It's debatable. Even though Linux and BSD's make their sources available to the public, there's not that many people willing or even able to read through the code and understand it well enough to even know what it takes (try to open up a text editor, make a change, and recompile the whole kernel with that change).
But what's the problem with that? I'm doing it all the time.
Somehow, the bugs seem to follow me like mosquitos do. This was the one from yesterday: https://www.postgresql.org/message-id/ZlHCxjfpxXyqS_wz@disp.intra.daemon.contact
I'm not really eager to start fixing that one, I hope the guys there will look into it. But I had some patches in the postgres at times when the query-optimizer did not produce the expected results.

I think this all is actually a matter of fear, and overcoming fear. Back when I, perchance, got a job as a business consultant, the other guys termed me "guru" - because I was not afraid of the technology. See, if you grab a personal-computer, install a unix onto it and then understand and configure it properly, then you can handle the compute-center environment of maybe a big bank as well, because it's all the same technology, only scaled up. You don't need to be afraid of it, you just need to read about the special features that are used in a corporate environment, and understand them. And obviousely you need to be careful.
But these guys tend to think, uuh this is big, so it goes over our understanding. And then they grab the manuals and try to enter these commands literally, without trying to understand what they do. Which doesn't always work.

It's probably the same now with the source base: uuh, this is so big and multi-leveled, it must go over our understanding. Which is not true, it's still simply make; make install.

Back in time, it was mandatory to compile your kernel first, before you could run a suitable system. And people didn't complain they would have a problem with that. So what has happened?

And the hardware that FreeBSD runs on is definitely NOT free. Somebody, somewhere, paid to produce it and sell it
That's true, I had to steal my first hardware (and often had to decide if I want something to eat or some hardware).

Maybe (some) people have just become too rich. And so they expect that everything has to be paid for, ignorantly assuming that everybody can pay (but how should the money magically appear?) - and decently ignoring that by tradition, over a million years, we always had built our necessities by ourselves (and not doing it yourself was only an option for the gentry).

Sure, you can normally not build computer hardware by yourself - but then, Seymour Cray could do it, so it is not impossible either.
 
Sure, you can normally not build computer hardware by yourself - but then, Seymour Cray could do it, so it is not impossible either.
You can trust a physics PhD to run a nuclear reactor AND to be able to learn to run a vacuum cleaner. But you can't trust a random janitor to run a vacuum cleaner AND be able to learn how to safely run a nuclear reactor. There's kind of a reason one has a physics PhD and the other does not... ?

Even if the janitor has no fear of flipping a switch on a nuclear reactor, you probably don't want a janitor anywhere near the switch panel for the facility. This is why appropriate expertise is such a difference maker - this is why huge nuclear facilities have tight security that only allow trained and credentialed staff in to work. If you have no idea why you absolutely cannot drop the uranium cartridge - you have no business being anywhere near it.

But even a rank-and-file physics PhD is not a sure bet for being capable of running a nuclear reactor - lecturing at a university and understanding the theoretical math is very different from realizing that if you don't build a huge fence around the facility, you'll end up with practically a 'death zone' around it.

But back to the thread:
See, if you grab a personal-computer, install a unix onto it and then understand and configure it properly, then you can handle the compute-center environment of maybe a big bank as well, because it's all the same technology, only scaled up. You don't need to be afraid of it, you just need to read about the special features that are used in a corporate environment, and understand them. And obviousely you need to be careful.
If that were the case, then anyone who wants a job at Intel or Oracle or Microsoft - they would have hand one by now. Yeah, info is out there - if you're not too lazy to dig it up, huh? Yeah, find me a random Joe who can learn Matlab in a day or so, documentation is out there, it must me so easy to learn, right? Yeah, how would you know you have your expert?
by tradition, over a million years, we always had built our necessities by ourselves (and not doing it yourself was only an option for the gentry).
Yeah, try writing your own OS - the TempleOS guy did that, y'know.
 
You can trust a physics PhD to run a nuclear reactor AND to be able to learn to run a vacuum cleaner. But you can't trust a random janitor to run a vacuum cleaner AND be able to learn how to safely run a nuclear reactor. There's kind of a reason one has a physics PhD and the other does not... ?
Yes, and that's exactly the cool thing with computers.
You cannot just build an aircraft DIY and fly around - or, few people can do that, and with a lot of people, I would just *not want* them to do it.
But you can always install something on a personal computer, and try things out - and either you're able to learn or not. There is very limited danger along.

