschmann, one of the Amarok leads, recently wrote:
Also, I like the way you show your appreciation for their work on a media player you obviously like by calling them "stupid". I'm sure it would make them feel really good about sharing their work with you.
...why, when something goes wrong in an organization, does the head of organization get called on to resign, when 90% of the time the incident didn't have anything to do with negligence or error on their part?
Can someone please explain for me?
Simple: so that you have a choice. Beneath the desktop environment there are many, many different ways of doing things. runit vs init, yum vs apt-get, etc, etc.
Not to mention that a distribution targetting ARM-based handheld systems is going to be totally different to one targetting tablets, which is in turn going to be very different to a distribution optimised for high-stability 99.99% uptime clustered servers.
Trying to make something that works for everyone will inevitably end up as something that isn't good enough for anyone.