Sorry, I meant to joke with the previous statement but upon reviewing it I came across as too snarky. I apologize, I meant to tease you and not insult you.
I think you're right. While there are a number of incredibly sophisticated free software projects - Gnu C Compiler, Linux kernel, GNOME, KDE, VLC, Postgres, etc... a lot of cutting edge stuff is proprietary and the free software equivalent comes out later and is often - though not always - inferior. Speech-to-text is one of those areas, as is Virtual Reality, and 3D Gaming.
This is what I get for reading about hosting email from public sites. I kept reading that it was all but impossible, and that messages are automatically dropped. I'll give it a try. Thanks for correcting me.
"Google Search is not done due to it requires millions of servers in order to handle the same load so the software part is of no interest since no one else have access to that amount of hardware."
No, there is interest and people are trying - Yacy is the example. The data and the service are distributed and decentralized, so it's technologically difficult to pinpoint the searches of individual users. And since there are hundreds of nodes, or at least there were, there are plenty of resources available. But even though, if I recall correctly, Yacy uses parts of the Lucene and Solr projects its search features are awful. I did a search for the cartoon Spongebob with the simple keyword "Spongebob" and got neither the official website nor the Wikipedia page in the first 50 responses.
So people care, and people are willing to contribute. But this is an incredibly difficult thing to tackle.
Eben Upton is looking at this from the perspective of someone buying CPUs to incorporate into some future version of the Raspberry Pi, not building or designing a new CPU for the Raspberry Pi or for any other use. I think, given that angle, the copyright and patent advantages of RISC-V are irrelevant to him.
I think the point you're getting at is valid in general. The PC market exploded when IBM allowed clones, and Apple and other players were crushed because the dozens of companies trying to make good products cheaply drove rapid innovation. Android launched as a piece of junk next to the iPhone and IOS, but the open nature of it let dozens or hundreds of companies use it, and the software and hardware evolved very rapidly. RISC-V could do the same thing for processors against x86 and ARM.
I disagree. I don't use Siri, or Google Now, or Cortana specifically because of the privacy implications. If I had available a completely open source, end-user-owned, private option I would use it all of the time for things like checking the weather, setting reminders, playing music, adjusting the temperature, assembling grocery lists, making purchases, etc... etc...
As an aside, your contempt for the average person won't help our cause. Elitism helps none of us - the best way to make open source stronger is to get more people using it, because some of those new users will in turn become contributors.
One of the fascinating things being tried right now is building decentralized applications on top of the cryptocurrency platforms. Ethereum, MaidSafe/Safecoin, Lisk, and others. The data is distributed in a distributed hash table (DHT). Users contribute CPU cycles, RAM usage, and disk space to the network in return for tiny portions of the respective digital currency. They're trying to build distributed autonomous organizations (DAOs), distributed alternatives to Twitter and Facebook, distributed alternatives to Dropbox and Google Drive, etc... etc... something like Siri could run on the same network.
If it ever works right. I'm hopeful but not willing to predict anything.
If I trusted the service completely, I would use it all of the time. "What's the weather tomorrow?" "Play Houses of the Holy." "Send a text to my wife asking her to get milk on the way home from work." "Find the cheapest price plus shipping of a DVD of Idiocracy on Ebay." etc... etc...
I never use Google Now/Siri/Cortana because I don't trust their respective owners with even more data about me than they already have.
Nonsense. There would be enormous use of fully open source alternatives to Google search, Gmail, Call of Duty, Starcraft 2, Destiny, and dozens of other similar projects. The best, to my knowledge, fully open source search engine is Yacy and it totally sucks. Running your own email server isn't too hard, but getting your mail to recipients on Gmail, Hotmail, or Yahoo mail without relaying through one of the big services is all but impossible. There are plenty of nice graphical fully open source video games out there, but nothing with the artwork or the voice acting or the visuals on par with a top of the line AAA game.
