Google ups its Linux Foundation membership to the $500,000/year Platinum level(techcrunch.com)
techcrunch.com
Google ups its Linux Foundation membership to the $500,000/year Platinum level
https://techcrunch.com/2018/06/27/google-ups-its-linux-foundation-membership-to-the-500000-year-platinum-level/
126 comments
The Linux Foundation is an umbrella organisation/lobby organisation, similar to the Apache project.
Most of the actual code is written by individual and corporate contributors. Google, in particular, contributes a LOT of code to both the Linux kernel and projects like Kubernetes.
Most of the actual code is written by individual and corporate contributors. Google, in particular, contributes a LOT of code to both the Linux kernel and projects like Kubernetes.
There is a major difference between the Apache Software Foundation and the Linux Foundation. The ASF is a 501(c)(3) charitable organization. The Linux Foundation is a 501(c)(6) business league.
Donations to a 501(c)(3) are tax-deductible but are very restricted, which donations to a 501(c)(6) are not tax-deductible but are far less restricted. Not coincidentally, the ASF's annual budget is a tiny fraction of the Linux Foundation's.
Donations to a 501(c)(3) are tax-deductible but are very restricted, which donations to a 501(c)(6) are not tax-deductible but are far less restricted. Not coincidentally, the ASF's annual budget is a tiny fraction of the Linux Foundation's.
See also Microsoft / clang
I thought Apple were the ones originally behind Clang?
What?
Serious question, what would the Linux foundation do with 5bn/year?
Plenty. E.g. pay for critical infrastructure projects[1][2]—OpenSSL was "manned" by just one or two developers who were grinding themselves to the ground. People barely realized it until "Heartbleed" CVE. Other things such as fund open source internships, improve CI and testing for relevant upstream projects, pay technical writers to make documentation top-notch, hardware resources, and so forth.
[1] https://www.coreinfrastructure.org/ [2] https://lwn.net/Articles/600571/
[1] https://www.coreinfrastructure.org/ [2] https://lwn.net/Articles/600571/
> People barely realized it until "Heartbleed" CVE
at which point they proceeded to shit on those devs for being so awful at maintaining a critical project. It was nasty and unwarranted.
at which point they proceeded to shit on those devs for being so awful at maintaining a critical project. It was nasty and unwarranted.
All the replies I got seem to address the question "What would LF do with $500MM/yr", not the question of "what would it do with the remaining 4.5bn".
Okay, here is something potentially bold (given Intel is also a Platinum Member of the Linux Foundation):
How about exploring the frontiers of open hardware (RISC-V)?
The back-of-the-envelope calculation from an LWN comment[+] (on the discussion "Is it time for open processors?") by Jon Masters, jcm, of Red Hat (but speaking for himself):
[quote] The problem with an "open" processor is the cost. It costs over $1.2 Billion to do a ground up 4 year OoO core of the kind of competitive performance that others have built. And even then, once silicon is deployed, security issues can still be found. In short, it will never happen. You'll get RISC-V cores like BOOM, you'll see some wonderful IoT designs, but you'll never see a high end Xeon-class core unless some billionaire funds it as a pet project, and keeps investing year after year for the greater good. And no, none of the major search/cloud vendors are going to go fund this - they want to make commercial vendors compete. [/quote]
I think it's a really "long shot", but we can dream. :-)
[+] https://lwn.net/Articles/743602/
How about exploring the frontiers of open hardware (RISC-V)?
The back-of-the-envelope calculation from an LWN comment[+] (on the discussion "Is it time for open processors?") by Jon Masters, jcm, of Red Hat (but speaking for himself):
[quote] The problem with an "open" processor is the cost. It costs over $1.2 Billion to do a ground up 4 year OoO core of the kind of competitive performance that others have built. And even then, once silicon is deployed, security issues can still be found. In short, it will never happen. You'll get RISC-V cores like BOOM, you'll see some wonderful IoT designs, but you'll never see a high end Xeon-class core unless some billionaire funds it as a pet project, and keeps investing year after year for the greater good. And no, none of the major search/cloud vendors are going to go fund this - they want to make commercial vendors compete. [/quote]
I think it's a really "long shot", but we can dream. :-)
[+] https://lwn.net/Articles/743602/
Love to see Linux Foundation to start do kernel equivalent projects for GPU.
GL driver, openGL Lib, etc all open source for all GPU.
Manage contributions from Intel, AMD, ARM, NVidia and other GPU silicon IP vendors.
The simple goal of opensource, advance the GPU driver, libraries for the benefits of everyone in the field.
After that, maybe define, manage the evolution of "AI" driver, library layers between the TensorFlow (or other AI framework) and the CPU, GPU, TPU.
GL driver, openGL Lib, etc all open source for all GPU.
Manage contributions from Intel, AMD, ARM, NVidia and other GPU silicon IP vendors.
The simple goal of opensource, advance the GPU driver, libraries for the benefits of everyone in the field.
