But I never expected to get anything of the company's either, because I already agreed to work for $DOLLARS per $TIME.
More importantly you were not *at risk*. You were paid for your time whether the company was profitable or not, whether the stock was up or down. Sure you can get laid off if things go really bad, been there done that, but people in that situation have generally been paid for the time they put in.
In general people prefer more security at the price of smaller payouts (salary and if lucky a modest bonus).
As I recall the plan for Free Software was to make money off of the support not the development, so things seem to be working according to the plan.;-)
The free CentOS distribution exists entirely because Red Hat Exists.
- Red Hat isn't just a middleman selling other people's work. Red Hat's employees work on - and contribute to - hundreds of different software packages. Red Hat is consistently one of the largest (and often THE largest) contributors of code to the Linux kernel, year after year.
and Red Hat contributes A LOT to open source too. If it wasn't for Red Hat there would be no "Linux" as we know it.
IBM too. Red Hat #1 and IBM #4 in terms of corporate development of Linux. All together 75% of kernel development is corporate. https://www.computerweekly.com...
Remember: not only RH pay salary for FLOSS engineers and supporters...
No, but Red Hat tops the list and IBM is #4:
"The top 10 organizations sponsoring Linux kernel development since the last report (or Linux kernel 2.6.36) are:
1. Red Hat,
2. Intel,
3. Novell,
4. IBM,
5. Texas Instruments,
6. Broadcom,
7. Nokia,
8. Samsung,
9. Oracle
10. and Google."
"... more than 7,800 developers from almost 800 different companies have contributed to the Linux kernel since tracking began in 2005. Of particular interest perhaps is the finding that — seventy-five percent of all kernel development is done by developers who are being paid for their work..."
There is another aspect people are ignoring. Linux has been corporate controlled and developed for years. A lot of work has been subsidized, and therefore directed, by various corporations. Linux is long past the point where it is primarily a "hobbyist" and "volunteer" effort.
The Linux foundation reports that 75% of kernel development is done by corporate sponsored developers. Who tops the list of these corporate sponsors? Red Hat. https://www.computerweekly.com...
Apple and 3rd parties sell wired headphones with a lighting connector. Apple did not drop wired headphones, the replaced the standard audio connector with a lightning connector.
Because a lot of the "real" work gets done at a desk, either at the office or at home, where people who care have external keyboards and monitors. The built-in keyboard is optimized for size, for people on-the-go, for short duration work. There is an inherent tradeoff that results from this. That said, a flakey built-in keyboard is something quite separate from a size optimized keyboard. Apple deserves criticism for a flakey built-in keyboard but not for optimizing for size rather than creating a typist's dream.
Personally I'm still using G4 PowerMac era external keyboards. Apple's current external keyboards are little different from laptop keyboards, however the design compromises that they embed are not justified as they are not mobile. On these externals Apple deserves criticism for being cheap and overvaluing visual fashion.
And for those who might comment that they use the built-in keyboard and monitor at work at their desk / work station, well then your employer sucks. The built-in's are great when visiting a colleague's office or for a meeting in the conference room, but your desk / workstation should come with a good chair, a good monitor and a good keyboard.
1. Its not the expense of mining that protects the integrity of the blockchain. It the distribute nature of the miners that protects the blockchain, the consensus of what the correct transactions are. The "work" in proof of work based system is merely how the mining rewards are randomized.
I used to think that was true. Turns out it's not. it's entirely the expense that prevents the double spend. Here's the proof.
take a network as distributed as you like. It has a certain total compute capacity. Now go out and rent amazon servers with four times that capacity...
Your argument is flawed. Take a network where the compute capacity is sufficiently concentrated that a 51% cartel can be formed from existing miners. The expense is irrelevant. Basically you are erroneously focused on a side effect, expense. The core condition is having a sufficiently sized and distributed network that the compute capacity necessary is not replicable. This is also an issue of difficulty, expense is a possible side effect of difficulty. You conflate the side effect, expense, with network size and distribution and work difficulty.
