In Missouri the courts told the state they couldn't keep a local chapter of the KKK from adopting a section of highway for trash removal. The courts said that unpopular speech was still protected speech, and putting up a sign saying who was part of the program didn't openly and directly endorse racial violence. The state could end the program, stop putting up signs thanking people, or give the KKK their sign.
The state legislature came to the rescue, though, with another sign. That whole section adopted by the KKK was designated an honorific route. They found themselves in charge of volunteer trash cleanup of the Rosa Parks Memorial Highway. After unsatisfactory participation in the program (one guy showed up once IIRC), the adoption was removed.
It doesn't take a whole state to do something that embarrasses people. It just takes a few neighbors you'd rather not have in the neighborhood practicing the same rights the non-bigots don't want to give up.
Patents are a significant barrier to broad sharing and quick turnaround. Even if you license a patent quite liberally, it takes years to get one and costs tens of thousands of dollars. Plus it has to be something novel enough to be patented. They really aren't the right tool to protect something the same way OSI-approved software licenses protect things.
I split two different points into two different replies. I should have made more clear this one wasn't a summary. The other one is a partial summary of it, and of the most important point in my opinion.
It shouldn't be too hard to read this one, though. I'm not sure how you're unsure of the point.
Individuals aren't the only ones who do open source software and won't be the only ones doing open source hardware if it does catch on. $50k in a company budget for the tools to work with a bunch of different designs, and $2k per prototype is well within the means of many medium-sized (50 to 500 person) companies.
I maintain the problem is when you put in $50k in tooling and test equipment plus $2k per turn, then that mid-sized company goes to market with a design tweaked from yours and never contributes back. Sure, you can reverse engineer it, and in fact you'd have a head start since it's based on your design. You'd still prefer that they honor your significant investment that made theirs cheaper because you did the initial design and the first few rounds of improvement.
If you put a turn in, it's $2000. If someone else makes a turn from your schematics, it's their $2000. If they share their process improvements back to the project, that works great. If they don't, that's where the weakness from the difference between this and software bites.
If I'm someone who only wrote gibberish (that's how it's spelled, by the way) in the thread, you sure are replying down-thread from me a fair bit.
Hint: I both summarized a point of your talk (to which you're replying to replies now) and explained another issue with open hardware. Other people seem to be understanding my other point just fine, and a discussion has started.
The two happen to be in different posts. Sorry for the confusion. This is an open forum. Sometimes the discussion is altered in a way useful to the participants that the original poster never foresaw.;-)
BTW, I think if you really care to lift the veil, I'm not hard to identify by something other than my Slashdot nickname.
He's not against it. He's describing problems that need to be overcome for it to work the way we'd like. For one, you can copyright a schematic but the copyright is on the schematic itself. To keep someone from building what the schematic describes or their own twist on it, you need a patent. Those are much harder to get for one thing. They also require different licensing language from copyright to allow people to use the patent.
short synopsis of one of the most important issues from the talk: Copyright and patent laws don't apply to hardware schematics the way they do to software under US law. You can copyright the schematic and keep people from reproducing it without following the license. You can't keep them from building the hardware the schematic describes.
Your open source software needs to be compiled to run on the hardware. If you and I have the same hardware, we can share a compiler. If you tweak Chip A to have features 1 and 2 while I tweak Chip A to have features 3 and 4, one of a few things happens. We could just not use those features on software we both use. We could fork the compiler. We could try to work out dynamically adjusting the compilation for our feature sets. That's all fine, even as rough as that last one starts to be. But then we have to consider the 18 other variations among our group of 20 installations using Chip A variants.
All this goes more or less smoothly until some well-meaning party comes along with Chip A.1 that does 95% of what Chip A does, but a different way, and then re-expands the additional features their own direction. Then the cycle starts over. Meanwhile, we're struggling to maintain compatibility across Chip A, and since A.1 isn't too much incompatible we decide one compiler should work for both. Then along comes A.2 two weeks later...
So, yeah, information does want to be free. Platforms also want a target that while not entirely stationary can at least have some chance to adapt to the levels beneath them. Some licenses allow you to completely change a work and keep calling it the same thing. With open hardware changing the underlying implementation is fine. If you change the instruction set or change side effects of one instruction being issued after another for any pair of instructions then you've forked the entire environment that sits on top.
TL;DR: It's not that some sort of open license won't work for hardware. It's that it has to be a carefully worded license that fully considers how hardware is different from software.