Nobody is making them because it's too damn difficult.
I think the critical difference has less to do with the advantages of 4k over 2k and more to do with the fact that 4k prices are dropping rapidly. I waited a long time to get a 37 inch 1080p television because for a long time it cost more than twice as much as a 37 inch 480p television. Today the price premium for 4K is already only 30-40%, sometimes less.
I've been using it on one of my desktops for over a year without any problems. One headache is that they have no upgrade option between major releases beyond back up and reinstall. So I'm going to set aside a few hours to replace 0.3 Freya with 0.4 Loki.
I haven't tried it for any large scale teams, but Rocket.chat is a great open source Slack replacement for small teams. It may scale well, I just don't know.
We can sit in the corner happy with BlackBox window manager or similar, and they're functional enough to get the job done. But if you want free software to conquer the world - which I very much do - we're going to have to make something that the other 99.7% of the population wants to use for their desktop. That requires the eyecandy, sorry.
If FVWM was going to conquer the world, it would have done it twenty years ago. Again, no disrespect to fans of BlackBox or FVWM or anything else that uses the X protocol efficiently. But the goal is to make something our colleagues, friends, and family members use without hating and in my experience 1990s Unix desktop graphics don't win any fans.
I guess the study has some value. But I have two immediate questions:
1. Is immersing a hand in cold water equivalent to all other forms of pain? What if marijuana increases the average man's ability to tolerate cold hands 20% and has an insignificant affect for women in that area, but increases a woman's tolerance for bruises, or stomach cancer pain, more than it does for men?
2. In my experience, women are bothered more by cold than men in general. I realize the study is supposedly double blind, so they tried to account for that. But I think that could be a factor to include. If the average guy in the placebo portion could tolerate 90 seconds and the average woman in the placebo portion could tolerate 60, and then the average guy on pot can tolerate 120 and the average woman on pot can tolerate 63, maybe the difference in baselines is important.
The Broadwell-E part they benchmarked against is probably a (slightly underclocked) Core i7-6900k, and it's $1100.
I'm taking wild guesses with the numbers here, but "15% slower than a Broadwell-E at a 45% lower cost and a similar TDP" is a valid market strategy. I haven't spent more than $240 for a CPU in over ten years, if in spring 2017 there are Zen parts at the $250 price point that are 15% behind the 2017 spring equivalent of the Intel i7-6700k or i7-6800k I will buy it.
Don't mess with the tinfoil hat, man. (Or woman, whatever.)
Our smart phones have software running in the wireless carrier cellular modem that end users can't access or control. Almost all recent x86 laptops and desktops have Intel's TPM which the regular users can't control. Our ISPs track the websites we visit. Our credit card and debit card companies track every purchase we make. Even our grocery stores use loyalty card programs to see what we do. And browser fingerprinting tech like the "Evercookie" mean that even if you run all of the ad-blockers you want and a VPN or Tor, some companies have a detailed profile on your browsing habits on public sites.
So yeah, if free-from-top-to-bottom computing devices and discussion forums and web stores running on top of distributed decentralized platforms like Ethereum become popular, I fully expect governments to throw up roadblocks to maintain surveillance and businesses to throw up roadblocks to maintain their advertising and tracking revenue. Maybe shipments get held up for nonsense reasons in customs, maybe the projects get buried in nonsense patent lawsuits (small companies often fold under those pressures even if the plaintiff is full of shit, because the defendant can't afford the legal fees), maybe Ecmascript 8 becomes a power-hungry monster solely so that users need to buy the latest Intel/AMD 50 watt whatever-it-is to browse a common website and the EOMA68 just can't compete no matter how efficiently it uses its 3 watt power input.
I misunderstood the parent post author and thought he was referring to malicious code in the compiled binaries for the EOMA68. Obviously if there is malicious code compiled into the compiler binaries, the problem is orders of magnitude harder to address. Are you suggesting we just give up on computing?