After that, maybe define, manage the evolution of "AI" driver, library layers between the TensorFlow (or other AI framework) and the CPU, GPU, TPU.
These all already exist in some form. Mesa plus in-tree drivers.
Probably R&D like Mozilla did with Rust and Quantum. (Long-term pay off)
Things they can't do without 5bn/year.
Sometimes you can think about fixing problems, only when you have the resources to do it.
Alternatively, its also possible the administration can go corrupt when there is so much money flying around. Non-profits can do mismanagement their funds.
Alternatively, its also possible the administration can go corrupt when there is so much money flying around. Non-profits can do mismanagement their funds.
$500M in charitable donations to Haiti through the Red Cross seems to have gone missing: https://news.ycombinator.com/item?id=15326335
The average person deserves to know how badly charity, foreign aid, and government programs can often be at improving human welfare. Not always, but often.
The average person deserves to know how badly charity, foreign aid, and government programs can often be at improving human welfare. Not always, but often.
Very true[1] but I would suggest that the Red Cross is unusually unreliable (I always wince when I see people earnestly recommending them, there are hundreds of NGOs which are more efficient and transparent).
1: https://www.npr.org/2015/06/03/411524156/in-search-of-the-re...
1: https://www.npr.org/2015/06/03/411524156/in-search-of-the-re...
You're right that it should be a consideration, but Red Cross is an egregious example. Any human organization is inefficient and leaky, though. That doesn't mean we have a significantly better way to fix unprofitable problems than the types of orgs you mentioned.
It would probably improve the code quality quite a bit, but mostly I think they'd use the money for other projects, which is something that happens with big companies, too. The reason you see Google, Microsoft, Amazon, and Apple get into "everything" is because they make so much more money than they need for their primary businesses.
> The reason you see Google, Microsoft, Amazon, and Apple get into "everything" is because they make so much more money than they need for their primary businesses.
The reason? No, at least one major reason is they need to diversify their revenue streams. Google for example has been desperately trying to find something to offset their lopsided dependence on ad revenue for ages, because their party will come to a grinding halt real quick if they lose a large share of the eyeballs. Why do you think Google has given so much money to Apple to stay on their platform as the search provider? It's not a good position to be in.
The reason? No, at least one major reason is they need to diversify their revenue streams. Google for example has been desperately trying to find something to offset their lopsided dependence on ad revenue for ages, because their party will come to a grinding halt real quick if they lose a large share of the eyeballs. Why do you think Google has given so much money to Apple to stay on their platform as the search provider? It's not a good position to be in.
> It's not a good position to be in.
I mean, I get what you're saying... but honestly making 15b profit per quarter is really a pretty good position to be in.
I mean, I get what you're saying... but honestly making 15b profit per quarter is really a pretty good position to be in.
Target code needed by municipalities.
A long while back, my city paid for a billing system. Was something like 30, 40 million bucks. They did not own the code, or if they did, were not in a position to make that the benefit it should have been.
Long story short, they ended up paying a lot more for support, updates, bug fixes and the like. And then did it again to finally arrive at something workable!
A group of people, myself included took this to government as an example where open code could be a much better investment.
For that same sum, an entity could be created to develop that code, making for local jobs, local revenue and a very likely savings over the long life possible.
Maintenance fees could provide the entity with the staff needed to maintain that code, until it pays for itself.
Additionally, that entity could make it available to other municipalities, and or consult on that problem domain. That turns what us currently a cost center into revenue that could, at a minimum, lower the cost of the utility currently paying way too much for software not really adding that much value.
We got bipartisan support. Industry lobbyists killed it off back then. Our bad. At the time, the thrust was to get OSS written into law as a must consider. The industry arguments and ability to lobby won the day, and were compelling to legislstors who were not well equipped to really grok the potential benefits.
What we shouod have done was push for and more fully develop the plans and non profit model needed to sell it as a project, or investment.
This all happened in the early 00's too. Its a bit different and even more plausible today, and legislators are more likely to understand it all better too.
Open code would mean other projects could build on that effort, and open data would mean the business of the people would be available where and how needed.
A project of this kind would end up sustainable and would yield consistent savings to every municipality using it, whether they developed it or not.
Once done, its a model for similar needs and costs and could be repeated anywhere. An umbrella foundation could unify these things, or several sprout up, each owning a needed domain, each making revenue and ending up self sustaining.
Over time, there would be various entities of this kind solving not sexy, but quite expensive and painful problems that really do need solving.
This kind of thing is not well matched to investors and founders looking to make big returns. The primary value is making public works, utilities, etc. lean, efficient.
It's also not well matched to proprietary software houses, who may well deliver great solutions, but at high margins and with closed and often expensive dependencies. Their solutions could be a bust too.
It's high risk.
Oregon vs Oracle comes to mind.
This kind of thing is well matched to a non profit type effort, or foundation and could easily be seen as a public work, of benefit to everyone, in that taxes, the cost of running the basics in society would be reduced for everyone.