But back to the thread:

If that were the case, then anyone who wants a job at Intel or Oracle or Microsoft - they would have hand one by now. Yeah, info is out there - if you're not too lazy to dig it up, huh?
Mayby that's indeed the point - lazyness. But also, rules have changed nowadays. Getting a job seems no longer a matter of technical skill, but rather of buerocratic compliance.

Yeah, find me a random Joe who can learn Matlab in a day or so, documentation is out there, it must me so easy to learn, right? Yeah, how would you know you have your expert?
I have no idea what Matlab can do. I'm already too lazy to configure my office suite appropriately. But I love this OS. When I got my first CDrom with Rel. 2.1.6., I was thinking how many thousands of hours of effort were condensed in that little shiny thing, already back then in 1995. And that's a gift, and one appreciates such a gift by playing with it, looking into it, understanding how it works, and utilizing the fact that it is all open software, can all be changed and adapted.

That's what it is about. At least I think so.
 
There is very limited danger along.
Until you ask that machine to control a nuclear reactor and collect all kinds of logs. If you have the brains, you can adapt Open Source components to do the job - or closed-source stuff that just might be a specially built tool for that job.

Getting a job seems no longer a matter of technical skill, but rather of buerocratic compliance.
My impression is that it used to be 80% technical skill and 20% industry skill. Like a database admin in a lawyer's office, who can query a database for a given case, but have no idea if the query returned any useful results. Seems like (in a law firm setting) these days, it's better to have a lawyer who's at least a little bit technically literate than a skilled sysadmin who can fix a booting problem - but has no idea how a law office functions. Certain data needs to be protected from unauthorized access for a variety of reasons. And yes, you gotta have such reasons in mind when setting up the authentication scheme or design storage for that data. Preferences are changing towards 80% industry skill and 20% technical skill. If you're wondering where I get my numbers from - google the term Pareto Principle.
 
My impression is that it used to be 80% technical skill and 20% industry skill. [...] Preferences are changing towards 80% industry skill and 20% technical skill.

Thats similar to what I observed. The technical gurus got fired and preference moved to the business consultants, i.e. those who can throw around three-letter-acronyms without having a clue what they mean.
I preferred to term the outcome "skillfree delivery".
 
Thank you for sharing your idea/concept of "Skill Free Delivery".

How do you market an object like FreeBSD, that takes a fair amount of effort to read and study and try things out? I have used the Clang LLVM compiler and the work other Developers completed in porting FreeBSD software to Raspberry Pi 4B , 400 SBCs. to port https://github.com/ghostbsd source code (https://ghostbsd.org which is a copy of FreeBSD source code , with modifications) to run on a Raspberry Pi 4B hardware. Like you did with CD rom containing that gift of software operating system, I , too, enjoy FreeBSD or GhostBSD running on a Raspberry Pi computer connected to my HDMI Television. Yet, I must not be saying the 3 letter buzz words to attract attention to my free offering, Like Robonuggie YouTube channel can.
https://ghostbsd-arm64.blogspot.com/2024/01/january-19-2024-howto-download.html
View: https://www.youtube.com/watch?v=uJtLJGujjhM
Yellow USB flash drive, UGreen USB SSD Flash drive, Raspberry Pi 4B sbc.
View: https://www.youtube.com/watch?v=FqPuSd1DnZc
Items necessary to create a Raspberry Pi 4B or 400 Keyboard GhostBSD-Arm64 workstation


https://ghostbsd-arm64.blogspot.com/2024/05/netbsd-aarch64-100-install-on-raspberry.html NetBSD 10.0 Install to Raspberry Pi 4B Aarch64 hardware.
 
I am so gonna steal this...
You're very welcome :)

Thank you for sharing your idea/concept of "Skill Free Delivery".

How do you market an object like FreeBSD, that takes a fair amount of effort to read and study and try things out?
A good question, but sadly I'm the wrong guy to ask that. I've given up on that matter.
Nowadays everything has to be "marketed" in order to grab some attention in a world that is already overcrowded with attention-grabbers. And I don't really know how to cope with that.
 
