French Appeal Court affirms decision that copyright claims on GPL are invalid | Hacker News Reader

French Appeal Court affirms decision that copyright claims on GPL are invalid

thehftguy.com

297 pointsuser5994461198 comments

Full title too long for HN "French Appeal Court affirms decision that copyright claims on GPL are invalid; must be enforced via contractual dispute"

There was a brief HN discussion on the initial ruling 2 years ago https://news.ycombinator.com/item?id=24478769 the case just went to appeal earlier this year.

Note that the last HN discussion never really happened, it gathered both too many votes and flags in 20 minutes and got automatically hidden, only to be unblocked by a mod later. It would be great if you could refrain from flagging this time, even if the ruling comes as a culture shock to you.

I assure you it is not clickbait. The authors of GPL software are really spending years and a fortune in court only to have all their copyright claims dismissed. The ruling passed the High Court (TGI), the Appeal Court and the CJUE.

Did they ask the FSF to chip in?

What for? What's the relevance of the FSF here?

Well, if you really want an anecdote with the FSF. I happen to have one :D

The court used the French GPL translation from the FSF. The very translations saying that they are not official translations and cannot be used in court. Well, the FSF doesn't get to decide what's admissible in court.

https://www.gnu.org/licenses/translations.en.html

That's not that unusual. If you have a contract written in something other than English and it ended up in a US court, the court would probably look at this translation and essentially be like, well, even though it says that it was translated by a non-party to this case and there's no reason to believe it's intentionally mistranslated, so good enough. Especially if the parties to the dispute didn't have any objections.

What I think is the real meat of the story:

> That is to conclude, by having a contract (a license is a contract), you lose your rights to pursue any IP/copyright claims (the two are exclusive in French law).

GPL was written by lawyers for the American system of law and France is not part of the USA. The systems of law are different and therefore the implications of GPL-licensing code are different.

I'd be very surprised if France is the only country that treats software licenses like this. The basis of the USA system of law is completely different from even other European courts, let alone non-Western systems of law. Assuming you can write a license for use in the USA and enforce it the world around is quite foolish.

Should've gotten some Canadian lawyers involved in GPL. Due to the combined heritage of English and French (Quebec) Canada has both Civil law (French) and Common Law (English).

I believe Louisiana uses a combination of civil and common law as well for the same reason.

As I understand it, Louisiana is a civil law state as well: https://www.bloomlegal.com/blog/what-is-unique-about-louisia...

While this is a true factoid, civil law in Canada is restricted to Quebec, and is ultimately subordinate to the constitution, which is interpreted in a nearly identical manner to American law, in the common law style. The civil law influence is mostly restricted to Quebec in areas of provincial jurisdiction (family law, small claims, etc.) The Copyright Act is federal and would get the common law handling.

(Approximately the same situation is true in the USA with Louisiana, whose state laws also have a form inherited or at least strongly influenced by the Napoleonic code, ultimately, but since harmonized to fit into a common law system.)

> and is ultimately subordinate to the constitution,

Didn't like, a third or more of the population just didn't sign it?

It was a single province, which at the time had around a third of the population indeed.

Their signature was not required for the constitution to enter in effect. It was desired (it still is) but since the refusal was (and still is) mostly political, it is not regarded as a major impediment.

Keep in mind two-thirds is a qualified majority.

> Their signature was not required for the constitution to enter in effect.

That's... weird. I mean, I couldn't really imagine one of the 13 colonies not signing the declaration or constitution.

> but since the refusal was (and still is) mostly political

I mean, it's a political document... anything having to do with it will be political!

> That's... weird. I mean, I couldn't really imagine one of the 13 colonies not signing the declaration or constitution.

The US Constitution went into effect on ratification of 9 of 13 states. Virginia and New York were numbers 10 and 11, but ratified it before a government under the new constitution was elected. North Carolina waited about a year later to ratify it, and Rhode Island even several months after that (it ratified it only after the United States threatened to embargo them), and consequently were not part of the first government under the new constitution.

At a time where interstate travel could take several months, That doesn't surprise me.

But they did end-up all signing in a relatively short timeframe.

> I couldn't really imagine one of the 13 colonies not signing the declaration or constitution.

Canada never declared independence from the UK. That's probably the difference in thinking. The constitution was not revolutionary. It was a reorganization, and that reorganization was carried out according to UK law (which was equivalent to Canadian law as a UK colony) at the time. The law did not require the consent of the provinces to amend the constitution at that time. In fact, the law allowed the UK parliament to simply rewrite Canada's constitution unilaterally. So they did, upon the request of the Canadian federal government after years of consultations in Canada to propose the amendments the UK rubber-stamped, when they renounced all further claim to that in 1982.

