PGP has a much easier task, though: it only needs to ensure that people with the key can decrypt content, while people without the key cannot. DRM schemes need to ensure that the same person can only decrypt given content for certain purposes, and not for other purposes.
Maybe it's a nitpick, but the headline "BBC's Plan To Kick Open Source Out of UK TV" to me sounds like someone is against open-source software, and has conjured up a scheme, the primary purpose of which is to harm it.
From the article, though, it seems more likely that the BBC is worried about copyright infringement, and as with many companies, the only sort-of-half-assed solution they can think of to combat it is to introduce some DRM, and the only even-more-half-assed solution they can think of to make it hard to crack the DRM is security-through-obscurity. That's incompatible with OSS, as Cory Doctorow points out, but I think out of a misplaced attempt to use security-through-obscurity, not out of an actual antipathy to open-source vs. proprietary software as licensing models. Who knows if they even realized that: 1) lots of open-source software is used in conjunction with receiving TV broadcasts (and not just by warez groups); and 2) their scheme would therefore harm an important segment of the public.
What's the "true profit motive in a free market" for space exploration and colonization? Mining? Hotels?
Or are you suggesting government bounties of some sort? Because artificial incentives ponied up by government with taxpayer money hardly count as true profit motives in a free market, though they may in some cases be reasonable ways for the government to get something done.
It seems to vary, like the throughput does, and roughly correlated with it. For example, AT&T latency seems better at all the locations in Atlanta. It does have some absolutely terrible ones in some cities, though (700ms+ latencies).
It doesn't really, though. OpenCL "decides" based on some very high-level, high-granularity features of devices it can enumerate. In practice, if you want your code to run reasonably well, you know which parts are going to run on the GPU and which on the CPU. OpenCL isn't an auto-parallelization solution, just a set of primitives for parallel programming--- more like an MPI or OpenMP that also supports GPGPU than the old 70s holy grain of auto-parallelizing where the compiler or runtime magically figures out how to chunk up your computation and where to send the chunks.
As far as I can tell, OpenCL is pretty much based on CUDA, not on an attempt to unify CUDA and CTM/Brook+/FireStream. That's partly because ATI's solutions never really caught on, and have been sorta ignored.
That seemed to be the consensus of the Slashdot discussion after their sudden announcement that they weren't going to pay people to make it after all, but wanted proposals for someone willing to do it for free. Did they actually find someone willing to do it for free? Or was funding restored?
Those laws are widely expected to be struck down as a violation of the First Amendment in the next month or two, though. Cases like this seem to be one of the better arguments for the "strike them down" side. I'm not a big fan of corporate money influencing politics, but throwing someone in jail because they're a "former republican party official" with an anti-Congressman blog seems pretty repugnant to free-speech principles.
All Mr Grayson is doing is what Democrats should have been doing years ago. Republicans want to talk about death panels and pulling the plug on grandma? Fine, they should be prepared to listen to the other side using the same kind of emotional language.
I typically vote for Democrats rather than Republicans because they don't do the same hyperbolic bullshit, or at least don't do it nearly as often. If they're going to start pulling this crap, then I'll henceforth start filing "voting for a Democrat" in the same category as "voting for Sarah Palin".
Lying in a political context is hard to prosecute, unless it rises to the level of libel, which has a pretty high bar for public officials, and an even higher bar for political speech about public officials.
Lying on forms filed with the government is illegal, though, under a blanket "don't lie to the government" law. The jail part of the complaint seems to be for allegedly misrepresenting the PAC on the filing documents with the FEC: the filed documents claim the PAC isn't aimed at any particular opponent, but the website clearly is aimed at one opponent.
IRB is more weird than lax in my experience. I've seen hassles over a protocol that asked adults to use some experimental software, asking for some BS to qualm the IRB's fears about whether it could cause any psychological harm. Perhaps thrown-together research-software GUI is so horrible someone will be traumatized, but it seems unlikely that any possible such experiment will really rise to the level of risk that even low-risk medical experiments have.
Tax software is hardly simple, though I agree it is on the technical side. Not only is there no "internationally agreed XML schema for the taxman to publish this years' tax legislation", but there is no formal record of this year's tax schema at all. The tax laws are riddled with vaguely worded provisions, some of which (in the US) have been clarified by IRS advisory opinions, tax court precedent, or regular court precedent. You basically need a team of lawyers from every jurisdiction to even figure out what the tax code is. Encoding it in software is the relatively easy part.