Cheaper, efficient government basically.
Redirect the savings into things like potholes, schools, health care maybe.
With these kinds of resources, or even a modest fraction, a whole lot of good can be done.
Pair developers with domain experts and legislators, city council people and solve some problems. Each solution woukd likely displace closed, expensive software and doing that would be a public good where it's most needed.
While doing this would not make the Microsofts, Oracles of the world happy, it would mean a lot of people get to make a great living, and we all see the benefit of a more lean, efficient use of taxes actually adding value instead of mostly cost to our lives.
A long while back, my city paid for a billing system. Was something like 30, 40 million bucks. They did not own the code, or if they did, were not in a position to make that the benefit it should have been.
Long story short, they ended up paying a lot more for support, updates, bug fixes and the like. And then did it again to finally arrive at something workable!
A group of people, myself included took this to government as an example where open code could be a much better investment.
For that same sum, an entity could be created to develop that code, making for local jobs, local revenue and a very likely savings over the long life possible.
Maintenance fees could provide the entity with the staff needed to maintain that code, until it pays for itself.
Additionally, that entity could make it available to other municipalities, and or consult on that problem domain. That turns what us currently a cost center into revenue that could, at a minimum, lower the cost of the utility currently paying way too much for software not really adding that much value.
We got bipartisan support. Industry lobbyists killed it off back then. Our bad. At the time, the thrust was to get OSS written into law as a must consider. The industry arguments and ability to lobby won the day, and were compelling to legislstors who were not well equipped to really grok the potential benefits.
What we shouod have done was push for and more fully develop the plans and non profit model needed to sell it as a project, or investment.
This all happened in the early 00's too. Its a bit different and even more plausible today, and legislators are more likely to understand it all better too.
Open code would mean other projects could build on that effort, and open data would mean the business of the people would be available where and how needed.
A project of this kind would end up sustainable and would yield consistent savings to every municipality using it, whether they developed it or not.
Once done, its a model for similar needs and costs and could be repeated anywhere. An umbrella foundation could unify these things, or several sprout up, each owning a needed domain, each making revenue and ending up self sustaining.
Over time, there would be various entities of this kind solving not sexy, but quite expensive and painful problems that really do need solving.
This kind of thing is not well matched to investors and founders looking to make big returns. The primary value is making public works, utilities, etc. lean, efficient.
It's also not well matched to proprietary software houses, who may well deliver great solutions, but at high margins and with closed and often expensive dependencies. Their solutions could be a bust too.
It's high risk.
Oregon vs Oracle comes to mind.
This kind of thing is well matched to a non profit type effort, or foundation and could easily be seen as a public work, of benefit to everyone, in that taxes, the cost of running the basics in society would be reduced for everyone.
Cheaper, efficient government basically.
Redirect the savings into things like potholes, schools, health care maybe.
With these kinds of resources, or even a modest fraction, a whole lot of good can be done.
Pair developers with domain experts and legislators, city council people and solve some problems. Each solution woukd likely displace closed, expensive software and doing that would be a public good where it's most needed.
While doing this would not make the Microsofts, Oracles of the world happy, it would mean a lot of people get to make a great living, and we all see the benefit of a more lean, efficient use of taxes actually adding value instead of mostly cost to our lives.
It's not the code, it's the red tape. That's where the expense comes in on that kind of stuff.
Yes, there are a number of compliance issues, and a level bureaucracy to deal with. But that's not germane to this discussion. We've got those in play no matter what. The real game here is setting up an entity to develop the code. Maybe even at similar cost. The savings is in the ownership of the code over long periods of time, and that other municipalities could then use the code.
Not to mention the numerous opportunities to take what that energy is learned in built and apply it all over the place Consulting, handling special cases and what not.
Not to mention the numerous opportunities to take what that energy is learned in built and apply it all over the place Consulting, handling special cases and what not.
Governments mandating OS code and ownership by the people could go a long way.
Thanks was a great story. Thanks for sharing.
Thanks was a great story. Thanks for sharing.
The mandating of open code is a major league sticking point.
I think personally, mandating open data is probably easier, and could be successful.
Industry lobbies will beat that open code mandate back hard. And I'm not sure we can win on that basis even today. Open data however, is a no-brainer.
I think the real answer is to just compete. If one of these projects were started, using the model that I put in my other comment, it would compete very well. I don't think it's even really a contest. And from there the savings and the economic forces would push it where it needs to go.
That would set the market opposition to the side, where it really should be. It may well be that closed proprietary solutions really are the best, and this would be one way to find that out, and find out where it's appropriate and where it's not. That's a big win for everyone.
I think personally, mandating open data is probably easier, and could be successful.
Industry lobbies will beat that open code mandate back hard. And I'm not sure we can win on that basis even today. Open data however, is a no-brainer.
I think the real answer is to just compete. If one of these projects were started, using the model that I put in my other comment, it would compete very well. I don't think it's even really a contest. And from there the savings and the economic forces would push it where it needs to go.