If you want an analogy, some US states rejected the prohibition amendment, but they were bound by it anyway. Established constitutional law allowed it to be imposed on those states.

> If you want an analogy, some US states rejected the prohibition amendment, but they were bound by it anyway. Established constitutional law allowed it to be imposed on those states.

Except it was a constitutional amendment. And the process to create such an amendment is described... in the Constitution itself, which, naturally, every state signed.

Yes. Exactly the same was true in 1982 in Canada. The UK Parliament commissioned a board in 1867 of elected representatives from the regions of what now compromise eastern and central Canada. They drafted a constitution. Quebec was involved and signed off on the Constitution Act 1867. That constitution was passed into law by the UK Parliament.

Per the established constitutional law operating in Canada in the 1980s, following the convention of the previous centuries, the government petitioned the UK government to amend the constitution yet again, this time moving the process of amendment to the Canadian legislatures and fully severing the connection.

Constitutional rule in Canada begins when the British first appear and have jurisdiction, as far back as the 1600s in the eastern parts of the country. It was never interrupted. The Crown has always reigned. 18th century treaties and proclamations in UK law are an inherited part of our unwritten constitutional body. It was simply a different historical course from the USA, but not fundamentally less democratic. IMO.

(Although I risk sounding like a mad Canadian nationalist at this point, I would note that under the modern constitutional system, Quebec could just exit the confederation. That is not a right of American states.)

The word "political" has changed meaning recently to exclusively refer to discussion of minority rights and issues. If you asked (Americans at least) with no context, whether people wanted politics out of government, probably 50% would say yes.

That’s wild. “What the government does” is the definition of politics.

A number of US states did not ratify some constitutional amendments and they still entered in effect. Minimum requirement is two-thirds of the states not unanimity.

I meant "political" in the sense of "it's a gesture designed to cause some political effect or to be used in the future for political negotiations" - maybe not the best word but couldn't think of a better one when writing that.

Quebec never signed off on the 1982 constitutional reform, which created a mechanism for amending the constitution and finally fully separated Canada from the UK government. Until that point, the constitution of Canada was technically part of UK law. Not entirely unlike the devolved Parliament of Scotland today. The UK Parliament could amend it, in theory unilaterally. (It was long-standing convention not to do so except when requested, by the 20th century.)

So Quebec's signature was not required, legally, for the current constitution to be imposed upon them. (Long ago, Quebec's elected delegates did approve of the original Constitution Act in 1867.) Per the amendment formula, each province including Quebec has veto power over any further amendment that requires unanimous consent (affecting distribution of government powers, etc.) Which is now something of a stand-off. Getting an amendment that triggers that clause passed, seems all but impossible in contemporary Canada politically.

> So Quebec's signature was not required, legally, for the current constitution to be imposed upon them.

That sounds completely alien.

If you’re familiar with the US Constitutional Amendment process, not all states have to sign onto the amendment for it to go into affect. And when it does, even the non-signers are bound by it.

Correct... Except the process itself is described in the constitution, which every state signed.

That is true. I forgot about that part.

US Constitution had to be ratified by a minimum of 9 states (out of 13 total) before it went into effect. In practice, the Continental Congress made it operable with 11 signatures; the remaining 2 happened after that.

> Canada has both Civil law (French) and Common Law (English).

So does the US; Louisiana is America’s Quebec.

Not just legally, culinary too...

Why do you assume that they don't know what they are doing?

https://fsffrance.org/news/article2009-09-22.en.html

https://www.pillsburylaw.com/images/content/1/6/v2/1655/A9A2... (Doesn't address France specifically.)

That doesn't seem right. Does that mean when I buy any proprietary software and then reverse engineer and leak the source code, they can't sue me?

Rather it seems like they might have to sue you for breach of contract, instead of breach of copyright.

Per the article the court didn't rule that licenses were unenforceable, they ruled that they were enforceable via contract law and that the plaintiffs were incorrectly attempting to enforce them via copyright law.

But can you enforce click-through licenses in France?

If Entr'Ouvert had chosen the alternative route and sued as a contract violation, then Orange would have claimed that they never accepted the contract and the matter could only be decided by the copyright court. But (according to the article) you only get one shot at the litigation. So I guess you can never enforce the GPL in France. Maybe?

I don't think taking and distributing a library is comparable to a click-through license. There's a relatively simple case to be made that you could have only copied the code if you had accepted the contract.

Not that Orange wouldn't have tried to claim what you are saying, but as a layman, the case for contract law sounds stronger than copyright infringement.

The point of the GPL is that, if you don't accept the license, then you don't have any rights to distribute the code and doing so would be a copyright violation. It's not a EULA that has extra restrictions on what you can do.