Github is about code. I should be able to tell you that your code sucks. You should be allowed to tell me my code sucks. If one of us writes enough bad code, the other may even be able to make a claim of "you're a dirty hack and I doubt any of your code will every be useful to me." It's not a site about race, gender, age, national origin, sexual orientation, or general discussion. Anything that's not about the code should be left by the wayside pragmatically. Harassment of any kind is not productive and shouldn't be allowed to get in the way of discussing the code.
4chan is a general discussion forum that's specifically for people to feel free to say anything. If that includes harassment up to the threshold where someone is legitimately threatened, so be it. People should still be held responsible if they are stalking, swatting, threatening, assaulting, or otherwise legitimately harassing people crossing over into real life. They shouldn't be bothered for calling someone hurtful names or for offending people, though.
I don't want 4chan to be Github and I don't want Github to be 4chan. I do think both have the right to exist, both have the right to their own rules on their own sites, and people should stop bitching about the policies of both.
Let's say Bob runs a small business. It's not in the fanciest part of town, but it's a safe enough mixed-use neighborhood. Some whore gets busted on the corner of the block. Now Bob and all his customers are defamed by the city as being criminals. How does Bob's business fair?
There's nothing wrong with having a "safe space" for sensitive discussions. That should be encouraged. What should be discouraged is forcing your entire community to be one so you're never offended, embarrassed, or shocked out in public. You have your rights. I have mine. You can go to your safe space. I can go to mine. When we're out in public, neither of us has the right to tell each other what's acceptable speech.
Yes, but you're not using a scrum of scrums for operational deployments if you're deploying hundreds of times per day. You're using automated integration, which is not a meeting people have. It's Jenkins, Travis, Bamboo, or something. It's not a scrum of scrums, which is a way for multiple teams to send a representative to a meta-scrum team to do the integration and deployment and to remove any cross-team impediments.
You shouldn't have a scrum of scrums every day unless you're running into cross-team impediments every day or are shipping / deploying every day. A couple of times per sprint is plenty for some places.
You also shouldn't have every member of every team at the scrum of scrums. If you have thirty people total, the scrum of scrums should be no more than 7 people or so as it's a representative from each team.
Thirty people is way too large for a scrum team. You'd be better off with five six-person teams and therefore five six-minute meetings. Or maybe six five-minute meetings for six five-person teams. No team of thirty people is going to grab tasks off the board without contention and be as focused as a smaller team.
The main experiments were back in the 1960s. There are some proof-of-concepts for future commercial plants from what I've heard and read. There are some being used to provide power to high-use single users like high-energy research labs I think.
Nobody's producing power to sell just yet. It's supposed to be soon, though. A Canadian company has a design they're putting into pre-licensing review in the coming months to hopefully be online around 2020. The US DoE which first developed MSRs (a program which Nixon axed) is helping China build a full-scaled 100 MW preview unit to be operational by 2024.
These things are safer (thorium vs. uranium for the bulk of the fuel, lower pressure inside the reactor), more efficient (higher temperatures transferred to the water/steam so more work gets to the generators), have easier spent fuel requirements (the half-lives are much shorter and it's much easier to keep them from breeding bomb-grade elements). They'll be cheaper to operate and produce cheaper, safer electricity. China's into the hundreds of millions researching building these things. It should happen.
The main problem for the IRA for decades has been the Six Counties. Ireland is a republic. Northern Ireland is still held and occupied by the UK. The reason things have been so bloody is that part of the local population wants to join Ireland and part wants to stay in the UK. The UK army was then stationed there to keep the peace, and neither local side really wanted them there. The IRA wanted them there even less than the UK loyalists.
You see, Northern Ireland is almost all Irish with some Scottish and other ethnicities. It's on the island of Eire. But it's not part of Eire politically. That's the contentious issue. From one side the question is should they be part of the country with which some of them yearn to rejoin or stay part of an occupying foreign empire. From the other side it's should they stay with the country they've been part of for a long time and are happy to be part of, or should they be ripped away and forced to join some other country due to ethnic and historical reasons. It's really not an easy thing to solve. It's amazing it's been this peaceful for this long in recent years.
Slashdot Mirror
So it's hackaday raiding Reddit for story ideas this time. Sorry, Buzzfeed, I blamed you out of habit.
Seriously, https://www.reddit.com/r/histo... was yesterday.
In Missouri the courts told the state they couldn't keep a local chapter of the KKK from adopting a section of highway for trash removal. The courts said that unpopular speech was still protected speech, and putting up a sign saying who was part of the program didn't openly and directly endorse racial violence. The state could end the program, stop putting up signs thanking people, or give the KKK their sign.
