RFC: PHP license update
wiki.php.net
wiki.php.net
Licensing: Meta and Google[1], and likely Microsoft, Apple, and most other megacorps explicitly forbid any use of AGPL software because it cannot be proven to be prevented from being invoked according to the vagueness of the "Remote Network Interaction" clause. So if you never want megacorps or anyone who runs a business to ever use your code, choose AGPL.
1. https://opensource.google/documentation/reference/using/agpl...
Also grafana has "observability as code" since release 12: https://grafana.com/docs/grafana/latest/whatsnew/whats-new-i...
And it seems there used to be something similar from earlier as well: https://grafana.com/blog/2022/12/06/a-complete-guide-to-mana...
> I do not understand why you say it's a worse licence?
Apache license is inferior to the AGPL that Grafana is using. If the project is any good it's just a matter of time before Amazon or some other big company starts selling a fork of the codebase and draining development resources, as it happened with Redis, ElasticSearch and other companies.
Other than that, younger generations are getting why the old models from shareware and trial demos, source available have made a comeback under apparently new models.
Not "anyone who runs a business" but "any business that uses your software to provide a proprietary network service".
That is the whole point of the AGPL!
Google's reasoning in your link is pretty clear that the problem is that they are a provider of network services. Most non tech businesses wold be completely unaffected by these issues and have no reason to care.
As far as many people are concerned, AGPL is vague enough it would end up in court, so the question becomes "do I want to have this sword of Damocles of unnecessary complications and upsets?"
Plenty of businesses run AGPL software (e.g. Grafana, Mastodon or Mattermost). Fewer run AGPL software for external paying customers.
As a developer, I care about the freedom of my users; I don’t much care about the paranoia of a megacorp who wants to restrict the freedom of its customers.
It's sold to us as non-news, which is good. No change for contributors, no change for end users, rights wise.
Funny you should mention stabilizer control (I don't think that is an aeronautic term). I recently visited the Battle of Britain Memorial Flight hanger at RAF Conningsby. It turns out that the Hurricane and Spitfire had unusual (by today's standards but normal for the times) ways of applying trim to control surfaces.
One of them - you glue a piece of string on top of an aileron and on the other you smack it with a hammer to bend it (that must be the Spitfire) and then you test it out and keep fettling until the job is done.
Well, that's roll sorted out, I'm not sure what trim for the other two axes (pitch, yaw) involves. Probably knicker elastic.
We didn't have time to polish ours like that what with the enemy shooting at them all the time. Oh and most of them were made of wood and wishes! The Hurricane was a bit of a crossover from WWI to WW2, yet rather a lot of Brits and Poles (int al) flew them against the Luftwaffe, who had rather more modern ME109s etc. It must have been a bit of a pain to see your enemy deploy a cannon when you only have a .303.
There is absolutely no doubt that if the US hadn't rocked up, we would have been ... well I'm not sure what Europe might mean nowadays if things had been different.
Possibly, our most lethal aircraft was the Mosquito. It was made of wood and twin enginned. So it was light and very, very fast and could pack a real punch - a quad cannon or lots of machine guns in the nose and carry bombs too. Parts were made in furniture factories and garden sheds - it was totally mad and rather hard to target manufacturing.
I think that the Nazi regime gave two kills for downing a Mosquito because it was considered a bit of a handful.
It's not multithreaded by default (as long as you're not running it in an environment like Apache), no async stuff going on, so the major roadblocks for real-time guarantees are already out of the picture.
The things I would be worried about is garbage collection and array handling. The former should be able to be configured (or in the worst case, rewritten) in a way that provides upper bounds on performance, people have already managed to do that for Java, the latter should be manageable by enforcing boundaries in the user code (e.g. disallow stuff like $foo[]='bar' that dynamically extends the length of an array).
--
Basically, these two clauses (first from PHP, second from Zend) are removed:
The name “PHP” must not be used to endorse or promote products derived from this software without prior written permission. For written permission, please contact group@php.net.
The names “Zend” and “Zend Engine” must not be used to endorse or promote products derived from this software without prior permission from Zend Technologies Ltd. For written permission, please contact license@zend.com.
And replaced with:
Neither the name of the copyright holder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
--
Then the following three terms (4-6) are removed from PHP:
4. Products derived from this software may not be called “PHP”, nor may “PHP” appear in their name, without prior written permission from group@php.net. You may indicate that your software works in conjunction with PHP by saying “Foo for PHP” instead of calling it “PHP Foo” or “phpfoo”
5. The PHP Group may publish revised and/or new versions of the license from time to time. Each version will be given a distinguishing version number. Once covered code has been published under a particular version of the license, you may always continue to use it under the terms of that version. You may also choose to use such covered code under the terms of any subsequent version of the license published by the PHP Group. No one other than the PHP Group has the right to modify the terms applicable to covered code created under this License.