That would set the market opposition to the side, where it really should be. It may well be that closed proprietary solutions really are the best, and this would be one way to find that out, and find out where it's appropriate and where it's not. That's a big win for everyone.
I would love to see it happen. And like I said, I think the whole thing is a huge win, and a great living for a bunch of developers and subject matter experts.
It would do the same as Mozilla probably. It would refocus to try to fix the world and social problems while drifting away from its core products.
With some of the money, they could fund a team of people to work on system integration, and device-driver reliability for desktop use-cases.
Advertise for Microsoft all of the great open source work they've contributed to a proprietary OS.
You're grossly misinformed about the importance of corporate sponsorship and involvement.
1. Google's use of Linux legitimizes Linux (further)
2. Corporations have contributed far more to Linux's source code than individual volunteers
3. As much as you'd like to paint Linux as a socialist, anti-corporate savior, it's symbiotic relationship with capitalism is what makes it thrive and relevant.
1. Google's use of Linux legitimizes Linux (further)
2. Corporations have contributed far more to Linux's source code than individual volunteers
3. As much as you'd like to paint Linux as a socialist, anti-corporate savior, it's symbiotic relationship with capitalism is what makes it thrive and relevant.
>As much as you'd like to paint Linux as a socialist, anti-corporate savior, it's symbiotic relationship with capitalism is what makes it thrive and relevant.
I have been thinking whether this model is purely because of GPL, or would it have been the same if it was Apache 2.0 or other more permissive license?
Or was it because it was so widely used you cant ignore it.
I have been thinking whether this model is purely because of GPL, or would it have been the same if it was Apache 2.0 or other more permissive license?
Or was it because it was so widely used you cant ignore it.
> Something has to be said for people who pour their hearts and lives to making a great OS project that's used under the hood almost everywhere.
I'm sure many of those developers have an email @google.com.
I'm sure many of those developers have an email @google.com.
I think Torvalds is worshiped well enough
A billion+/year would be reasonable, considering what they've done with Android.
Microsoft is routinely the biggest contributor to the Linux source, so it's not like big companies are just getting the stuff for nothing.
I keep hearing this, but the "biggest" contribution was Hyper-V drivers so that Linux could run on Hyper-V. They came in, dumped some bloated drivers, then left and became unresponsive.
Microsoft and other big companies are literally getting the stuff for zero dollars.
That they choose to contribute is because it is in their best interest to do so.
Free software is sort of a brain hack in that way: because Linux is successful, it is in Microsoft's best interest to contribute. But that's very different than saying, for example, that Microsoft owes dues and pays them.
That they choose to contribute is because it is in their best interest to do so.
Free software is sort of a brain hack in that way: because Linux is successful, it is in Microsoft's best interest to contribute. But that's very different than saying, for example, that Microsoft owes dues and pays them.
The Linux Foundation is a corporate PR joke. Even Oracle, slayer of FLOSS projects for decades, and vmware, violating the GPL for the Linux kernel for years with ESX, are 'platinum members'[1].
At this point, any membership in this 'foundation' should be viewed a shallow attempt at marketing.
1. https://www.linuxfoundation.org/membership/members/
At this point, any membership in this 'foundation' should be viewed a shallow attempt at marketing.
1. https://www.linuxfoundation.org/membership/members/
The Linux Foundation does do some pretty useful work. For instance, iirc, they facilitate much of the work on Let's Encrypt.
Yeah, I agree. FWIW, I've been participating in Linux Foundation's technical events since late 2012. Not everything they do is golden (w.r.t coordinating with open source communities or partnerships with companies), but for now their strengths outweigh their weaknesses.
AFAIK, they provide only administrative support (payroll, HR, etc). That's not really "much of" the work on Let's Encrypt.
There's a whole lot of other facilitating not included in that.. like infrastructure (hosting, development, test), management, technical support, etc.
There's a whole lot of other facilitating not included in that.. like infrastructure (hosting, development, test), management, technical support, etc.
I don't understand the angst, are you saying they should not accept any donation for those companies?
They should accept the donation but perhaps think twice about giving away board seats for them
They should not grant membership to companies that have a demonstrated history of being hostile to FLOSS. But hey, who would say no to more money, amirite?
>But hey, who would say no to more money, amirite?
Yes, you are, and luckily those in charge of the foundation are capable of understanding that inclusion (and money) serve a greater good than does being spiteful.
Yes, you are, and luckily those in charge of the foundation are capable of understanding that inclusion (and money) serve a greater good than does being spiteful.
Much like the UN security council, it helps to bring all players to the table rather than ignoring them, as you say, out of spite.
Would you say they should go as far as appending license that would prevent companies from using their software if they have a history of hostility towards them? Or only when it comes to receiving donations?
> appending license that would prevent companies from using their software
No, that would defeat the purpose of FLOSS entirely.