2. The expense of the work required for mining is actually making bitcoins less secure, increasing the risk of a manipulated blockchain. This expense has led to the development of ASIC based mining. By moving away from CPU and GPU based mining we no longer have a distributed system. ASIC based mining is far more centralized in terms of commercial mining operations so cartels are plausible. They are also highly regionalized, mostly in a single country near cheap government supplied hydroelectric power so government intervention is more plausible.
Yes the market is unstable. Not because the asics per se, but because there is a growing body of obsolete mining capacity as the hash cost rises above the reward level. When the price falls in a price swing much more of the capacity goes off line, focusing the power of hashing into fewer miners, making the prospect of buying enough capacity to overhwelm the system easier. Asics are just a symtom of this thin margin on an expensive operation not so much the cause of this.
Here too you conflate the side effect, expense, with the core conditions of network size and work difficulty. You also ignore the core problem that a focus on expense increases the obsolescence of mining capacity, reducing network size and distribution. Focusing on the expense side effect increases the risk to blockchain security, the focus should be on network size and distribution.
With respect to the bitcoin fork, the "fix", it would have to be an ASIC-resistant algorithm unrelated to SHA-256. If so existing ASIC miners will not be able to jump in and dominate, their existing high performance expensive hardware would be unusable.
1. Its not the expense of mining that protects the integrity of the blockchain. It the distribute nature of the miners that protects the blockchain, the consensus of what the correct transactions are. The "work" in proof of work based system is merely how the mining rewards are randomized.
2. The expense of the work required for mining is actually making bitcoins less secure, increasing the risk of a manipulated blockchain. This expense has led to the development of ASIC based mining. By moving away from CPU and GPU based mining we no longer have a distributed system. ASIC based mining is far more centralized in terms of commercial mining operations so cartels are plausible. They are also highly regionalized, mostly in a single country near cheap government supplied hydroelectric power so government intervention is more plausible.
Now I am not saying that cartels or government intervention is likely, the point is that the **reality** of bitcoin has deviated from the **theory** of bitcoin. In theory cartels and government could do nothing because the blockchain is maintained by ordinary users and their ordinary CPUs/GPUs and a 51% attack would be implausible. But with mining consolidate both in terms of commercial operations and regionally, cartel and government manipulation is increasingly plausible.
Bitcoin is at risk because it has deviated from its design, a system of distributed blockchain maintenance and consensus, the distributed nature lost.
Fortunately this can be fixed with a software update, one that brings mining back to the realm of CPUs and GPUs, ie regular users and their regular computers. Unfortunately miners would have to agree with that update and why would ASIC miners do that? So if bitcoin is to be "fixed" it will have to fork and be a different type of coin. The blockchain could fork with the code so no holdings are lost.
People aren't buying as many traditional personal computers.
While there is some truth to the idea that some activities are more convenient on a phone or tablet, I think the larger culprit in declining PC sales is their increased longevity. We are long past the point where the computational power of a PC has exceeded the needs of many users, where a new computer has no perceptible performance increase over a three year old computer for many users. Now granted I installed ample RAM in my 8 year old PC when I built it but it is still a useful machine, even for many video games with a video card upgrade every 2-3 years.
Phones are now personal computers--while it can only do about 50% of what a personal computer can do, it can do 100% of what most people want to do with their personal computer.
Perhaps "many" not "most". For longer endurance activities, outside of gaming, larger screens and real keyboards are more necessary. Phones and PCs will likely remain complementary devices, in the developed world people will likely continue to have both. Tablets and PCs, there we may have convergence, a "laptop" becoming a "dock" and a detachable "screen".
My Mom is no longer subsidizing my cheap hardware by buying the three year-old model of what I bought.
Similar story in my family. The "retired" folks who just wants email, Skype, web browsing and online shopping in moderate proportions is finding a tablet quite satisfactory. And this includes people who had used computers for many years at work.
But for people in school or still working, I think PCs will be hard to replace with tablets.
They are just a different form factor for convenience when performing different tasks. PCs, tablets, and smartphones are destined for convergence.
Those two statements appear to be at opposition with each other. The fact that these devices have different form factors to conveniently perform different tasks is why they will not converge. They will continue to be very similar, and perhaps the OS and apps on the devices will converge, but ultimately there is a reason each form factor exists.