6. Redistributions of any form whatsoever must retain the following acknowledgment: “This product includes PHP software, freely available from http://www.php.net/software/”.
--
And the following three terms (4-6) are removed from Zend:
4. Zend Technologies Ltd. may publish revised and/or new versions of the license from time to time. Each version will be given a distinguishing version number. Once covered code has been published under a particular version of the license, you may always continue to use it under the terms of that version. You may also choose to use such covered code under the terms of any subsequent version of the license published by Zend Technologies Ltd. No one other than Zend Technologies Ltd. has the right to modify the terms applicable to covered code created under this License.
5. Redistributions of any form whatsoever must retain the following acknowledgment: "This product includes the Zend Engine, freely available at http://www.zend.com"
6. All advertising materials mentioning features or use of this software must display the following acknowledgment: "The Zend Engine is freely available at http://www.zend.com"
Now "Windows Subsystem for Linux" makes even less sense than it already did.
Aside:
"Meanwhile, Richard Stallman, author of the GPL and founder of the FSF, had significant disagreements with the PHP project over their use of the GPL, so the PHP project discontinued the dual-licensing approach, removing the GPL license as an option"
Haha, such classic Stallman.
And reading the full license text, I don't see "or later" there either?
It amounts to the same thing, but the former framing means they're covered by the "or later".
The main link that cites is primarily about the MySQL license change to GPL and the impact on PHP licensing of tha. I also do not understand why anyone would drop the GPL because Stallman does not like their licensing.
Stallman is said to be notoriously difficult to work with, and IMHO he's quite an extremist.
He tried to veto emoji support in Emacs, because at the time it's only been supported on macOS. The reasonable recourse would've been add emoji support on the free platforms, but he seemingly can't see that far. So he's been ousted from the project.
In situations like this, I honestly would prefer to shut him off as well. I have huge respect for his past work, but at this point he's doing more harm than good.
Without Stallman, the moderate would be cast as the extremist.
Stallman’s positive contribution to humanity cannot be overstated by any measure. Without him and the like of him, it is not improbable that the whole software industry would be like Apple ecosystem.
That, and the uh, "it's not rape if..." comments. I wouldn't want anything to do with this guy following that, just no.
You do not have to work with him to use the GPL. You can use the GPL and ignore him completely. Its just a license.
Because he can be so obnoxious that people want to sever all ties with him and anything that has to do with him so that they never ever have to talk to him ever again in any capacity.
This goes back a long time. For example Keith Packard on why X11 isn't GPL (from https://youtu.be/cj02_UeUnGQ?t=1705):
> Richard Stallman, the author of the GPL and quite an interesting individual lived at 5405 DEC square, he lived up on the sixth floor I think? Had an office up there; he did not have an apartment. And we knew him extremely well. He was a challenging individual to get along with. He would regularly come down to our offices and ask us, or kind of rail at us, for not using the GPL.
> This did not make a positive impression on me; this was my first interaction with Richard directly and I remember thinking at the time, “this guy is a little, you know, I’m not interesting in talking to him because he’s so challenging to work with.”
> And so, we should have listened to him then but we did not because, we know him too well, I guess, and met him as well.
There are many anecdotes like this over the years from many people, often people who in broad lines actually agree with Stallman.
Same. Wrote a web application using Catalyst and a client for mobile phones using J2ME.
Writing projects in obscure languages is the best way to avoid dumb professors/examiners ask you stupid questions about your codebase lol.
I remember reading the Wikipedia page of Perl the night before the oral exhibition, just to be sure and guess what... The examiner asked me about the difference between "Perl" (capitalized) and "perl" uncapitalized. Which was one of the first lines in the Wikipedia page.
The guy literally went fishing for stupid and pedantic questions to ask, on Wikipedia, the night before.
Think of what a pointer is. Picture a struct; accessing a member is dereference + offset. It makes sense, you know there's a car, and the car has a steering wheel, and the steering wheel has a horn button. Simple. A struct with one field, that field has an offset of zero - it fits our "naked pointer" case 1:1, but is more readable. The compiler will do the rest.
As my dear friend often noted, "why simple?"
"'Conspicuous', with reference to a term, means so written, displayed, or presented that, based on the totality of the circumstances, a reasonable person against which it is to operate ought to have noticed it."