They should aggressively defend FLOSS licenses, which would include prosecuting some of their current members. If they allow members to take, for example, the Linux kernel, modify it, and not redistribute modifications to customers (as required by the GPL), then what exactly is the LF doing other than collecting membership dues from literally anyone that can afford it, and using funds to help members promote their 'we love open source' initiatives/marketing campaigns?
No, that would defeat the purpose of FLOSS entirely.
They should aggressively defend FLOSS licenses, which would include prosecuting some of their current members. If they allow members to take, for example, the Linux kernel, modify it, and not redistribute modifications to customers (as required by the GPL), then what exactly is the LF doing other than collecting membership dues from literally anyone that can afford it, and using funds to help members promote their 'we love open source' initiatives/marketing campaigns?
> and not redistribute modifications to customers (as required by the GPL)
Upon request the code should be made available, specifically if it's been modified and you are running the software. But yeah they should be at least open towards customers about it (those who have a right to said source, but also they can redistribute it legally under the GPL license terms).
Upon request the code should be made available, specifically if it's been modified and you are running the software. But yeah they should be at least open towards customers about it (those who have a right to said source, but also they can redistribute it legally under the GPL license terms).
The point is, at least two of the 'platinum members' of the Linux Foundation have publicly violated this. And yet, there they are.
GPL and other libre licenses have termination clauses. Eg,
> 8. Termination.
> You may not propagate or modify a covered work except as expressly provided under this License. Any attempt otherwise to propagate or modify it is void, and will automatically terminate your rights under this License (including any patent licenses granted under the third paragraph of section 11).
> However, if you cease all violation of this License, then your license from a particular copyright holder is reinstated (a) provisionally, unless and until the copyright holder explicitly and finally terminates your license, and (b) permanently, if the copyright holder fails to notify you of the violation by some reasonable means prior to 60 days after the cessation.
> Moreover, your license from a particular copyright holder is reinstated permanently if the copyright holder notifies you of the violation by some reasonable means, this is the first time you have received notice of violation of this License (for any work) from that copyright holder, and you cure the violation prior to 30 days after your receipt of the notice.
> Termination of your rights under this section does not terminate the licenses of parties who have received copies or rights from you under this License. If your rights have been terminated and not permanently reinstated, you do not qualify to receive new licenses for the same material under section 10.
> 8. Termination.
> You may not propagate or modify a covered work except as expressly provided under this License. Any attempt otherwise to propagate or modify it is void, and will automatically terminate your rights under this License (including any patent licenses granted under the third paragraph of section 11).
> However, if you cease all violation of this License, then your license from a particular copyright holder is reinstated (a) provisionally, unless and until the copyright holder explicitly and finally terminates your license, and (b) permanently, if the copyright holder fails to notify you of the violation by some reasonable means prior to 60 days after the cessation.
> Moreover, your license from a particular copyright holder is reinstated permanently if the copyright holder notifies you of the violation by some reasonable means, this is the first time you have received notice of violation of this License (for any work) from that copyright holder, and you cure the violation prior to 30 days after your receipt of the notice.
> Termination of your rights under this section does not terminate the licenses of parties who have received copies or rights from you under this License. If your rights have been terminated and not permanently reinstated, you do not qualify to receive new licenses for the same material under section 10.
Why should the Linux foundation be full of FLOSS sticklers? Not even Linus cares that much about freedom, as his refusal to licence Linux under the GPL3 shows
Linus explains his reasoning on that here: https://www.youtube.com/watch?v=PaKIZ7gJlRU
Relicensing Linux as GPLv3, even if actually possible (GPL is incompatible with everything that's not a subset of GPL, including other versions of GPL: GPLv2 is not compatible with GPLv3) would be unacceptable to many commercial players. It wouldn't be end of the world - most big customers, from RedHat to Android folks already have to maintain their own forks anyway - but it would hurt in the long run.
Taking their money means they have less money to be "hostile to FLOSS".
You're operating off the assumption that open source foundations aren't mainly a racket themselves.
So are you saying it's wrong for a company who had a history of being hostile to want to support open source now? Isn't a company that used to be hostile towards FLOSS switching to investing half a million a year into it... a good thing?
VMWare violating the GPL license? No offence, but it's just mixing up FSF's PR with reality. Which is: VMWare can afford much better lawyers than FSF could, and if there really was a violation, they wouldn't give the project a green light, given that alternatives (operating systems under less restrictive licenses) are freely available.
It wasn't really just EFF PR, but it, for what it's worth (IANAL and this is just my opinion), rested on some very subtle points about which reasonable people disagreed. Here's something I wrote at the time:
https://www.cnet.com/news/vmware-and-the-gpl-round-two/
https://www.cnet.com/news/vmware-and-the-gpl-round-two/
Being reasonable person is one thing, understanding licensing issues is something completely different.
The whole thing boils down to one thing: whether something becomes a "derivative work" just due to being linked with code licensed under GPL. FSF says it does - and one could assume they know what they are talking about; after all it's their own license, right?