I'm leaning towards the hardware converging but the software not. For example a tablet can simply be a removable laptop screen, or if you want to look from the "other" side laptops become docks for tablets. To use Apple as an example I would expect the screen to run iOS when undocked and macOS when docked.
Smartphone being part of the convergence is trickier given the pocket sized requirement. For longer work sessions that laptop sized screen would seem a necessity. There might be convergence in the sense that one day nearly all laptops and tablets may have cellular capability but I expect two devices to persist, one a 24/7 pocket companion, one for more serious work sessions.
The point of my post is that JSF software problems are not necessarily evidence of language based coding difficulties.
Regarding your question of adding coding problems into the mix, the answer is a common one. Tradeoffs. For performance reasons you may need to use C. There is no universal answer to what language to use, its a matter of best fit, and sometimes that best fit is C. I realize some might argue otherwise, but I've also noticed that many people merely argue for the language they are most familiar with.
They used to, sometimes. But modern dictators ain't what they used to be.;-)
In the Roman Republic (emphasize Republic, after the kings, before the emperors) the dictator had a temporary appointment and absolute authority limited to the territory in crisis, for example a region with active warfare. An interesting story:
Rome was invaded. The Senate appointed a man named Cincinnatus dictator for six months. On his first day he appointed a military commander and ordered all able bodied males in Rome to report for military service. The next day they marched to meet the enemy. He outmaneuvered the enemy and put them in a very bad position, they begged for mercy. The deal was to execute the top three enemy leaders and grant amnesty to the bulk of the enemy army. Cincinnatus then disbanded his Roman army and resigned the dictatorship. He was dictator for about two weeks and then returned to his farm outside of Rome.
Well, this isn't quite the same comment, but if the language is compatible with C, or some subset of C, couldn't you compile the "safe version", run your tests, and then, when you were satisfied, compile with standard C? Surely the answers ought to be guaranteed to be the same if there's no error.
Only for real world inputs that match your test inputs. If you compile with standard C you lose the run time checks, array bounds for example. If these check only have a 1% penalty then for many apps that might be quite acceptable.
How about these explanations: 1) If a user has a wierd problem, a professional developer has to fix it promptly or give a refund. An open source developer doesn't have to do anything.
The open source developer is free to say: you have the source code, fix it yourself. Or hire me to fix it, or hire someone else to fix it.
Free open-source software supporter by a small number of dev without monetary incensive can do it. So now, how can you explain that a profitable company can't ?
You answered your own question: "devs without monetary incentive" vs "profitable company". The former are unconcerned with their efforts being profitable. The ability to do something is a secondary issue for the later, profitability is the primary issue.
I'd believe this, except for the whole android is linux argument.
Android is not Linux, it is merely hosted on Linux. If Android provides sufficient APIs an app doesn't know or care about Linux, the hosting kernel underneath could be switched to BSD (or Fuchsia someday) and the Android app would not care. Admittedly, dropbox probably does use the NDK but...
I don't see dropbox abandoning android support in the summary.
If dropbox needs to use the NDK and access Linux directly there is only one distribution to be compatible with. That is quite different than supporting desktop Linux in general.
If google started deleting or blocking the apps that I choose to sideload onto my android, I would ditch them ASAP and join the class action lawsuit which would surely follow.
If they were doing so only in cases of genuine malware, you would lose. And the terms of service could easily be updated to permit this if not already allowed to further weaken your actions.
Slashdot Mirror
But I never expected to get anything of the company's either, because I already agreed to work for $DOLLARS per $TIME.
More importantly you were not *at risk*. You were paid for your time whether the company was profitable or not, whether the stock was up or down. Sure you can get laid off if things go really bad, been there done that, but people in that situation have generally been paid for the time they put in.
In general people prefer more security at the price of smaller payouts (salary and if lucky a modest bonus).
As I recall the plan for Free Software was to make money off of the support not the development, so things seem to be working according to the plan. ;-)
The free CentOS distribution exists entirely because Red Hat Exists.