The 'ideal' mix of technical literacy vs 'industry-related competence'... An admin is supposed to be able to respect and support the workflow of the office. If info is meant to be shared, where's the competence of Apache and IPv6? If info is meant to be protected, where's the ability to set up MFA correctly? Who can you trust with that kind of judgement, to be able to tell which is which without getting the rest of the office into trouble?

This is partly why licenses exist, why there's a bewildering mix of them, and they tell you different things. Yeah, you want an expert, you want some kind of limits to avoid getting into trouble.
 
At the risk of going on a tangent ...

Well thought-out and well-written software that just-works doesn't stir up much noise in public forums while quirky, hard-to-use, and buggy software gets lots of attention - creating the false impression that it is more popular.
 
How do you market an object like FreeBSD, that takes a fair amount of effort to read and study and try things out?
1) why do you want to market it? Marketing can mean many things, but often it is trying to create a demand where none exists.
2) If you want to tell a person about FreeBSD, you can try the competence / skill approach: "FreeBSD requires learning the skill of how to use it and become familiar with it. When you have done that, you have at your disposal a tool for very many computing tasks".
 
'industry-related competence'
Does that skill come with kneepads? Inquireing minds dont't want to know...

I have realized that most companies have an age limit of maybe 60 years, after which they can only implode under their own incompetence, nepotism, greed and bureaucracy. There is no good in fighting that. When you have companies that are around for longer they are either still in the hands of the founders, small enough not to have the critical mass for the collapse or they are kept alive from the outside. Well, more like "kept around" than "kept alive". Skilled people start jumping ship long before, or go start-up and make the old company obsolete. We already have a good licence for that, you can take any work under that with you when you leave. GPL3 tries that with the anti-patent part. Patents are doing severe harm these days.
 
When I was in college (just after y2k), selection of software to perform particular tasks was kind of limited, so people talked about 'software patents', rather than 'software licenses'. There were Patent Offices in many countries, and they were well suited to using the legal system to fight against copycats and make money for the inventors/programmers. Patents as such made sense back then.

By contrast, the very concept of 'software licenses' started making sense because of proliferation of software. 'Prior art' is a term you often see when doing research on how to file a patent. In case of software, there's just too much of that 'Prior art' out there to be able to make a case that a patent should be granted for your efforts. The earliest known software patent was granted in 1950s to General Motors (to cover an OS they wrote for IBM's 701 computer)... precisely because nothing like that ever existed before. Even Microsoft Windows could not get an OS patent, they ended up with a 'GUI patent' in 1985. Licenses started making sense on many accounts when patenting stopped making sense for software.

Skilled people start jumping ship long before, or go start-up and make the old company obsolete. We already have a good licence for that, you can take any work under that with you when you leave. GPL3 tries that with the anti-patent part.
When I was in college, one of my Computer Science professors remarked that according to university policy, rights to software (written by a professor) belong to the university. But by contrast, the same policy document also specified that if a student wrote some decent software, the student retains the rights to that software, not the university. That basically means that if a professor leaves the university, it's the university calling the shots on what to do with the software the professor wrote - patent it, publish/distribute the source, and more. By contrast, if a student wrote some software and then leaves the university - it's the student calling the shots on what happens to the software they wrote (patent/publish/distribute). (yeah, provided the student did not graduate with a crushing debt, as is common within US ? )

My point is, software licenses are easier to deal with than software patents. There's been quite a bit of criticism of the whole 'patent system', btw, coming to light at a time when the software licenses have been a huge point of contention around 2010 (plus-minus a few years). Who can forget the rampage that SCO went on? or Sony/MPAA?
 
To me what Perens' idea is all about are people like Werner Koch, the developer of GNU Privacy Guard, the nowaday standard PGP program.

Koch might be a good developer, but he sucks at money. So Koch always failed to ensure that he can make a living by e.g. getting hired by a big company, even for some time was short in the verge of bankruptcy. The is why he considered quitting the project at that point. His project also seems to be more or less a one man show.
What if he had offered $1-2 license for GPG? Probably good income and motivation for improvements. There will be disgruntled people arguing that software should be free. Even just $2 would be considered cheeky. In the same time same people will pay $10 for phone glass protector or colored case.
 
What if he had offered $1-2 license for GPG? Probably good income and motivation for improvements. There will be disgruntled people arguing that software should be free. Even just $2 would be considered cheeky. In the same time same people will pay $10 for phone glass protector or colored case.
That's the playbook that Apple uses for the iPhone app developers already. And there are glass protectors that cost $30-40 on Amazon easy - already. ?
 