Except that the GPL license doesn't really say that; it's just FSF's interpretation. While e.g. the Mozilla license states very clearly what is and what is not covered by the license, thus avoiding the problem, the GPL is unclear.
There is, however, one very simple way to win any lawsuit that would claim that your binary blob violates the GPL: just make sure it's possible for you to demonstrate in court that your code can run with something else instead of the GPL thingie. If someone sued, say, NVidia, for putting their binary drivers into people's GNU kernels, they can point at the FreeBSD version of their driver: if it works without Linux, it clearly doesn't constitute a derived code; Linux is just one of the platforms it works with.
The whole thing boils down to one thing: whether something becomes a "derivative work" just due to being linked with code licensed under GPL. FSF says it does - and one could assume they know what they are talking about; after all it's their own license, right?
Except that the GPL license doesn't really say that; it's just FSF's interpretation. While e.g. the Mozilla license states very clearly what is and what is not covered by the license, thus avoiding the problem, the GPL is unclear.
There is, however, one very simple way to win any lawsuit that would claim that your binary blob violates the GPL: just make sure it's possible for you to demonstrate in court that your code can run with something else instead of the GPL thingie. If someone sued, say, NVidia, for putting their binary drivers into people's GNU kernels, they can point at the FreeBSD version of their driver: if it works without Linux, it clearly doesn't constitute a derived code; Linux is just one of the platforms it works with.
No real disagreement.
Some combination of the license (and existence of the LGPL), other things the FSF has written, practice, and history suggest that somewhere between statically linking your program together with GPL programs and shipping your program on a disk alongside unrelated GPL programs, you go from having to comply with the GPL and not having to comply with the GPL (for your own code).
However, much of the space between those two points is a matter of conjecture, legal theory, and opinion.
Some combination of the license (and existence of the LGPL), other things the FSF has written, practice, and history suggest that somewhere between statically linking your program together with GPL programs and shipping your program on a disk alongside unrelated GPL programs, you go from having to comply with the GPL and not having to comply with the GPL (for your own code).
However, much of the space between those two points is a matter of conjecture, legal theory, and opinion.
By this logic no company that can afford good lawyers would violate any laws.
That clearly is not the case, just look at VW diesel emissions scandal for instance.
Are the funds not used to develop Linux?
You're judging the foundation by who decides to give them half a million $ per year?
If Oracle, or anyone else you don't like, donates to the EFF, or the Red Cross or whatever: does that also taint those organizations?
If Oracle, or anyone else you don't like, donates to the EFF, or the Red Cross or whatever: does that also taint those organizations?
Sure, take their money, but don't give them a seat on the board of the organisation.
It's funny and sad how Linux Foundation gets all the attention from these corporations, when GNU/FSF have been just as (or more) enabling for their success (ignoring Android for a moment).
Imagine what GNU could do if they had just a fraction of the donations LF receives. Perhaps we could finally get good, affordable hardware that respects our freedoms.
Misunderstand me right, Linux Foundation does great work and deserve the success, but they likely would not exist if it wasn't for the FSF.
Imagine what GNU could do if they had just a fraction of the donations LF receives. Perhaps we could finally get good, affordable hardware that respects our freedoms.
Misunderstand me right, Linux Foundation does great work and deserve the success, but they likely would not exist if it wasn't for the FSF.
Imagine what GNU could do if they had just a fraction of the donations LF receives.
Publish more rants telling people to stop using all software and services they enjoy? Fork already-free software to be more free? Rewrite popular BSD-licensed software as GPLv3?
Perhaps we could finally get good, affordable hardware that respects our freedoms.
The way to get freedom-respecting hardware is to build something 80% free, then invest the profits into a second generation that's 90% free, then 95%, etc. Companies like Purism are doing this but the FSF isn't willing to compromise so they make no progress.
Publish more rants telling people to stop using all software and services they enjoy? Fork already-free software to be more free? Rewrite popular BSD-licensed software as GPLv3?
Perhaps we could finally get good, affordable hardware that respects our freedoms.
The way to get freedom-respecting hardware is to build something 80% free, then invest the profits into a second generation that's 90% free, then 95%, etc. Companies like Purism are doing this but the FSF isn't willing to compromise so they make no progress.
>Rewrite popular BSD-licensed software as GPLv3?
What do they rewrite? To the best of my knowledge, they have no problem using BSD licensed software, they just won't develop BSD license software[1] (mostly, though they do approve it here and there for new projects).
[1]. More accurately, they won't approve it for new projects. But the projects are fine.
What do they rewrite? To the best of my knowledge, they have no problem using BSD licensed software, they just won't develop BSD license software[1] (mostly, though they do approve it here and there for new projects).
[1]. More accurately, they won't approve it for new projects. But the projects are fine.
It's neither funny, nor sad, it's a pretty direct reflection of the fact that GNU/FSF is pretty ineffective at achieving any of those goals over time themselves.
While I agree with you, GNU/FSF is only one player. Userspace isn't as important and is frankly replaceable in Enterprise.