- Red Hat isn't just a middleman selling other people's work. Red Hat's employees work on - and contribute to - hundreds of different software packages. Red Hat is consistently one of the largest (and often THE largest) contributors of code to the Linux kernel, year after year.
And IBM is #4. 75% of kernel dev is corporate.
https://www.computerweekly.com...
Um, I don't know what crazy world you live on, but we pay our scientists.
And 75% of Linux kernel development is paid for by corporations.
https://www.computerweekly.com...
and Red Hat contributes A LOT to open source too. If it wasn't for Red Hat there would be no "Linux" as we know it.
IBM too. Red Hat #1 and IBM #4 in terms of corporate development of Linux. All together 75% of kernel development is corporate.
https://www.computerweekly.com...
Remember: not only RH pay salary for FLOSS engineers and supporters...
No, but Red Hat tops the list and IBM is #4:
..."
"The top 10 organizations sponsoring Linux kernel development since the last report (or Linux kernel 2.6.36) are:
1. Red Hat,
2. Intel,
3. Novell,
4. IBM,
5. Texas Instruments,
6. Broadcom,
7. Nokia,
8. Samsung,
9. Oracle
10. and Google."
"... more than 7,800 developers from almost 800 different companies have contributed to the Linux kernel since tracking began in 2005. Of particular interest perhaps is the finding that — seventy-five percent of all kernel development is done by developers who are being paid for their work
https://www.computerweekly.com...
There is another aspect people are ignoring. Linux has been corporate controlled and developed for years. A lot of work has been subsidized, and therefore directed, by various corporations. Linux is long past the point where it is primarily a "hobbyist" and "volunteer" effort.
The Linux foundation reports that 75% of kernel development is done by corporate sponsored developers. Who tops the list of these corporate sponsors? Red Hat.
https://www.computerweekly.com...
Apple and 3rd parties sell wired headphones with a lighting connector. Apple did not drop wired headphones, the replaced the standard audio connector with a lightning connector.
Because a lot of the "real" work gets done at a desk, either at the office or at home, where people who care have external keyboards and monitors. The built-in keyboard is optimized for size, for people on-the-go, for short duration work. There is an inherent tradeoff that results from this. That said, a flakey built-in keyboard is something quite separate from a size optimized keyboard. Apple deserves criticism for a flakey built-in keyboard but not for optimizing for size rather than creating a typist's dream.
Personally I'm still using G4 PowerMac era external keyboards. Apple's current external keyboards are little different from laptop keyboards, however the design compromises that they embed are not justified as they are not mobile. On these externals Apple deserves criticism for being cheap and overvaluing visual fashion.
And for those who might comment that they use the built-in keyboard and monitor at work at their desk / work station, well then your employer sucks. The built-in's are great when visiting a colleague's office or for a meeting in the conference room, but your desk / workstation should come with a good chair, a good monitor and a good keyboard.
Two flaws in your comment.
1. Its not the expense of mining that protects the integrity of the blockchain. It the distribute nature of the miners that protects the blockchain, the consensus of what the correct transactions are. The "work" in proof of work based system is merely how the mining rewards are randomized.
I used to think that was true. Turns out it's not. it's entirely the expense that prevents the double spend. Here's the proof. take a network as distributed as you like. It has a certain total compute capacity. Now go out and rent amazon servers with four times that capacity ...
Your argument is flawed. Take a network where the compute capacity is sufficiently concentrated that a 51% cartel can be formed from existing miners. The expense is irrelevant. Basically you are erroneously focused on a side effect, expense. The core condition is having a sufficiently sized and distributed network that the compute capacity necessary is not replicable. This is also an issue of difficulty, expense is a possible side effect of difficulty. You conflate the side effect, expense, with network size and distribution and work difficulty.
2. The expense of the work required for mining is actually making bitcoins less secure, increasing the risk of a manipulated blockchain. This expense has led to the development of ASIC based mining. By moving away from CPU and GPU based mining we no longer have a distributed system. ASIC based mining is far more centralized in terms of commercial mining operations so cartels are plausible. They are also highly regionalized, mostly in a single country near cheap government supplied hydroelectric power so government intervention is more plausible.