Oh, and there are constant updates to it, some of which come out so late that patches ideally should be out 1-2 days after the new regulation is announced. Who is the volunteer who is going to spend their life monitoring the IRS's press releases and patching tax software?
There's lots of web-based stuff too, which is increasingly being offered as a part of a package with other services. Payroll processors are common, for example: they handle paying your employees (check or direct deposit), and as part of the deal withhold payroll taxes and applicable income taxes, send out W-2s to the employees, send the appropriate filings to the IRS, and keep the appropriate records. It's not just the software, but the fact that they also handle things like doing the direct deposit, which otherwise you'd need to set up and keep maintained.
I'm usually a fan of doing things yourself, but for $300-600/yr for a small company for services like those, I would probably just let them handle it, because the fixed costs for dealing with payroll for only a handful of employees are too high.
Here's a plot (thankfully, they give out the raw CSV data) with the "all versions" included. Firefox has a ways to go. http://yfrog.com/j5temptlp
Statcounter also plots that, fwiw. (Click on the dropdown box after "Statistic:" at the bottom-left of the graph to get other views and data sets as well.)
I'm not sure "many" of them are. It's hard to estimate, but most estimates for the proportion of users using some form of ad-blocking software are only in the 3-5% range. Even if every one of those is a Firefox 3.5 user, that would only nudge up the 21% market share to the mid-20%s, not totally rearrange the curve or anything.
Perelman's proof is fairly skeletal, though most/all now agree it contains all the required components and enough of a sketch of the missing details. However,some Chinese mathematicians (Cao and Zhu) filled in some of the details in a massive 300-page journal article. A famous Chinese mathematician, Shing-Tung Yao, was accused of promoting the Cao-Zhu article as the real proof, and taking away credit rightfully due to Perelman. There were other shenanigans alleged on both sides.
To some extent it comes down to a question of insight vs. work, with some on the Chinese mathematicians' side claiming that Perelman basically came up with the high-level breakthrough, but didn't follow through with the work to actually prove the theorem, which they claim is non-trivial--- and so the credit for the proof should go to Cao-Zhu, while Perelman gets credit for coming up with the major ideas that inspired the proof. Others view Perelman as essentially coming up with the proof.
Replying to myself, here's a thread buried in the Amazon reviews for the N900 that seems to have mixed experiences of people getting various tricks to work. It sounds, based on that, like T-Mobile is just being somewhat lax about checking what phones are allowed to connect to the $10 plan, so I'm not sure I'd count on it as a long-term or generally available solution for cheap-data smartphones.
Tmobile has an "unlimited web for phones" for $9.99/mo. It's intended for non-smartphones, basically so you can browse the web on normal phones' tiny screens, or use a Google Maps app. But it can also, apparently, be used with unlocked smartphones, like the N900, that have no way of enforcing a specific premium data plan. Judging by forum chatter, people with jailbroken iPhones are also successfully using the $10/mo plan.
That's certainly the precedent set by JSTOR, a more traditional non-profit initiative with closer ties to libraries. An individual not affiliated with a subscribing institution basically can't get access, outside a few narrow exceptions (like access to a specific journal if you're an individual subscriber to the paper version of the journal). They won't even allow public access to old journals that are in the public domain! Google so far is being much more public-friendly.
The downside of running an old version of os-x is that in the not too distant future both security updates and new versions of applications are likely to dry up.
I suppose it depends on your definition of "distant". I'm still running 10.4 on one machine, and it gets periodic security updates, and all the apps I use still work on it. I imagine 10.5 won't be cut off until at least sometime after 10.4 gets cut off.
Yeah, but OS X 10.5 works fine, and so does Linux. (Older Apples are actually better for Linux, because there's been plenty of time for all the driver support to mature.)
You could get a used 12" Aluminum PowerBook G4 for about the same price if you really want. The CPU and battery life are roughly comparable. The Asus has a newer/better video card, and somewhat more RAM (2gb versus a typical max of 1.25gb for PowerBooks). The Asus is also lighter (3.2 lbs vs. 5.6 lbs), although some of that is due to, well, metal v. plastic.