The state legislature came to the rescue, though, with another sign. That whole section adopted by the KKK was designated an honorific route. They found themselves in charge of volunteer trash cleanup of the Rosa Parks Memorial Highway. After unsatisfactory participation in the program (one guy showed up once IIRC), the adoption was removed.
It doesn't take a whole state to do something that embarrasses people. It just takes a few neighbors you'd rather not have in the neighborhood practicing the same rights the non-bigots don't want to give up.
Patents are a significant barrier to broad sharing and quick turnaround. Even if you license a patent quite liberally, it takes years to get one and costs tens of thousands of dollars. Plus it has to be something novel enough to be patented. They really aren't the right tool to protect something the same way OSI-approved software licenses protect things.
http://slashdot.org/comments.p...
I split two different points into two different replies. I should have made more clear this one wasn't a summary. The other one is a partial summary of it, and of the most important point in my opinion.
It shouldn't be too hard to read this one, though. I'm not sure how you're unsure of the point.
Individuals aren't the only ones who do open source software and won't be the only ones doing open source hardware if it does catch on. $50k in a company budget for the tools to work with a bunch of different designs, and $2k per prototype is well within the means of many medium-sized (50 to 500 person) companies.
I maintain the problem is when you put in $50k in tooling and test equipment plus $2k per turn, then that mid-sized company goes to market with a design tweaked from yours and never contributes back. Sure, you can reverse engineer it, and in fact you'd have a head start since it's based on your design. You'd still prefer that they honor your significant investment that made theirs cheaper because you did the initial design and the first few rounds of improvement.
You need to make one of the reward levels being on the panel that chooses which women are attractive.
Strawmen make terrible marketers and worse engineers.
If you put a turn in, it's $2000. If someone else makes a turn from your schematics, it's their $2000. If they share their process improvements back to the project, that works great. If they don't, that's where the weakness from the difference between this and software bites.
If I'm someone who only wrote gibberish (that's how it's spelled, by the way) in the thread, you sure are replying down-thread from me a fair bit.
Hint: I both summarized a point of your talk (to which you're replying to replies now) and explained another issue with open hardware. Other people seem to be understanding my other point just fine, and a discussion has started.
The two happen to be in different posts. Sorry for the confusion. This is an open forum. Sometimes the discussion is altered in a way useful to the participants that the original poster never foresaw. ;-)
BTW, I think if you really care to lift the veil, I'm not hard to identify by something other than my Slashdot nickname.
He's not against it. He's describing problems that need to be overcome for it to work the way we'd like. For one, you can copyright a schematic but the copyright is on the schematic itself. To keep someone from building what the schematic describes or their own twist on it, you need a patent. Those are much harder to get for one thing. They also require different licensing language from copyright to allow people to use the patent.
short synopsis of one of the most important issues from the talk: Copyright and patent laws don't apply to hardware schematics the way they do to software under US law. You can copyright the schematic and keep people from reproducing it without following the license. You can't keep them from building the hardware the schematic describes.
Your open source software needs to be compiled to run on the hardware. If you and I have the same hardware, we can share a compiler. If you tweak Chip A to have features 1 and 2 while I tweak Chip A to have features 3 and 4, one of a few things happens. We could just not use those features on software we both use. We could fork the compiler. We could try to work out dynamically adjusting the compilation for our feature sets. That's all fine, even as rough as that last one starts to be. But then we have to consider the 18 other variations among our group of 20 installations using Chip A variants.
All this goes more or less smoothly until some well-meaning party comes along with Chip A.1 that does 95% of what Chip A does, but a different way, and then re-expands the additional features their own direction. Then the cycle starts over. Meanwhile, we're struggling to maintain compatibility across Chip A, and since A.1 isn't too much incompatible we decide one compiler should work for both. Then along comes A.2 two weeks later...
So, yeah, information does want to be free. Platforms also want a target that while not entirely stationary can at least have some chance to adapt to the levels beneath them. Some licenses allow you to completely change a work and keep calling it the same thing. With open hardware changing the underlying implementation is fine. If you change the instruction set or change side effects of one instruction being issued after another for any pair of instructions then you've forked the entire environment that sits on top.
TL;DR: It's not that some sort of open license won't work for hardware. It's that it has to be a carefully worded license that fully considers how hardware is different from software.
Well, they considered Chicago but those folks are too mired in the details to be impartial about exhibits.
Github is about code. I should be able to tell you that your code sucks. You should be allowed to tell me my code sucks. If one of us writes enough bad code, the other may even be able to make a claim of "you're a dirty hack and I doubt any of your code will every be useful to me." It's not a site about race, gender, age, national origin, sexual orientation, or general discussion. Anything that's not about the code should be left by the wayside pragmatically. Harassment of any kind is not productive and shouldn't be allowed to get in the way of discussing the code.