Yes the market is unstable. Not because the asics per se, but because there is a growing body of obsolete mining capacity as the hash cost rises above the reward level. When the price falls in a price swing much more of the capacity goes off line, focusing the power of hashing into fewer miners, making the prospect of buying enough capacity to overhwelm the system easier. Asics are just a symtom of this thin margin on an expensive operation not so much the cause of this.
Here too you conflate the side effect, expense, with the core conditions of network size and work difficulty. You also ignore the core problem that a focus on expense increases the obsolescence of mining capacity, reducing network size and distribution. Focusing on the expense side effect increases the risk to blockchain security, the focus should be on network size and distribution.
With respect to the bitcoin fork, the "fix", it would have to be an ASIC-resistant algorithm unrelated to SHA-256. If so existing ASIC miners will not be able to jump in and dominate, their existing high performance expensive hardware would be unusable.
Two flaws in your comment.
1. Its not the expense of mining that protects the integrity of the blockchain. It the distribute nature of the miners that protects the blockchain, the consensus of what the correct transactions are. The "work" in proof of work based system is merely how the mining rewards are randomized.
2. The expense of the work required for mining is actually making bitcoins less secure, increasing the risk of a manipulated blockchain. This expense has led to the development of ASIC based mining. By moving away from CPU and GPU based mining we no longer have a distributed system. ASIC based mining is far more centralized in terms of commercial mining operations so cartels are plausible. They are also highly regionalized, mostly in a single country near cheap government supplied hydroelectric power so government intervention is more plausible. Now I am not saying that cartels or government intervention is likely, the point is that the **reality** of bitcoin has deviated from the **theory** of bitcoin. In theory cartels and government could do nothing because the blockchain is maintained by ordinary users and their ordinary CPUs/GPUs and a 51% attack would be implausible. But with mining consolidate both in terms of commercial operations and regionally, cartel and government manipulation is increasingly plausible.
Bitcoin is at risk because it has deviated from its design, a system of distributed blockchain maintenance and consensus, the distributed nature lost.
Fortunately this can be fixed with a software update, one that brings mining back to the realm of CPUs and GPUs, ie regular users and their regular computers. Unfortunately miners would have to agree with that update and why would ASIC miners do that? So if bitcoin is to be "fixed" it will have to fork and be a different type of coin. The blockchain could fork with the code so no holdings are lost.
People aren't buying as many traditional personal computers.
While there is some truth to the idea that some activities are more convenient on a phone or tablet, I think the larger culprit in declining PC sales is their increased longevity. We are long past the point where the computational power of a PC has exceeded the needs of many users, where a new computer has no perceptible performance increase over a three year old computer for many users. Now granted I installed ample RAM in my 8 year old PC when I built it but it is still a useful machine, even for many video games with a video card upgrade every 2-3 years.
Phones are now personal computers--while it can only do about 50% of what a personal computer can do, it can do 100% of what most people want to do with their personal computer.
Perhaps "many" not "most". For longer endurance activities, outside of gaming, larger screens and real keyboards are more necessary. Phones and PCs will likely remain complementary devices, in the developed world people will likely continue to have both. Tablets and PCs, there we may have convergence, a "laptop" becoming a "dock" and a detachable "screen".
My Mom is no longer subsidizing my cheap hardware by buying the three year-old model of what I bought.
Similar story in my family. The "retired" folks who just wants email, Skype, web browsing and online shopping in moderate proportions is finding a tablet quite satisfactory. And this includes people who had used computers for many years at work.
But for people in school or still working, I think PCs will be hard to replace with tablets.
They are just a different form factor for convenience when performing different tasks. PCs, tablets, and smartphones are destined for convergence.
Those two statements appear to be at opposition with each other. The fact that these devices have different form factors to conveniently perform different tasks is why they will not converge. They will continue to be very similar, and perhaps the OS and apps on the devices will converge, but ultimately there is a reason each form factor exists.
I'm leaning towards the hardware converging but the software not. For example a tablet can simply be a removable laptop screen, or if you want to look from the "other" side laptops become docks for tablets. To use Apple as an example I would expect the screen to run iOS when undocked and macOS when docked.