What if he had offered $1-2 license for GPG? Probably good income and motivation for improvements.
It would then have failed within the industry since no-one would want to pay a license for something so trivial.

However, it would have meant that there was now a gap in the market for an open-source/non-license encumbered one to be developed. And hopefully by a developer who is a bit more financially stable.

If developers can't afford to work for free on something they enjoy, that is absolutely fine, they should focus on stability first. What they can't assume is that they will make money on working on something for free. That is a gamble. I believe there is a massive push for monetizing free software in the last 5 years (including new licenses such as Post Open and Business Source license).
 
Here's what I've come up with:

Code:
Copyright <YEAR> <COPYRIGHT HOLDER>

Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.

2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
This first part is a direct copy of the 2-clause BSD license, also known as the Simplified BSD license or FreeBSD license.

Code:
3. File-based directory wide - Modifications made to this source code directory belong under the terms of this license.

4. Preservation of dynamic linking - Code under this license cannot be restricted from use as dynamically linked libraries. Also, code under this license, must be permitted to be dynamically linked to libraries which allow it.

These are the customizations. File-based directory wide clause helps keep the code organized and separated from other code. It also preserves this code's terms through many iterations.
Code:
5. Patent Grant - Grant of Patent License. Subject to the terms and conditions of this License, each Contributor hereby grants to You a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable (except as stated in this section) patent license to make, have made, use, offer to sell, sell, import, and otherwise transfer the Work, where such license applies only to those patent claims licensable by such Contributor that are necessarily infringed by their Contribution(s) alone or by combination of their Contribution(s) with the Work to which such Contribution(s) was submitted. If You institute patent litigation against any entity (including a cross-claim or counterclaim, but excluding declaratory judgment actions) alleging that the Work or a Contribution incorporated within the Work constitutes direct or contributory patent infringement, then any patent licenses granted to You under this License for that Work shall terminate as of the date such litigation is filed.
This is almost an exact replication of the patent clause from Apache 2.0. The only modification here is: "but excluding declaratory judgment actions". I might stay with the original Apache 2.0 patent clause. I only want to allow that someone can ask what their rights are. and maybe judgement actions is excluded, as a judge can make a wrong call, and they should lose their rights to the code, if they try to infringe on the opensource.
Code:
6, If code under this license is sublicensed, it must be with a compatible license, where this licensed code retains its full terms. If sublicensed, these terms must be kept, and also be shown through NOTICE to keep the terms of this licensed code intact from that of other source directories used in ways compatible to this license. The sublicense will pertain to this directory with the combined license of those other directories.
This clause may not be needed, as obviously code under this license must fall under these terms, this is more of an explanation for clarity than a rule. A sublicense will pertain to multiple sourcecode directories, which includes this one which retains its terms. An author can add more warranties or restrictions with the combined directories, which includes those and this one together. This doesn't affect code which is connected through dynamic linking, and depending on the sublicense other linking. That code of the sublicense, falls under those terms, which are only a variant of these terms. The original sourcecode under this license which is the directory must be marked.

Code:
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS “AS IS” AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
This part is from the same BSD license.

Further Notes:
If you want to add your own code through your licensing terms, do so outside of the license directory.

Preservation of dynamic linking makes it compatible with more licenses, except the more viral ones. Viral licenses need to be updated to at least allow dynamic linking exceptions to make use of other opensource code. The onus is on them to remove those specific restrictions, through a standard dynamic linking exception to use other libraries. The largest example is that GPL2 needs to be updated to allow code to be dynamically linked to use LGPL3.

Here's what I came up with. There was a contradiction in sublicensing, but I think I got that worked out in the above. Code used in a compatible sublicense, can also be permitted to be linked and used with this license, without necessarily requiring dynamic linking. I may need to add this in to a rule. The purpose is that, code under this license and with sublicenses can have additional warranties or restrictions that only apply to the combination of those works that are under this license and allowed sublicense. That author can offer that.

Dynamic linking is guaranteed, no matter if it's not under this license or an accepted sublicense. Other linking can be made, but isn't always guaranteed depending on the combination of the sublicense with this license. I may need to rough out this proposed license, to maintain this idea and intent.