4chan is a general discussion forum that's specifically for people to feel free to say anything. If that includes harassment up to the threshold where someone is legitimately threatened, so be it. People should still be held responsible if they are stalking, swatting, threatening, assaulting, or otherwise legitimately harassing people crossing over into real life. They shouldn't be bothered for calling someone hurtful names or for offending people, though.
I don't want 4chan to be Github and I don't want Github to be 4chan. I do think both have the right to exist, both have the right to their own rules on their own sites, and people should stop bitching about the policies of both.
Let's say Bob runs a small business. It's not in the fanciest part of town, but it's a safe enough mixed-use neighborhood. Some whore gets busted on the corner of the block. Now Bob and all his customers are defamed by the city as being criminals. How does Bob's business fair?
There's nothing wrong with having a "safe space" for sensitive discussions. That should be encouraged. What should be discouraged is forcing your entire community to be one so you're never offended, embarrassed, or shocked out in public. You have your rights. I have mine. You can go to your safe space. I can go to mine. When we're out in public, neither of us has the right to tell each other what's acceptable speech.
Yes, but you're not using a scrum of scrums for operational deployments if you're deploying hundreds of times per day. You're using automated integration, which is not a meeting people have. It's Jenkins, Travis, Bamboo, or something. It's not a scrum of scrums, which is a way for multiple teams to send a representative to a meta-scrum team to do the integration and deployment and to remove any cross-team impediments.
You shouldn't have a scrum of scrums every day unless you're running into cross-team impediments every day or are shipping / deploying every day. A couple of times per sprint is plenty for some places.
You also shouldn't have every member of every team at the scrum of scrums. If you have thirty people total, the scrum of scrums should be no more than 7 people or so as it's a representative from each team.
According to Sturgeon, 90% of all software teams are crap whether or not they use any particular methodology.
A sole developer usually doesn't need team meetings so rules on how to conduct those meetings would be kind of pointless.
Thirty people is way too large for a scrum team. You'd be better off with five six-person teams and therefore five six-minute meetings. Or maybe six five-minute meetings for six five-person teams. No team of thirty people is going to grab tasks off the board without contention and be as focused as a smaller team.
"Flexible" and "agile" used to mean very similar things. I think in the real world they still may.
The main experiments were back in the 1960s. There are some proof-of-concepts for future commercial plants from what I've heard and read. There are some being used to provide power to high-use single users like high-energy research labs I think.
Nobody's producing power to sell just yet. It's supposed to be soon, though. A Canadian company has a design they're putting into pre-licensing review in the coming months to hopefully be online around 2020. The US DoE which first developed MSRs (a program which Nixon axed) is helping China build a full-scaled 100 MW preview unit to be operational by 2024.
These things are safer (thorium vs. uranium for the bulk of the fuel, lower pressure inside the reactor), more efficient (higher temperatures transferred to the water/steam so more work gets to the generators), have easier spent fuel requirements (the half-lives are much shorter and it's much easier to keep them from breeding bomb-grade elements). They'll be cheaper to operate and produce cheaper, safer electricity. China's into the hundreds of millions researching building these things. It should happen.
http://www.world-nuclear.org/i...
http://fukushimaupdate.com/tho...
http://www.technologyreview.co...
http://fortune.com/2015/02/02/...
http://www.forbes.com/sites/ke...
https://en.wikipedia.org/wiki/...
Molten salt reactors are said to have, in theory, fuel remnants with storage needs of a few centuries rather than tens of millennia.
The main problem for the IRA for decades has been the Six Counties. Ireland is a republic. Northern Ireland is still held and occupied by the UK. The reason things have been so bloody is that part of the local population wants to join Ireland and part wants to stay in the UK. The UK army was then stationed there to keep the peace, and neither local side really wanted them there. The IRA wanted them there even less than the UK loyalists.
https://en.wikipedia.org/wiki/...
You see, Northern Ireland is almost all Irish with some Scottish and other ethnicities. It's on the island of Eire. But it's not part of Eire politically. That's the contentious issue. From one side the question is should they be part of the country with which some of them yearn to rejoin or stay part of an occupying foreign empire. From the other side it's should they stay with the country they've been part of for a long time and are happy to be part of, or should they be ripped away and forced to join some other country due to ethnic and historical reasons. It's really not an easy thing to solve. It's amazing it's been this peaceful for this long in recent years.