What are some of the ways you can tell if a term in a legal document is conspicuous? One of the ways mentioned in the UCC:
"a heading in capitals equal to or greater in size than the surrounding text"
So apparently the thinking is that if you make everything in the license all caps, it is all a conspicuous term, and thus a court will be convinced that a reasonable person ought to have noticed it.
The main issue with BSD is its proliferation of versions, incompatible with each other.
If you're comparing a specific version straight with MIT, this is not evident.
The fact that the OSI didn't approve of the PHP License until pressured shows the wayward nature of their "stewardship" of "open source". As does their wonky and rights-eroding definition of "open source AI".
> The proposed license does not reduce any user rights or add any new restrictions on the use of code previously licensed under the PHP License, version 3.01,
Yes, it does. Modified BSD Clause 3 (copied below).
> 3. Neither the name of the copyright holder nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
I know I'm being pedantic, but this is a narrowing of rights.
> Do We Require Permission From All Contributors? The short answer is, “No.”
I think that they can get away with this change since the original license doesn't preclude a narrowing of rights on derivatives.
It would be interesting if a contributor protested the additional burden and headache of having to deal with a torrent of snail mail asking for endorsement.
3. The name "PHP" must not be used to endorse or promote products
derived from this software without prior written permission. For
written permission, please contact group@php.net.
4. Products derived from this software may not be called "PHP", nor
may "PHP" appear in their name, without prior written permission
from group@php.net. You may indicate that your software works in
conjunction with PHP by saying "Foo for PHP" instead of calling
it "PHP Foo" or "phpfoo"
PHP License Clause 3 & 4 are about protecting PHP branding. Modified BSD Clause 3 is about using the software author's name or likeness as endorsement. For example, it limits putting antirez's face and name on our managed Redis product without obtaining his permission.
However, since the PHP and Zend licences both permit the user to use PHP under the terms of whatever licence version was applied to that PHP version or any later version, the point is essentially moot, since a user can choose to use the new version of the PHP/Zend licence once published, which will give them the same rights.
You can totally change the license of already released code, if the change is compatible with the precious license or if you have permission from all the contributors whose code is still present in significant amount. (However, you can't prevent people from using the released code under the former license)
That said, my point was: as the author, if you have previously released software S version V under license L1, nothing prevents you from releasing software S version V again under a new license L2 provided all the contributors of significant portions of software S version 1 agree or L1 happens to allow this additional license (because it's permissive, or because it has a "or later" clause or some other means).
Of course, re-release under a new license or not, software S version V can be used under license L1 "forever" and users can choose to ignore the L2 release completely. You cannot remove license L1, you can only offer an additional possibility (using software S version V under the new license, which you didn't allow before the re-release unless the relicensing was allowed by L1).
I've not seen this done, but I can imagine this being useful if someone needs specifically software S version V under the new license. Usually, people can just use newer versions under the new license though.
I admit my comment was terse (and a further edition probably removed important phrasing), I hope this one makes things clearer.
And that's the default. Trademark laws and laws that protect individuals already work like this. I'm not even sure this clause is strictly necessary in the BSD license.
I assume they've carefully evaluated this change with a lawyer.
Even if this restriction does align with US law, I will be flabbergasted if it aligns with the laws of every other country as well.
No, it's not. Explicitly stating which rights you don't grant is not more narrow than implicitly not granting them, it's just clearer. Copyrights and trademark rights are different.
Implicitly not granted? You mean, not mentioned at all? Imagine a world in which the modified BSD license exists in a vacuum. This license restricts how a product can be endorsed/promoted as per the clause. Granted, additional restrictions are removed in regard to "PHP" et al.
The shape is different, not just clearer.
> This license restricts how a product can be endorsed/promoted as per the clause.
...is technically not true for licenses, because they do not _restrict_ usage, rather they _permit_ usage. The usage is, by default, always restricted by automatic copyright and authorship laws, so any license (including the GPL) is not a restriction but a permission, since it has _granted_ permissions that were implicitly not there.
So if you change your license in a way where it appears more restrictions have been added, but those restrictions were already implicit because they were not even covered or taken into account in your original license, no new restrictions were added, your "grant" was just made easier to understand.
"Not mentioned at all" does not mean you can do it. Licenses do not restrict, they permit, from a default of "you may not use my stuff".
The license text could also say "you may not break into my house". That would not make it a "narrowing of rights", and that doesn't mean other licenses implicitly grant you the right to break into the software author's house if you use their software.