Slashdot Mirror
PGP has a much easier task, though: it only needs to ensure that people with the key can decrypt content, while people without the key cannot. DRM schemes need to ensure that the same person can only decrypt given content for certain purposes, and not for other purposes.
Maybe it's a nitpick, but the headline "BBC's Plan To Kick Open Source Out of UK TV" to me sounds like someone is against open-source software, and has conjured up a scheme, the primary purpose of which is to harm it.
From the article, though, it seems more likely that the BBC is worried about copyright infringement, and as with many companies, the only sort-of-half-assed solution they can think of to combat it is to introduce some DRM, and the only even-more-half-assed solution they can think of to make it hard to crack the DRM is security-through-obscurity. That's incompatible with OSS, as Cory Doctorow points out, but I think out of a misplaced attempt to use security-through-obscurity, not out of an actual antipathy to open-source vs. proprietary software as licensing models. Who knows if they even realized that: 1) lots of open-source software is used in conjunction with receiving TV broadcasts (and not just by warez groups); and 2) their scheme would therefore harm an important segment of the public.
What's the "true profit motive in a free market" for space exploration and colonization? Mining? Hotels?
Or are you suggesting government bounties of some sort? Because artificial incentives ponied up by government with taxpayer money hardly count as true profit motives in a free market, though they may in some cases be reasonable ways for the government to get something done.
It seems to vary, like the throughput does, and roughly correlated with it. For example, AT&T latency seems better at all the locations in Atlanta. It does have some absolutely terrible ones in some cities, though (700ms+ latencies).
It doesn't really, though. OpenCL "decides" based on some very high-level, high-granularity features of devices it can enumerate. In practice, if you want your code to run reasonably well, you know which parts are going to run on the GPU and which on the CPU. OpenCL isn't an auto-parallelization solution, just a set of primitives for parallel programming--- more like an MPI or OpenMP that also supports GPGPU than the old 70s holy grain of auto-parallelizing where the compiler or runtime magically figures out how to chunk up your computation and where to send the chunks.
As far as I can tell, OpenCL is pretty much based on CUDA, not on an attempt to unify CUDA and CTM/Brook+/FireStream. That's partly because ATI's solutions never really caught on, and have been sorta ignored.
That seemed to be the consensus of the Slashdot discussion after their sudden announcement that they weren't going to pay people to make it after all, but wanted proposals for someone willing to do it for free. Did they actually find someone willing to do it for free? Or was funding restored?
They probably do; I don't really watch television. The actual people in Congress seem a bit more restrained, though, with some exceptions.
Those laws are widely expected to be struck down as a violation of the First Amendment in the next month or two, though. Cases like this seem to be one of the better arguments for the "strike them down" side. I'm not a big fan of corporate money influencing politics, but throwing someone in jail because they're a "former republican party official" with an anti-Congressman blog seems pretty repugnant to free-speech principles.
I typically vote for Democrats rather than Republicans because they don't do the same hyperbolic bullshit, or at least don't do it nearly as often. If they're going to start pulling this crap, then I'll henceforth start filing "voting for a Democrat" in the same category as "voting for Sarah Palin".
If you prefer, this and this predate the Fox story by several days.
Lying in a political context is hard to prosecute, unless it rises to the level of libel, which has a pretty high bar for public officials, and an even higher bar for political speech about public officials.
Lying on forms filed with the government is illegal, though, under a blanket "don't lie to the government" law. The jail part of the complaint seems to be for allegedly misrepresenting the PAC on the filing documents with the FEC: the filed documents claim the PAC isn't aimed at any particular opponent, but the website clearly is aimed at one opponent.
IRB is more weird than lax in my experience. I've seen hassles over a protocol that asked adults to use some experimental software, asking for some BS to qualm the IRB's fears about whether it could cause any psychological harm. Perhaps thrown-together research-software GUI is so horrible someone will be traumatized, but it seems unlikely that any possible such experiment will really rise to the level of risk that even low-risk medical experiments have.
Tax software is hardly simple, though I agree it is on the technical side. Not only is there no "internationally agreed XML schema for the taxman to publish this years' tax legislation", but there is no formal record of this year's tax schema at all. The tax laws are riddled with vaguely worded provisions, some of which (in the US) have been clarified by IRS advisory opinions, tax court precedent, or regular court precedent. You basically need a team of lawyers from every jurisdiction to even figure out what the tax code is. Encoding it in software is the relatively easy part.