I should have worded my post differently and qualified my statements. My best guess from the sidelines is that you're approaching this from a well-informed, well-planned angle and everything I've read and watched built my trust. I pledged to support the project.
But I still think that those angles are the most likely cause of trouble. Hopefully nothing comes up or you're able to adjust for everything that does come up. I appreciate all of your free software work, by the way - I've used Samba before myself.
That's the idea, actually - a whole new computer that slots into the case.
The sacrifice is performance - if you read comments by the project founder, in order to have tiny swappable cards they're targeting something like a 3.5 watt power draw. So the device is a few generations behind the latest ARM chips, and is running at smart phone power levels. So the big wins are modularity and freedom, the big loss is that your modular 2017 mini-computer or laptop has the computing power of a 2013 smart phone.
I think the inclination for most people is to stop reading there and forget about the project. And I understand that. But if a Chromebook and a smart phone is sufficient computing power for more than half the population, then this idea genuinely could have a future. The 2017 EOMA68 with 2GB of RAM and a dual core ARM processor that launched in 2013 might be a novelty for hobbyists and fanatic FSF nerds (I am one of the latter). But the 2021 EOMA68 with maybe 6GB of RAM - which you can get today in the OnePlus Three - and a hexa-core or octo-core ARM processor from 2017? That's probably fine for 90% of the computer users in the country. Software developers, graphic artists, video editors, and PC gamers wouldn't use it. Everyone else could.
The FSF-approved Linux distribution (or GNU/Linux, whatever) "Parabola" that they offer won't include the firmware for the GPU, and does all graphics processing and calculations on the CPU. So the GPU is included on the chip but it's not used.
If the device runs open source software from top to bottom, the way to escape your malicious binaries is just to recompile everything myself. Problem solved.
Slashdot Mirror
Sorry, I meant to joke with the previous statement but upon reviewing it I came across as too snarky. I apologize, I meant to tease you and not insult you.
I think you're right. While there are a number of incredibly sophisticated free software projects - Gnu C Compiler, Linux kernel, GNOME, KDE, VLC, Postgres, etc... a lot of cutting edge stuff is proprietary and the free software equivalent comes out later and is often - though not always - inferior. Speech-to-text is one of those areas, as is Virtual Reality, and 3D Gaming.
This is what I get for reading about hosting email from public sites. I kept reading that it was all but impossible, and that messages are automatically dropped. I'll give it a try. Thanks for correcting me.
"Google Search is not done due to it requires millions of servers in order to handle the same load so the software part is of no interest since no one else have access to that amount of hardware."
No, there is interest and people are trying - Yacy is the example. The data and the service are distributed and decentralized, so it's technologically difficult to pinpoint the searches of individual users. And since there are hundreds of nodes, or at least there were, there are plenty of resources available. But even though, if I recall correctly, Yacy uses parts of the Lucene and Solr projects its search features are awful. I did a search for the cartoon Spongebob with the simple keyword "Spongebob" and got neither the official website nor the Wikipedia page in the first 50 responses.
So people care, and people are willing to contribute. But this is an incredibly difficult thing to tackle.
Yup can't be done. That would be like having a bunch of volunteers writing a really good C compiler, or a kernel. It'll never happen, right?
Eben Upton is looking at this from the perspective of someone buying CPUs to incorporate into some future version of the Raspberry Pi, not building or designing a new CPU for the Raspberry Pi or for any other use. I think, given that angle, the copyright and patent advantages of RISC-V are irrelevant to him.
I think the point you're getting at is valid in general. The PC market exploded when IBM allowed clones, and Apple and other players were crushed because the dozens of companies trying to make good products cheaply drove rapid innovation. Android launched as a piece of junk next to the iPhone and IOS, but the open nature of it let dozens or hundreds of companies use it, and the software and hardware evolved very rapidly. RISC-V could do the same thing for processors against x86 and ARM.