I've worked a few gigs, and shipped some products, where we didn't use any GNU software at all.
Additionally, userspace isn't where a lot of the development needs to happen. We don't need new features or performance tuning of existing tools. We usually just write new ones.
I've worked a few gigs, and shipped some products, where we didn't use any GNU software at all.
Additionally, userspace isn't where a lot of the development needs to happen. We don't need new features or performance tuning of existing tools. We usually just write new ones.
> I've worked a few gigs, and shipped some products, where we didn't use any GNU software at all.
That's very impressive (assuming you were developing for Linux). I guess no one used Emacs, or GDB, and you were cross-compiling (using Clang and LLD) from Windows or macOS?
That's very impressive (assuming you were developing for Linux). I guess no one used Emacs, or GDB, and you were cross-compiling (using Clang and LLD) from Windows or macOS?
I am a FSF supporter generally and financially but let's be entirely honest about what the FSF has done in the last twenty years, and that's actually very little in terms of making good Free Software.
They lack a cohesive vision or the expertise to move projects forward. I'd love if that were different - we *need a Free Software org with vision, but the FSF hasn't had any for a long time.
They lack a cohesive vision or the expertise to move projects forward. I'd love if that were different - we *need a Free Software org with vision, but the FSF hasn't had any for a long time.
What happened there? In the 80s they were the next big thing. Then they finished their OS project (when the Linux kernel became available they finally had a FLOSS OS). But they never took off with their high priority projects. Is it that they can't find programmers? Is it that they pay so little that you might as well write your own code and host it on Git*.com?
Yea, we could, but that's not what their sponsors want.
You mean, GCC and the GNU toolchain, as it's the only actually crucial thing that can be attributed to FSF? It already seems to be quite well supported by hardware vendors, even if more "researchy" projects are slowly migrating towards LLVM.
When I read things like this, it makes appreciate OpenBSD (and the rest of projects under the OpenBSD Foundation) even more, considering how it is still actively and fearlessly maintained by few individuals (less than 40), motivated mainly by their own enthusiast and passion, investing their own time and money on it, just for the purpose of producing a bloat-free OS focused on security and correctness, that can rival GNU/Linux in terms of performance.
Um... How do the 40+ OpenBSD make living? I first thought donations to FreeBSD were already tiny compared to Linux, and then OpenBSD seems to be even smaller.
I am amazed at how the BSD community continue to survive when Linux has literally suck out all the OS OS development fundings.
I am amazed at how the BSD community continue to survive when Linux has literally suck out all the OS OS development fundings.
FreeBSD's kernel offers better performance (on the server) compared to Linux, and the licensing model makes it more attractive to some industries. Whatsapp (server side), Netflix, and the internals of PS3/4 all use FreeBSD code. OpenBSD is actively used in networks, as a firewall (where iptables from Linux is a mess) and other security oriented aspects where stability is also crucial. Linux is popular because it became available to users first than the others back in the 90s, OpenBSD's contributors, on the contrary, are comfortable on not implementing overengineered "features" with the sole purpose on keeping the source clean, stable, and with as less bugs as possible, as a UNIX system should be.
I may be cynical, but I'll make a bet that there's a big Fuchsia (https://en.wikipedia.org/wiki/Google_Fuchsia) announcement on the horizon and Google wants to have some evidence to point to that shows they still care about Linux to combat the inevitable "Google is abandoning Linux" FUD/headlines.
To add to the cynicism...
Maybe Linux Foundation will host that as one of their Collaborative Projects? :)
Seems like a ridiculously low amount of money from the world's largest company. Doesn't Linux power their entire business?
They also contributed a lot of work to the linux kernel like initial support for containers with cgroups in 2006-2007.
I've a rule: don't consider "give back" to the community things that are strictly aligned to what you need. Google contributes a lot to the Linux kernel by other means as well, but the work to the containers was strategic for them. To really provide something back is to give $$$ or work to just make a project better regardless of your end goals. Otherwise it's great that you are doing it as OSS, but I would not consider it a "give back" from what you got.
Isn't that the entire point of open source? I need a project, but it lacks the one feature I need. So I build the feature, contribute it back, and everyone wins. Is Google supposed to work on Amazon's needs and vice versa for them to be considered good patrons?
You are right downvoting me, I did not specify my reasoning. Basically contributing features you need in a very large project that you did not initiate yourself, leads to "featurism", more code to handle in the future, certain specific needs you accept mostly since they were contributed and saying no sometimes is hard. Instead work not interested to solve specific use cases tend to be consolidation work, or big general features work at least, where the goal is to fix something or do the right thing in some area, considering the whole public of such system, not your local needs. So the two approaches lead to different software philosophies and outcomes.
Do you have examples of companies that "work to just make a project better regardless of [their] end goals"? I sincerely can't think of any and don't see a problem with it either.
I skipped the part about giving $$$ because they already did that.
I skipped the part about giving $$$ because they already did that.
