Ok, granted, abolishing 90% of software patents based on quality would improve some things, but it's the least efficient way to fix the system. We'd be fools to aim for this.
If you've ten fields with ten land in mines each, the worst way to clear 90% of them would be to remove nine from each field. The optimum way is to completely rid nine fields of mines.
Raising quality is equivalent to removing 90% but leaving a mine in each field. The equivalent to the optimum situation is: exclude certain domains from patentability. Imagine being able to say "My software has to play videos, and I know reading, writing, and transmitting videos can't infringe patents, so I'm good". That's what we need. Even going 20% in this direction would be better than going 90% in the quality direction.
(An example of this working in other domains is: writing a novel. No author ever has to thank his lucky stars that he only has to face 30 patents.)
Discussion of trolls and quality is a bad use of time. Both have been hot topics for more than a decade, and the problem has only been getting worse.
> It's the invalid and questionable patents that create expense and work.
On the contrary, those are the easiest ones. At least you've a chance against those ones. It's the clearly valid ones that will kill your project.
> Even if there's a genuine legitimate patent in that haystack, > it helps a lot if you can find it and act accordingly.
This is the same myth as above: that "appropriate action" (or "sufficient work") can solve patent problems. Fact is, if the patented idea is necessary for your project, and it's a "genuine legitimate patent", then you're at the mercy of the patent holder.
("Necessary" can be absolute, such as the need to read a certain file format, or it can be a practical matter such as an inconvenience or a 20% speed decrease which would cause a sufficient loss of customers to make your business unprofitable.)
Maybe you can pay them. Maybe. But a lot of software is distributed in ways which aren't compatible with paying the usual per-copy fees. Not just free software, but also freemium models and software which is developed as a byproduct of the person doing their real job. Even if they do give you a payment option that works for you, you don't know if they'll still be offering this option next year.
Maybe they don't want to be paid. They could prefer being the only company that can do this job.
The "act accordingly" options available are pretty lousy, and it's out of your hands.
I can't see how the price difference could be significant.
If I'm on a 747 and it's 80% full, that's 292 people.
The plane weighs 162,400Kg (358,000 lbs). 292 people @ 80 kg (170lbs) is 23,400kg, for a plane+people weight of 185,000kg, or 636kg per person.
How big would someone have to be before you charge them more? 200 kg (440 lbs) ? When blended in with their portion of the plane, we're only talking an increase from 636 to 756 kg, or roughly 20%.
So you want a small number of people to pay 20% more for the *fuel portion* of their ticket price?
This is a big discussion for a small number of euros/dollars from a small number of people. Move on.
(Asking a "two-seat person" to pay for two seats is a different question.)
That's not how it works. Fighting patents isn't a question of "work". If a patent is valid, and it's necessary for what you're doing, then you're screwed. You can't just throw "work" at the problem.
If you're faced with one to five patents, and you think they're invalid or you think you don't infringe them, *then* it's a question of work (and crossing you're fingers you don't get unlucky with the judge/jury). 30 patents? The 30 which are more "legit" than the 270 we already discarded? Forget it.
That's not the problem either. Besides the practical problems of defining "actual inventions" or on subjectively interpreting whether an application meets this standard, a small number of software patents is almost as bad as a very large number.
Imagine we raise the obviousness bar so high that 90% of software patents get invalidated. The Mpeg video formats would be covered by 100 patents instead of 1000. What does that change? Nothing.
Microsoft has, IIRC, 300 patents is uses when shaking down distributors of smartphones. If that was cut to 30, would anything change?
For new formats, even if Microsoft had only one single software patent that they knew was valid, they can design their format in a way that requires it.
The problem isn't lax evaluations, it's that patents don't work in a domain where development can be cheap, distribution can be free, both can be done by average people, and compatibility is so important (with the implied tendency to foster monopolies).
* Microsoft getting royalties from hundreds of millions of smartphones that contain no Microsoft software (and we're lucky they're currently only asking for royalties - they have an equal right to simply tell others to stop developing!)