The numbered ones are conditions and preserved permissions. The last one is the part similar to do what you want, use as is and what's not covered. The last part which isn't numbered is like the MIT style part.

In my opinion, true opensource means that dynamic linking from libraries should always be permitted, and that opensource should be able to use opensource libraries within reason, through dynamic linking.

The directory based clauses are intended to preserve the opensource, through many iterations as long as any iteration is preserved, which lost code can be made up for to replicate the original function of lost code. It is also intended to keep a clear separation of what falls under this license.

Not fully thought out, what can be done, if an author wants to add additional warranties. Those warranties will be for their own snapshot, with the combination with their code. Static linking is allowed, so it can be used through that. sublicensing can remove the ability to static link for use with their code combination. Then, the code under this license, will always be as free as above, though, may not necessarily have the same access to the sublicense, depending on their terms, but the terms to use with this one permit the rules of this license, when separated.
 
I've been thinking about CDDL, and it's actually more honest and more functional for dual business opensource/share than dual licensed GPL to proprietary. CDDL 1.1 isn't OSI approved, and it was speculated that, it was because they didn't apply for approval. I wonder if it's because the patent clause, while is really basic as necessary, protects the shared product over that of other opensource. CDDL1.0 is OSI approved, yet the difference is that it lacks a patent clause. CDDL doesn't infringe on other opensource, however, it's incompatible with GPL, so that in itself might be perceived as an attempted infringement. Actually, the viralness of GPL is more infringing compared to other opensource. For instance, when a product is dual licensed, to proprietary and GPL, they get to receive improvements from competitors, but they get to keep their code from the GPL. This is where CDDL is better by being fair, and more honest. So what if company code is protected!? That's a good thing! That's the purpose of it, so they can share it, and neither side (company and free use) is infringed upon.

We should embrace CDDL, because that's one of the best ways for companies to share code freely, while at the same time, protecting their assets. When, thinking about GPL, all that's really missing is, for it to allow use of other opensource code through dynamic linking, but it doesn't. Arguably, this could be for use of only opensource code, or also dynamic linking of GPL code using any code which allows it, including proprietary code. They'll define what constitutes opensource, and whether CDDL is defined as approved to them. They likely won't allow it. However, GPL2.1 needs to be allowed for use with both major versions of LGPL. It also seems wrong that the only reason that LGPL is accepted with certain versions of GPL, is that LGPL can be absorbed directly to be GPL. Opensource, should allow the use of dynamic linking, but it should NOT force the absorption of libraries which can be dynamically linked, at minimum! Until GPL updates a minor revision for standardized acceptance of dynamic linking at minimum of other opensource, I can't embrace it like better opensource. I tolerate it, and have to pretend it's the way, it should be (while not pretending to the extent of going against its terms, which require complexity and additional explanations to make up for its shortfalls, and how it chooses to be and how its intended use). I disagree fundamentally with terms that it doesn't allow the use of other libraries through dynamic linking. I largely agree with if the part that restricts dynamic linking from it, as that's the author's choice on that part.

The other case about allowing dynamic linking to proprietary code which allows it, there's hocus pocus of how GPL is allowed to run on top of Microsoft Windows. There's no clear statement of separation, that GPL is used on top of a proprietary system. They'll say there is, but the fact is, even if Microsoft libraries aren't somehow used, somehow, the proprietary Microsoft system is used underneath. For the Linux kernel, the separation is in the license itself, as on top of a specific API, that noncompatible code is allowed beyond that. Another case is that, GPL could use something like ZFS libraries, without making hocus pocus statements, in a clear and way that simply avoids being challenged, if they allowed dynamic linking. They don't have to use ZFS itself, they could have made their own product, only using its libraries. Though they have their own alternative product now. Allowing use of dynamic linking, would also make it worthwhile to improve code, which currently has obstacles by the redtape that it has now.

In addition, GPL gets to virally pollute code intended under permissive license. One minor insertion, renders it all having to go by having GPL terms, unless it's cleared out, and people insist on keeping that GPL piece in that product. No one ever clears it out, such as to Xorg, with the rare exception of Xenocara which through using the minimalist of options. CDDL and Apache don't have the capability to do this. MPL2.0 only has the capability to do this, by its terms of making a description of when it's dual licensed with GPL, but it's not done through MPL at all by itself. GPL made that hurdle for which MPL tries to make a defined way to be compatible with it, which was already available through combo licensing. Unless an organization wholly keeps its terms to permissive or limited in not allowing viral license insertions, such as FreeBSD, NetBSD, OpenBSD, Postgresql, Apache Foundation, Mozilla Foundation and others do, massive pieces of code cannot be properly maintained. They'll continue to make excuses as to why Xorg includes unwarranted pieces of GPL. Yet, those orgs keep their code maintained, and keep it out, while when they can do that, it's ok for them to add GPL, outside of those large pieces of maintained code.