Meh. Doing the right thing for the wrong reason is still doing the right thing. If everyone benefits from their additions, it's immaterial to me what the intent of those additions were.
And even then is it "the wrong reasons"? Surely most third party contributions to the Linux core are from people that need a feature or support for a specific use case.
> Doesn't Linux power their entire business?
Doesn't they power you and your entire business and the web ? Search , Gmail , G Suite , Docs , Chrome , G Analytics , Firebase , Angular etc... For almost nothing per month ?
Money isn't everything , it's nice that they give more money but when you see their contributions level in Github ... They are literaly one of the biggest open source contributors on earth so I don't even see why they bother with that.
Doesn't they power you and your entire business and the web ? Search , Gmail , G Suite , Docs , Chrome , G Analytics , Firebase , Angular etc... For almost nothing per month ?
Money isn't everything , it's nice that they give more money but when you see their contributions level in Github ... They are literaly one of the biggest open source contributors on earth so I don't even see why they bother with that.
They aren't compelled to give anything -- though they might embarrassed that they attained Platinum level after Microsoft did.
It's pretty small but they do also contribute to the community in other ways through programs like GSOC. I think they could and should definitely do more though.
I have a mixed view about this. On one hand I'm happy to see money going back to open source, but then I'm not sure those foundations are a good thing at the end of the day.
They bring politics, ego, and other undesired side-effects to open source. I saw that happening for a lot of projects that were very well self-managed until they grew to become part of a foundation, with all the politics that come with it.
They bring politics, ego, and other undesired side-effects to open source. I saw that happening for a lot of projects that were very well self-managed until they grew to become part of a foundation, with all the politics that come with it.
IMO, politics are an unfortunate result of any kind of growth. Even self-managed organizations become political when they scale up -- it's just the nature of society.
I think the best organizations are the ones that actively manage their political complexity.
I think the best organizations are the ones that actively manage their political complexity.
500k lol that's peanuts. That's basically what they spend yearly on a good engineer.
keep in mind they also pay engineer salaries to work on stuff that gets contributed back to the kernel (and other projects). of course that's with a healthy dose of self interest, but this 500k isn't the real bottom line
Indeed Google has hundreds of engineers dedicated to kernel projects, not just people who are allowed to dabble in it.
While I'm glad that they're increasing their donation (and donating at all), I'm surprised it's only to $0.5 million per year for a company as big, profitable, and Linux-dependent as Google is. Though I suppose they also donate labor, which might not be quantified publicly.
How many Googlers contribute on corporate time or in their spare time? If you offer FT employment to a maintainer, that's just as meaningful.
For those questioning the donation/what it would be useful, or if you disagree with me that employment of oss contributors is an important form of support, please go back to recent history and look at what led up to Heartbleed in the OpenSSL project from an internal maintainer/organization perspective and what the Linux Foundation did in response to it.
For those questioning the donation/what it would be useful, or if you disagree with me that employment of oss contributors is an important form of support, please go back to recent history and look at what led up to Heartbleed in the OpenSSL project from an internal maintainer/organization perspective and what the Linux Foundation did in response to it.
They're also one of the biggest contributors of code, donating significantly more in engineering time and pay.
I've been thinking about it as well and came to an idea that excessive donation can be damning to the recipient. Tougher constraints keep you sharper.
That's true, but there's so much open source around and so many critical projects are developed on a shoestring or by time-pressed volunteers. We learned the consequences that hard way with OpenSSL (and at least partially mitigated them since).
An excessive donation could pay for some needed but more workmanlike development, like better open source drivers (or hardware reverse-engineering to create them).
An excessive donation could pay for some needed but more workmanlike development, like better open source drivers (or hardware reverse-engineering to create them).
Tencent became a platinum member a mere three days ago: https://techcrunch.com/2018/06/24/tencent-increases-its-focu...
I wonder if that may have prompted Google to up it's place on the list. Regardless, more money to open source is a good thing.
I wonder if that may have prompted Google to up it's place on the list. Regardless, more money to open source is a good thing.
I dont think you drop half a million dollars because a news article came out. My guess is this was in the works for a while, carefully assessing impact, risks and potential reactions.
Does the Linux Foundation really help Linux and software libre? If so, in what ways?
Everybody talking about money here, more important is the choice who will join the foundation. A precious position in your hand as employer.
Just a bit below two mid level developer salaries at Google.
Is that enough to make the best Linux desktop ever?
$500k per is a rounding error for Google. Why didn't they do this earlier?
Meanwhile, IBM has poured Billions with a capital B to turn Linux from a hobbyist me-too to a top-tier OS.
Wasn't that mostly stategic, as a counter to Microsoft?
I believe you can't really put a $$ amount on good OS projects. They really do fundamentally shift how things work for an entire industry and make a tiny dent in the universe.
While the world worships the Tech billionaires like Zuckerberg, Bezos, Page, Brin e.t.c Something has to be said for people who pour their hearts and lives to making a great OS project that's used under the hood almost everywhere.