I disagree. I don't use Siri, or Google Now, or Cortana specifically because of the privacy implications. If I had available a completely open source, end-user-owned, private option I would use it all of the time for things like checking the weather, setting reminders, playing music, adjusting the temperature, assembling grocery lists, making purchases, etc... etc...
As an aside, your contempt for the average person won't help our cause. Elitism helps none of us - the best way to make open source stronger is to get more people using it, because some of those new users will in turn become contributors.
One of the fascinating things being tried right now is building decentralized applications on top of the cryptocurrency platforms. Ethereum, MaidSafe/Safecoin, Lisk, and others. The data is distributed in a distributed hash table (DHT). Users contribute CPU cycles, RAM usage, and disk space to the network in return for tiny portions of the respective digital currency. They're trying to build distributed autonomous organizations (DAOs), distributed alternatives to Twitter and Facebook, distributed alternatives to Dropbox and Google Drive, etc... etc... something like Siri could run on the same network.
If it ever works right. I'm hopeful but not willing to predict anything.
If I trusted the service completely, I would use it all of the time. "What's the weather tomorrow?" "Play Houses of the Holy." "Send a text to my wife asking her to get milk on the way home from work." "Find the cheapest price plus shipping of a DVD of Idiocracy on Ebay." etc... etc...
I never use Google Now/Siri/Cortana because I don't trust their respective owners with even more data about me than they already have.
Nonsense. There would be enormous use of fully open source alternatives to Google search, Gmail, Call of Duty, Starcraft 2, Destiny, and dozens of other similar projects. The best, to my knowledge, fully open source search engine is Yacy and it totally sucks. Running your own email server isn't too hard, but getting your mail to recipients on Gmail, Hotmail, or Yahoo mail without relaying through one of the big services is all but impossible. There are plenty of nice graphical fully open source video games out there, but nothing with the artwork or the voice acting or the visuals on par with a top of the line AAA game.
Nobody is making them because it's too damn difficult.
I think the critical difference has less to do with the advantages of 4k over 2k and more to do with the fact that 4k prices are dropping rapidly. I waited a long time to get a 37 inch 1080p television because for a long time it cost more than twice as much as a 37 inch 480p television. Today the price premium for 4K is already only 30-40%, sometimes less.
Interesting. I've been using Elementary OS 0.3 for over a year with months between reboots, and it never crashed.
I've been using it on one of my desktops for over a year without any problems. One headache is that they have no upgrade option between major releases beyond back up and reinstall. So I'm going to set aside a few hours to replace 0.3 Freya with 0.4 Loki.
I haven't tried it for any large scale teams, but Rocket.chat is a great open source Slack replacement for small teams. It may scale well, I just don't know.
We can sit in the corner happy with BlackBox window manager or similar, and they're functional enough to get the job done. But if you want free software to conquer the world - which I very much do - we're going to have to make something that the other 99.7% of the population wants to use for their desktop. That requires the eyecandy, sorry.
If FVWM was going to conquer the world, it would have done it twenty years ago. Again, no disrespect to fans of BlackBox or FVWM or anything else that uses the X protocol efficiently. But the goal is to make something our colleagues, friends, and family members use without hating and in my experience 1990s Unix desktop graphics don't win any fans.
I guess the study has some value. But I have two immediate questions:
1. Is immersing a hand in cold water equivalent to all other forms of pain? What if marijuana increases the average man's ability to tolerate cold hands 20% and has an insignificant affect for women in that area, but increases a woman's tolerance for bruises, or stomach cancer pain, more than it does for men?
2. In my experience, women are bothered more by cold than men in general. I realize the study is supposedly double blind, so they tried to account for that. But I think that could be a factor to include. If the average guy in the placebo portion could tolerate 90 seconds and the average woman in the placebo portion could tolerate 60, and then the average guy on pot can tolerate 120 and the average woman on pot can tolerate 63, maybe the difference in baselines is important.