If I were to Tier licenses: the directory wide license, which always allows dynamic linking to and from with a patent clause, and those which are directly compatible, aside from dynamic linking, would be 1st tier. This would be due to the protection from being aborbed and restricted. Permissive licenses would be second tier, because they allow relicensing. Apache also protects itself better than other permissive licenses, by requiring preservation through notation of the original code. MPL and CDDL would be 3rd tier. LGPL would be 4th tier, because it's made to be abosrbed into GPL, without benefits of other licenses. LGPL allows other code to use it through dynamic linking, until it allows GPL to absorb it. LGPL will still be embraced, due to what it allows. GPL would be at the bottom or not approved by a new standard, simply for its viralness beyond where necessary, and prevention of default use of dynamic linking. It would just be tolerated, until if it's ever improved. It needs a direct replacement, if they don't improve that part.


Back to my proposed license of directory wide, and unconditionally permitting dynamic linking. Two clauses (directory wide file-based and dynamic linking unconditionally allowed) may seem redundant to each other, but it's better that way. If a compatible license wants to absorb this license, that combination can have stricter terms, except cannot infringe on the directory license and must still allow dynamic linking, when used in that combination. Those additional rules won't apply to code which is only used through dynamic linking. If someone doesn't like the additional rules, they can always pull the code out that's under the original license from that of the overlapping license, because those two licenses were used on compatible terms, where it applied to that combination. Also, by allowing compatible terms on top of the license, they can give a warranty to any version and combination of their choice, and it doesn't have to apply to everything under the above directory/dynamic linking allowed license, because it permits it. It's up to them what they want to do, such as warrant under the terms.

Let's say, an overlapping compatible license wants to provide a static library based on the code, then wants to restrict static linking from beyond that combination use. Then, that's accepted, because dynamic linking is always allowed from that license and closely used compatible license combination. Depending, also, static linking is allowed, until a license code combination restricts it, but dynamic linking can never be restricted directly from that closely used license combination.

If a piece of code is used through dynamic linking or even static linking, when that's not restricted by additional terms, then, that code doesn't fall under the applicable full terms.

This proposed license doesn't also need all of the terms such as that Apache 2.0 or MPL2.0 has, because it's intended to be permissive in principle and allow freedoms beyond the set down terms of clauses 3 and 4. In addition, it has other fundamental parts: the protections through the patent clause, and parts taken from other permissive licenses. If someone has a reason for additional terms incorporating code from under this license, or reasons to add permitted clarifications, they're free to make their own incorporating license.

For this proposed license, there would be different levels of use with other licensed code:
  1. Code which is under and compatible with terms of both the license and sub-license.
    • Let's say, an organization wants to only give additional benefits like warranties to certain parts or versions. That would be fine, because the rest of their contribution to that directory still falls under the terms, and not their terms for additional benefits, which they would keep a clear additional separation for those benefits not restricted by that license.
    • Also, code which is outside of the licensed directory, absorbed by the sublicense, but that combination of work still adheres to both licenses. Still, everything from the directory can be separated back out, under the terms of the original license.
  2. Code which is outside of the directory, and used side by side, such as with Apache, MPL or CDDL.
    • That code doesn't fall under the license terms, except for allowed use.
    • This perhaps might require the use of permitted static linking or additional another way of use. Static linking allowed, unless sublicense combo prohibits it.
  3. Code which is used through dynamic linking. If permitting, static linking, but see point 2 above for that.
    • More freedom of use with that code through dynamic linking.
 
Very interesting thread. I moved my laptop from Linux to FreeBSD just because of ZFS, I was using it under Arch but got fed up by all this GPL vs. CDDL stuff, so FreeBSD was a pretty straightforward choice for me.

Now my question. I read that a few companies use FreeBSD as a base for their product because of the permissive license, the usual names are Netflix, NetApp, Juniper, Sony, Apple at one time and so on. As per the BSD license they are not pushed to contribute back anything but I'm curious to now if and what they eventually gave back to the FreeBSD community.
 