Oh, and there are constant updates to it, some of which come out so late that patches ideally should be out 1-2 days after the new regulation is announced. Who is the volunteer who is going to spend their life monitoring the IRS's press releases and patching tax software?
There's lots of web-based stuff too, which is increasingly being offered as a part of a package with other services. Payroll processors are common, for example: they handle paying your employees (check or direct deposit), and as part of the deal withhold payroll taxes and applicable income taxes, send out W-2s to the employees, send the appropriate filings to the IRS, and keep the appropriate records. It's not just the software, but the fact that they also handle things like doing the direct deposit, which otherwise you'd need to set up and keep maintained.
I'm usually a fan of doing things yourself, but for $300-600/yr for a small company for services like those, I would probably just let them handle it, because the fixed costs for dealing with payroll for only a handful of employees are too high.
(I have in mind stuff like Intuit Online Payroll and SurePayroll.)
How about making your own website? Or making a website with people you know?
Not everything has to clone the old-media approach of speech and media as "content" production.
Statcounter also plots that, fwiw. (Click on the dropdown box after "Statistic:" at the bottom-left of the graph to get other views and data sets as well.)
I'm not sure "many" of them are. It's hard to estimate, but most estimates for the proportion of users using some form of ad-blocking software are only in the 3-5% range. Even if every one of those is a Firefox 3.5 user, that would only nudge up the 21% market share to the mid-20%s, not totally rearrange the curve or anything.
Perelman's proof is fairly skeletal, though most/all now agree it contains all the required components and enough of a sketch of the missing details. However,some Chinese mathematicians (Cao and Zhu) filled in some of the details in a massive 300-page journal article. A famous Chinese mathematician, Shing-Tung Yao, was accused of promoting the Cao-Zhu article as the real proof, and taking away credit rightfully due to Perelman. There were other shenanigans alleged on both sides.
To some extent it comes down to a question of insight vs. work, with some on the Chinese mathematicians' side claiming that Perelman basically came up with the high-level breakthrough, but didn't follow through with the work to actually prove the theorem, which they claim is non-trivial--- and so the credit for the proof should go to Cao-Zhu, while Perelman gets credit for coming up with the major ideas that inspired the proof. Others view Perelman as essentially coming up with the proof.
Here's a brief bloggy summary with some links.
Replying to myself, here's a thread buried in the Amazon reviews for the N900 that seems to have mixed experiences of people getting various tricks to work. It sounds, based on that, like T-Mobile is just being somewhat lax about checking what phones are allowed to connect to the $10 plan, so I'm not sure I'd count on it as a long-term or generally available solution for cheap-data smartphones.
Tmobile has an "unlimited web for phones" for $9.99/mo. It's intended for non-smartphones, basically so you can browse the web on normal phones' tiny screens, or use a Google Maps app. But it can also, apparently, be used with unlocked smartphones, like the N900, that have no way of enforcing a specific premium data plan. Judging by forum chatter, people with jailbroken iPhones are also successfully using the $10/mo plan.
That's certainly the precedent set by JSTOR, a more traditional non-profit initiative with closer ties to libraries. An individual not affiliated with a subscribing institution basically can't get access, outside a few narrow exceptions (like access to a specific journal if you're an individual subscriber to the paper version of the journal). They won't even allow public access to old journals that are in the public domain! Google so far is being much more public-friendly.
The downside of running an old version of os-x is that in the not too distant future both security updates and new versions of applications are likely to dry up.
I suppose it depends on your definition of "distant". I'm still running 10.4 on one machine, and it gets periodic security updates, and all the apps I use still work on it. I imagine 10.5 won't be cut off until at least sometime after 10.4 gets cut off.
Yeah, but OS X 10.5 works fine, and so does Linux. (Older Apples are actually better for Linux, because there's been plenty of time for all the driver support to mature.)
You could get a used 12" Aluminum PowerBook G4 for about the same price if you really want. The CPU and battery life are roughly comparable. The Asus has a newer/better video card, and somewhat more RAM (2gb versus a typical max of 1.25gb for PowerBooks). The Asus is also lighter (3.2 lbs vs. 5.6 lbs), although some of that is due to, well, metal v. plastic.