Thanks for answering. I just wanted to make sure that wasn't the deciding factor.
Same video card?
The Broadwell-E part they benchmarked against is probably a (slightly underclocked) Core i7-6900k, and it's $1100.
I'm taking wild guesses with the numbers here, but "15% slower than a Broadwell-E at a 45% lower cost and a similar TDP" is a valid market strategy. I haven't spent more than $240 for a CPU in over ten years, if in spring 2017 there are Zen parts at the $250 price point that are 15% behind the 2017 spring equivalent of the Intel i7-6700k or i7-6800k I will buy it.
Don't mess with the tinfoil hat, man. (Or woman, whatever.)
Our smart phones have software running in the wireless carrier cellular modem that end users can't access or control. Almost all recent x86 laptops and desktops have Intel's TPM which the regular users can't control. Our ISPs track the websites we visit. Our credit card and debit card companies track every purchase we make. Even our grocery stores use loyalty card programs to see what we do. And browser fingerprinting tech like the "Evercookie" mean that even if you run all of the ad-blockers you want and a VPN or Tor, some companies have a detailed profile on your browsing habits on public sites.
So yeah, if free-from-top-to-bottom computing devices and discussion forums and web stores running on top of distributed decentralized platforms like Ethereum become popular, I fully expect governments to throw up roadblocks to maintain surveillance and businesses to throw up roadblocks to maintain their advertising and tracking revenue. Maybe shipments get held up for nonsense reasons in customs, maybe the projects get buried in nonsense patent lawsuits (small companies often fold under those pressures even if the plaintiff is full of shit, because the defendant can't afford the legal fees), maybe Ecmascript 8 becomes a power-hungry monster solely so that users need to buy the latest Intel/AMD 50 watt whatever-it-is to browse a common website and the EOMA68 just can't compete no matter how efficiently it uses its 3 watt power input.
I misunderstood the parent post author and thought he was referring to malicious code in the compiled binaries for the EOMA68. Obviously if there is malicious code compiled into the compiler binaries, the problem is orders of magnitude harder to address. Are you suggesting we just give up on computing?
I should have worded my post differently and qualified my statements. My best guess from the sidelines is that you're approaching this from a well-informed, well-planned angle and everything I've read and watched built my trust. I pledged to support the project.
But I still think that those angles are the most likely cause of trouble. Hopefully nothing comes up or you're able to adjust for everything that does come up. I appreciate all of your free software work, by the way - I've used Samba before myself.
Thanks for the correction, and I apologize for spreading misinformation!
That's the idea, actually - a whole new computer that slots into the case.
The sacrifice is performance - if you read comments by the project founder, in order to have tiny swappable cards they're targeting something like a 3.5 watt power draw. So the device is a few generations behind the latest ARM chips, and is running at smart phone power levels. So the big wins are modularity and freedom, the big loss is that your modular 2017 mini-computer or laptop has the computing power of a 2013 smart phone.
I think the inclination for most people is to stop reading there and forget about the project. And I understand that. But if a Chromebook and a smart phone is sufficient computing power for more than half the population, then this idea genuinely could have a future. The 2017 EOMA68 with 2GB of RAM and a dual core ARM processor that launched in 2013 might be a novelty for hobbyists and fanatic FSF nerds (I am one of the latter). But the 2021 EOMA68 with maybe 6GB of RAM - which you can get today in the OnePlus Three - and a hexa-core or octo-core ARM processor from 2017? That's probably fine for 90% of the computer users in the country. Software developers, graphic artists, video editors, and PC gamers wouldn't use it. Everyone else could.
The FSF-approved Linux distribution (or GNU/Linux, whatever) "Parabola" that they offer won't include the firmware for the GPU, and does all graphics processing and calculations on the CPU. So the GPU is included on the chip but it's not used.
If the device runs open source software from top to bottom, the way to escape your malicious binaries is just to recompile everything myself. Problem solved.