* IBM getting 1,200 patents on cloud computing
* Nokia and other companies with failing software divisions sitting on a mountain of software patents
* Video formats being covered by 1,000+ patents of telecoms, software companies, hardware companies...
* Facebook patenting social services, as if competing with their dominant position wasn't hard enough
Trolls are a real problem, but the topic is also used to draw people's attention away from the bigger problems created by software companies with big patent portfolios.
Yeah. It can't fuck with people who don't buy shiny eavesdropping devices. 1950's tech FTW as always! At least things from back then still work fine now.
Boycotting stuff doesn't work when your friends and other nearby people have shiny things that are recording the ambient audio etc.
Boycotting stuff also gets harder when there are no cars on sale (in reasonable price range) which don't include such devices. And it gets harder again when houses go that way.
Does this software do something that the Blender devs currently can't do (and aren't making good progress on)? Will Blender celebrate this release?
(Note: Renderman is still proprietary, it just costs nothing for a non-commercial licence. Not free software at all.)
(Note #2: The announcement is their "intent" to release it as open source by summer 2016. Currently no code has been released and no licence can be read.)
If you had to remote to another machine for every intensive task, do you think you'd maintain the same volume of intensive tasks?
In summary, he reduced a bunch of things to zero, and another heap got reduced greatly, and then some smart aleck comes along and says he did nothing because this thing over here didn't get reduced much. Someone's missing the point.
I haven't looked into how Mega worked, but with my limited knowledge I'm assuming end-to-end encryption via a website must require the website sending javascript to the user which the browser will run to encrypt the files before upload. If that's how Mega worked when Kim was there, then the operator would have no way to decrypt the files and national security letters wouldn't work. A few web searches suggest that this is indeed how it worked. (Example: articles say "don't lose your password, it's the only encryption key and the operators don't have a copy".)
The downside of this user-friendliness is that the operator could always start sending out new javascript which simply uploads the file and then encrypt with some other key. The user wouldn't notice this change. Then national security letters etc. would be a problem.
If users want encryption to be done for them (instead of using GnuPG themselves, or running their own local copy of the javascript), I don't see any way to avoid this problem.
If this javascript change has happened, and it was when Kim was in charge, then he's at fault. If it happened after he left then I don't see how he can be blamed. No?
It doesn't sound like that's what the OP was referring to. If that was what he was referring to, then I don't agree that someone's bad deeds when they were 20 imply that they can't be trusted when they're 41.
I'm not trying to defend the guy, I don't know much about him, but the jabs being levelled at him here just don't seem very well founded.
> Nor should we expect his next iteration of cloud filesharing to be fully encrypted.
I don't know the history of Mega's downfall. Have you a link to suggest that Kim didn't do his best?
In this interview he comes across well enough for me to take a wait-and-see attitude to his future businesses. (Although, IIRC, Mega required running non-free javascript in your browser. Sticking a GPL or another free licence on the code you have to run to use the service would be a big improvement.)
If Mega was serving javascript for client-side encryption, and the new owner replaced this with server-side encryption, then the lack of safety would only apply to post-Kim era uploads to Mega, and there's no way anyone could have future proofed against that possibility. (The closest thing to future proof would be to offer no encryption services at all, and tell the users to use GnuPG. In this way, the users would never have gotten used to Mega doing the encryption for them.)
I actually hope the guy who shot it down just gets a small fine and let go.
I hope he gets no fine. And the drone operator gets investigated for trespass, peeping tommery, or whatever other offenses exist to protect people's privacy.
Software is running our lives. Anyone who can write software knows that, for example, electronic voting can be easily fixed. People who haven't written basic software find this question hard to analyse. Even if these people don't become full-time coders, it's still good that society has more and more people who know how the software running our lives works.
Banning these things from the civilian economy, or placing restrictions which would reduce demand (example: need a licence), would certainly slow down development greatly. The military's ability to finance this sort of tech is small compared to society's. Computers are a example where the military benefits from development financed almost completely by society. (Computers are only an example of the funding model, I'm not suggesting limiting computers. They do way more good than bad, which probably won't be true for autonomous weapons.)