As per the BSD license they are not pushed to contribute back anything but I'm curious to now if and what they eventually gave back to the FreeBSD community.
Over time I have noticed that regardless of license, companies really can always find ways to avoid giving back unless it benefits them (i.e as a grand marketing gesture or free maintenance).

Perhaps the earliest I remember is the PSX development kit (developed by Psygnosis). All very much GNU GCC and bintools and yet no source code was ever upstreamed or offered to the licensees who purchased their tooling.

Likewise even for larger hardware companies like Allwinner:
Allwinner does not actively participate in or support this community. In fact, it is violating the GPLv2 license in several ways and has so far not shown willingness to resolve this.

Ironically, the only entities that have the ability to actively persue and defend the GPL violations are the commercial companies themselves.
 
I have the opposite impression. Many companies seem to give back changes so that they don't have to continue to manage them themselves. Let the upstream project deal with future changes to the contribution required by changes in the base software.

GPL compliance is in the hands of non-profit organizations. It seems to work at least sometimes. All that dd-wrt and openwrt stuff is a result of that success.

Overall the license seems to make little difference bringing back random improvements. Building entirely new products on top of existing open source software might be more sensitive.

There are also organizations that don't allow any GPLed stuff at all. Different lawyers, different viewpoints.
 
I'm not talking about GPL violations because I consider those like thefts. I'm more curious about BSD licensed software and tools, for which companies have no obligations at all and thus all they give back is only due to their willingness to do it (regardless of their motivations).
 
I'm not talking about GPL violations because I consider those like thefts.
"Theft" vs "Taking". Both licenses ultimately result in the same.
but I'm curious to now if and what they eventually gave back to the FreeBSD community
Both licensing approaches result in a strong "maybe". Depends how they feel. That said, I think we are at a slight singularity now where new companies born today can't hope to play catchup anymore with open-source and will have to engage with free software if they want any hope to become competitive. Likewise old companies like IBM and Microsoft are rapidly losing ground to open-source simply due to the natural "enshitification" process and may well fizzle out.

In 20+ years I could imagine most of the world being based on open-source code. It will be an absolute discusting cesspit of *terrible* open-source code but nonetheless, it will be free. Whether they give back or not... it will require so much work to clean it up that it might as well be a closed-source binary. IMO, is that really giving back?
 
I'm not talking about GPL violations because I consider those like thefts. I'm more curious about BSD licensed software and tools, for which companies have no obligations at all and thus all they give back is only due to their willingness to do it (regardless of their motivations).
Well, why should they?

I suggest you put this into a bit of a bigger picture that spans more of time - that helps to adjust reality and understand what is considered "normal"...

See, in ancient Egypt, we had the pharao who was so unbelievably rich they could afford to build the pyramids, and we had the peasants who, more or less, had barely enough to eat. And that was perfectly normal, because pharao was god.
( Oh, btw, I have one about that, let's see if that works...

Fast-forward... in the middle ages we had the gentry and clergy owning practically everything, and the peasants who, more or less, had barely enough to eat. And that was perfectly normal, because the former did own the land by the grace of god.

We resolved this by inventing the guillotine. But then industrialization came up, and consequentially everything was owned by the rich factory owners, and the peasants had to go and work in the factory and, more or less, had barely enough to eat. That was considered normal, at least by the factory owners, because of the grace of money.

That one was then more or less resolved by Karl Marx and the labour unions. But why should things stay that way? It seems to me we now reach a new ruling, where the peasants have to write the software which the rich guys need to get richer, and that will again be considered perfectly normal.

In 20+ years I could imagine most of the world being based on open-source code. It will be an absolute discusting cesspit of *terrible* open-source code but nonetheless, it will be free. Whether they give back or not... it will require so much work to clean it up that it might as well be a closed-source binary. IMO, is that really giving back?
I might agree. But that will then not be an issue of open-source or not. This will rather be the issue that we are sitting on a tremendous mess of an unintellegible software stack where nobody any longer knows how things might interplay on one hand, and on the other hand have all our infrastructure, organization and public supply made dependent of exactly this very unintellegible software-stack. And I think that will constitute a situation that brings a lot more severe problems than just the question of open-source.
 
Back
Top