... :-)
Smartphone being part of the convergence is trickier given the pocket sized requirement. For longer work sessions that laptop sized screen would seem a necessity. There might be convergence in the sense that one day nearly all laptops and tablets may have cellular capability but I expect two devices to persist, one a 24/7 pocket companion, one for more serious work sessions.
Oh damn, we forgot the watch
To be fair, anybody investing their student loan in crypto-currencies would probably not have benefited from an academic education anyways....
You sure, isn't the "greater fool" theory likely covered in microeconomics 101?
The point of my post is that JSF software problems are not necessarily evidence of language based coding difficulties.
Regarding your question of adding coding problems into the mix, the answer is a common one. Tradeoffs. For performance reasons you may need to use C. There is no universal answer to what language to use, its a matter of best fit, and sometimes that best fit is C. I realize some might argue otherwise, but I've also noticed that many people merely argue for the language they are most familiar with.
Dictators do not work for industry or countries.
They used to, sometimes. But modern dictators ain't what they used to be. ;-)
In the Roman Republic (emphasize Republic, after the kings, before the emperors) the dictator had a temporary appointment and absolute authority limited to the territory in crisis, for example a region with active warfare. An interesting story:
Rome was invaded. The Senate appointed a man named Cincinnatus dictator for six months. On his first day he appointed a military commander and ordered all able bodied males in Rome to report for military service. The next day they marched to meet the enemy. He outmaneuvered the enemy and put them in a very bad position, they begged for mercy. The deal was to execute the top three enemy leaders and grant amnesty to the bulk of the enemy army. Cincinnatus then disbanded his Roman army and resigned the dictatorship. He was dictator for about two weeks and then returned to his farm outside of Rome.
https://en.wikipedia.org/wiki/...
Bugs are not always coding errors, bugs may also be in the design and correct code of any language can manifest such design bugs
Well, this isn't quite the same comment, but if the language is compatible with C, or some subset of C, couldn't you compile the "safe version", run your tests, and then, when you were satisfied, compile with standard C? Surely the answers ought to be guaranteed to be the same if there's no error.
Only for real world inputs that match your test inputs. If you compile with standard C you lose the run time checks, array bounds for example. If these check only have a 1% penalty then for many apps that might be quite acceptable.
Its not just eventual mortality. There is also after 27 years that after work hobby may not be as much fun as it used to be.
;-)
Maybe he wants a new hobby, perhaps give microkernels a try.
He gave them two hints:
(1) I got my dates mixed up and I cannot attend due to prior commitments.
(2) Don't reschedule, just carry on without me.
Don't you guys get it? He may want the community to be less dependent, more able to stand on its own, not need the hand holding.
How about these explanations: 1) If a user has a wierd problem, a professional developer has to fix it promptly or give a refund. An open source developer doesn't have to do anything.
The open source developer is free to say: you have the source code, fix it yourself. Or hire me to fix it, or hire someone else to fix it.
Free open-source software supporter by a small number of dev without monetary incensive can do it. So now, how can you explain that a profitable company can't ?
You answered your own question: "devs without monetary incentive" vs "profitable company". The former are unconcerned with their efforts being profitable. The ability to do something is a secondary issue for the later, profitability is the primary issue.
I'd believe this, except for the whole android is linux argument.
Android is not Linux, it is merely hosted on Linux. If Android provides sufficient APIs an app doesn't know or care about Linux, the hosting kernel underneath could be switched to BSD (or Fuchsia someday) and the Android app would not care. Admittedly, dropbox probably does use the NDK but ...
I don't see dropbox abandoning android support in the summary.
If dropbox needs to use the NDK and access Linux directly there is only one distribution to be compatible with. That is quite different than supporting desktop Linux in general.
If google started deleting or blocking the apps that I choose to sideload onto my android, I would ditch them ASAP and join the class action lawsuit which would surely follow.
If they were doing so only in cases of genuine malware, you would lose. And the terms of service could easily be updated to permit this if not already allowed to further weaken your actions.