Slashdot Mirror
Ok, granted, abolishing 90% of software patents based on quality would improve some things, but it's the least efficient way to fix the system. We'd be fools to aim for this.
If you've ten fields with ten land in mines each, the worst way to clear 90% of them would be to remove nine from each field. The optimum way is to completely rid nine fields of mines.
Raising quality is equivalent to removing 90% but leaving a mine in each field. The equivalent to the optimum situation is: exclude certain domains from patentability. Imagine being able to say "My software has to play videos, and I know reading, writing, and transmitting videos can't infringe patents, so I'm good". That's what we need. Even going 20% in this direction would be better than going 90% in the quality direction.
(An example of this working in other domains is: writing a novel. No author ever has to thank his lucky stars that he only has to face 30 patents.)
Discussion of trolls and quality is a bad use of time. Both have been hot topics for more than a decade, and the problem has only been getting worse.
> It's the invalid and questionable patents that create expense and work.
On the contrary, those are the easiest ones. At least you've a chance against those ones. It's the clearly valid ones that will kill your project.
> Even if there's a genuine legitimate patent in that haystack,
> it helps a lot if you can find it and act accordingly.
This is the same myth as above: that "appropriate action" (or "sufficient work") can solve patent problems. Fact is, if the patented idea is necessary for your project, and it's a "genuine legitimate patent", then you're at the mercy of the patent holder.
("Necessary" can be absolute, such as the need to read a certain file format, or it can be a practical matter such as an inconvenience or a 20% speed decrease which would cause a sufficient loss of customers to make your business unprofitable.)
Maybe you can pay them. Maybe. But a lot of software is distributed in ways which aren't compatible with paying the usual per-copy fees. Not just free software, but also freemium models and software which is developed as a byproduct of the person doing their real job. Even if they do give you a payment option that works for you, you don't know if they'll still be offering this option next year.
Maybe they don't want to be paid. They could prefer being the only company that can do this job.
The "act accordingly" options available are pretty lousy, and it's out of your hands.
I can't see how the price difference could be significant.
If I'm on a 747 and it's 80% full, that's 292 people.
The plane weighs 162,400Kg (358,000 lbs). 292 people @ 80 kg (170lbs) is 23,400kg, for a plane+people weight of 185,000kg, or 636kg per person.
How big would someone have to be before you charge them more? 200 kg (440 lbs) ? When blended in with their portion of the plane, we're only talking an increase from 636 to 756 kg, or roughly 20%.
So you want a small number of people to pay 20% more for the *fuel portion* of their ticket price?
This is a big discussion for a small number of euros/dollars from a small number of people. Move on.
(Asking a "two-seat person" to pay for two seats is a different question.)
That's not how it works. Fighting patents isn't a question of "work". If a patent is valid, and it's necessary for what you're doing, then you're screwed. You can't just throw "work" at the problem.
If you're faced with one to five patents, and you think they're invalid or you think you don't infringe them, *then* it's a question of work (and crossing you're fingers you don't get unlucky with the judge/jury). 30 patents? The 30 which are more "legit" than the 270 we already discarded? Forget it.
That's not the problem either. Besides the practical problems of defining "actual inventions" or on subjectively interpreting whether an application meets this standard, a small number of software patents is almost as bad as a very large number.
Imagine we raise the obviousness bar so high that 90% of software patents get invalidated. The Mpeg video formats would be covered by 100 patents instead of 1000. What does that change? Nothing.
Microsoft has, IIRC, 300 patents is uses when shaking down distributors of smartphones. If that was cut to 30, would anything change?
For new formats, even if Microsoft had only one single software patent that they knew was valid, they can design their format in a way that requires it.
The problem isn't lax evaluations, it's that patents don't work in a domain where development can be cheap, distribution can be free, both can be done by average people, and compatibility is so important (with the implied tendency to foster monopolies).
In software, more problems come from non-trolls.
* Microsoft getting royalties from hundreds of millions of smartphones that contain no Microsoft software (and we're lucky they're currently only asking for royalties - they have an equal right to simply tell others to stop developing!)
* IBM getting 1,200 patents on cloud computing
* Nokia and other companies with failing software divisions sitting on a mountain of software patents
* Video formats being covered by 1,000+ patents of telecoms, software companies, hardware companies...
* Facebook patenting social services, as if competing with their dominant position wasn't hard enough
Trolls are a real problem, but the topic is also used to draw people's attention away from the bigger problems created by software companies with big patent portfolios.
Oh. Thanks for letting us know. Whatever the future brings, I hope you folk get more of a say.
C'mon Dice, making Slashdot Videos require proprietary software?!
(This is my first time voicing a Dice-era complaint.)
Yeah. It can't fuck with people who don't buy shiny eavesdropping devices. 1950's tech FTW as always! At least things from back then still work fine now.
Boycotting stuff doesn't work when your friends and other nearby people have shiny things that are recording the ambient audio etc.
Boycotting stuff also gets harder when there are no cars on sale (in reasonable price range) which don't include such devices. And it gets harder again when houses go that way.
Does this software do something that the Blender devs currently can't do (and aren't making good progress on)? Will Blender celebrate this release?
(Note: Renderman is still proprietary, it just costs nothing for a non-commercial licence. Not free software at all.)
(Note #2: The announcement is their "intent" to release it as open source by summer 2016. Currently no code has been released and no licence can be read.)
If you had to remote to another machine for every intensive task, do you think you'd maintain the same volume of intensive tasks?
In summary, he reduced a bunch of things to zero, and another heap got reduced greatly, and then some smart aleck comes along and says he did nothing because this thing over here didn't get reduced much. Someone's missing the point.
Why not run your own DNS and do
*.facebook.[com|net]
*.fb.com
etc ? While you're at it you can do *.cn and *.ru too.
Just never got around to it.
For me it's gone into my /etc/hosts
0.0.0.0 1-edge-chat.facebook.com
0.0.0.0 2-edge-chat.facebook.com
0.0.0.0 3-edge-chat.facebook.com
0.0.0.0 4-edge-chat.facebook.com
0.0.0.0 5-edge-chat.facebook.com
0.0.0.0 6-edge-chat.facebook.com
0.0.0.0 7-edge-chat.facebook.com
0.0.0.0 8-edge-chat.facebook.com
0.0.0.0 9-edge-chat.facebook.com
0.0.0.0 apps.facebook.com
0.0.0.0 connect.facebook.net
0.0.0.0 de-de.facebook.com
0.0.0.0 developers.facebook.com
0.0.0.0 error.facebook.com
0.0.0.0 es-es.facebook.com
0.0.0.0 facebook.com
0.0.0.0 facebook.net
0.0.0.0 fbcdn.com
0.0.0.0 fbcdn-creative-a.akamaihd.net
0.0.0.0 fbcdn.net
0.0.0.0 fbcdn-profile-a.akamaihd.net
0.0.0.0 fb.com
0.0.0.0 fbexternal-a.akamaihd.net
0.0.0.0 fbm.mysocialpixel.com
0.0.0.0 fbstatic-a.akamaihd.net
0.0.0.0 fr-fr.facebook.com
0.0.0.0 internet.org
0.0.0.0 l.facebook.com
0.0.0.0 login.facebook.com
0.0.0.0 newsroom.fb.com
0.0.0.0 nl-nl.facebook.com
0.0.0.0 pixel.facebook.com
0.0.0.0 scontent-ams.xx.fbcdn.net
0.0.0.0 s-static.ak.facebook.com
0.0.0.0 static.ak.connect.facebook.com
0.0.0.0 static.ak.facebook.com
0.0.0.0 static.ak.fbcdn.net
0.0.0.0 static.api.ak.facebook.com
0.0.0.0 threatexchange.fb.com
0.0.0.0 upload.facebook.com
0.0.0.0 www.connect.facebook.net
0.0.0.0 www.facebook.com
0.0.0.0 www.facebook.net
0.0.0.0 www.fbcdn.com
0.0.0.0 www.fbcdn.net
0.0.0.0 www.fb.com
0.0.0.0 www.login.facebook.com
0.0.0.0 www.static.ak.connect.facebook.com
0.0.0.0 www.static.ak.facebook.com
0.0.0.0 www.static.ak.fbcdn.net
I haven't looked into how Mega worked, but with my limited knowledge I'm assuming end-to-end encryption via a website must require the website sending javascript to the user which the browser will run to encrypt the files before upload. If that's how Mega worked when Kim was there, then the operator would have no way to decrypt the files and national security letters wouldn't work. A few web searches suggest that this is indeed how it worked. (Example: articles say "don't lose your password, it's the only encryption key and the operators don't have a copy".)
The downside of this user-friendliness is that the operator could always start sending out new javascript which simply uploads the file and then encrypt with some other key. The user wouldn't notice this change. Then national security letters etc. would be a problem.
If users want encryption to be done for them (instead of using GnuPG themselves, or running their own local copy of the javascript), I don't see any way to avoid this problem.
If this javascript change has happened, and it was when Kim was in charge, then he's at fault. If it happened after he left then I don't see how he can be blamed. No?
It doesn't sound like that's what the OP was referring to. If that was what he was referring to, then I don't agree that someone's bad deeds when they were 20 imply that they can't be trusted when they're 41.
I'm not trying to defend the guy, I don't know much about him, but the jabs being levelled at him here just don't seem very well founded.
> Why did you lie under oath in court to try and get John Banks put in jail?
Have you got a link? I did a quick search and found:
"Kim Dotcom challenges John Banks to sit-down interview"
http://www.nzherald.co.nz/nz/n...
And an article with plenty of details, but it boils down to each saying the other's lying:
http://www.stuff.co.nz/nationa...
> Nor should we expect his next iteration of cloud filesharing to be fully encrypted.
I don't know the history of Mega's downfall. Have you a link to suggest that Kim didn't do his best?
In this interview he comes across well enough for me to take a wait-and-see attitude to his future businesses. (Although, IIRC, Mega required running non-free javascript in your browser. Sticking a GPL or another free licence on the code you have to run to use the service would be a big improvement.)
If Mega was serving javascript for client-side encryption, and the new owner replaced this with server-side encryption, then the lack of safety would only apply to post-Kim era uploads to Mega, and there's no way anyone could have future proofed against that possibility. (The closest thing to future proof would be to offer no encryption services at all, and tell the users to use GnuPG. In this way, the users would never have gotten used to Mega doing the encryption for them.)
Yep. A good read, and he comes across as a lot more reasonable than I expected.
Is it bigger than 40?
Depends on the value of ^
Good luck with that! This is how drones fly:
https://www.youtube.com/watch?...
You're most likely to miss and end up as a youtube video titled "Wheezing middle-ager brings hose to a drone fight HAHAHA FAIL!!!"
So you follow the drone, remaining on public property and filming with your phone.
I might have agreed with you until I saw how fast and long range drones fly: https://www.youtube.com/watch?v=8p5uDf9i_Yc
And as for taking it down, suggestions of using a hose or a pool scraper or thread are non-starters. Even with a gun, you'll only get one shot.
I actually hope the guy who shot it down just gets a small fine and let go.
I hope he gets no fine. And the drone operator gets investigated for trespass, peeping tommery, or whatever other offenses exist to protect people's privacy.
Ah yes, "rigged" would have been clearer. Thanks.
Software is running our lives. Anyone who can write software knows that, for example, electronic voting can be easily fixed. People who haven't written basic software find this question hard to analyse. Even if these people don't become full-time coders, it's still good that society has more and more people who know how the software running our lives works.
Banning these things from the civilian economy, or placing restrictions which would reduce demand (example: need a licence), would certainly slow down development greatly. The military's ability to finance this sort of tech is small compared to society's. Computers are a example where the military benefits from development financed almost completely by society. (Computers are only an example of the funding model, I'm not suggesting limiting computers. They do way more good than bad, which probably won't be true for autonomous weapons.)