I'm totally unfamiliar with French Civil law; can contract law ever be punishable with any prison time? In the US copyright law is (at least in theory) a criminal offense, punishable with prison time, but my understanding is that breech of contract is not.

The implications, if true, would be that if you license something in France, and start selling copies in flagrant violation of the license, you cannot be threatened with criminal charges?

> I'm totally unfamiliar with French Civil law; can contract law ever be punishable with any prison time?

Private parties in most countries can’t prosecute criminal charges anyhow, and limitations on the mechanisms of pursuing civil [0] law claims by private parties don't generally have any effect on criminal prosecution by the State, since even with the same common name for an area of law, the precise definitions in criminal law are different than civil law.

[0] used here in contradistinction to “criminal” as an area of law, not to “Common” as a system of law, since both distinctions have been relevant in the thread.

But as the article also points out, the different ways of enforcement typically result in drastically different levels of compensation (read: incentive) with contract law being judged on missed profits, which is often zero for FOSS libraries.

During the piratebay case in Sweden, products that has never been for sale and would never be for sale were discussed in terms of missed profits. The argument that the Swedish courts accepted was a hypothetical sale of said product using the closest similar product, with an added multiplier for the fact that the product would never be allowed to be sold in the first place.

It is hard to know if the french courts would accept a similar argument, but as the author I would definitive argue it. Depending on what the FOSS library do and what a similar product would cost if acquired through legal purchase, the number could easy reach into millions of "lost profits".

Reverse engineering is allowed under French law, there are specific legal exemptions around that.

It's common to see EULA/license/contracts with clauses that prohibit reverse engineering, these clauses are void in France and many other jurisdictions.

Article L122-6-1 https://www.legifrance.gouv.fr/codes/article_lc/LEGIARTI0000...

IANAL but EU reverse engineering exception is granted for purposes of interoperability as I remember? Think API, protocol, file format.

You can't reverse engineer and share random internals.

In the EU, and probably elsewhere, there are entire engineering shops of which the one and only purpose is to reverse engineer competitors’ features and figure out a legal way to get around patents and replicate those features.

No interoperability there.

Some people I know do that for a living in the auto industry.

There is some variation within the EU, with different national laws. In French law at least you’re right: reverse engineering is allowed only for interoperability and you’re not allowed to distribute decompiled code.

> That doesn't seem right. Does that mean when I buy any proprietary software and then reverse engineer and leak the source code, they can't sue me?

Yes it's legal in France if it's for "interoperability purposes" (so depends on your specific case). Basically any DRM is breakable legally due to this exception.

Reverse engineering is always legal as well and no contract signed can remove that right.

My understanding is that no contract can prevent you from doing what you want with stuff you buy, including disassemble, study, repair, modify, or any use in a non-intended way. IP still apply so it doesn't give you the right to share any information you may find, except for "interoperability purposes".

No, it means that what happens to you is completely jurisdiction-dependent.

> GPL was written by lawyers for the American system of law and France is not part of the USA.

The GPL also has been ruled a valid contract in the US, and the US also has the rule that with a contract in place, conduct authorized conditionally by a contract but alleged to be done outside of its authorized conditions must generally be pursued as breach of contract.

Would love a citation where this was actually litigated.

Edit: And to be clear here, I'm not doubting you, I am actually quite interested in software license/copyright litigation. Having had long discussions over which is better "BSD" or "GPL" and discussing the litigation of the BSD vs AT&T established some solid case law on both the legitimacy of, and the way the BSD license was viewed. But for a long time there hasn't been good litigation of the various GPL flavors that actually went to court and came back with a decision that was appealed and affirmed. Would love to add such citations to my collection.

This is not the case. In the US the Copyright Act supersedes all other common law and State statutes, this is written into the Act itself. See: Title 17 USC Section 301(a). It is not possible to sue for breach of contract in relation to a copyright claim.

Nope. 301(a) is talking about things like common law copyright. The notes to that section from the Cornell LII [0]:

> Nothing in the bill derogates from the rights of parties to contract with each other and to sue for breaches of contract; however, to the extent that the unfair competition concept known as “interference with contract relations” is merely the equivalent of copyright protection, it would be preempted.

If you think about it, if what you said were true, then you couldn’t make any contracts regarding copyright works and expect them to ever be enforced. Disney wouldn’t work as a concept. Sounds a bit extreme, is a bit extreme.

[0]: https://www.law.cornell.edu/uscode/text/17/301

> he basis of the USA system of law is completely different from even other European courts, let alone non-Western systems of law.

This suggests that it's meaningful to divide legal systems into “Western” vs. “Non-western”.

The most meaningful divide is rather “civil law” vs. “common law”. Japanese civil law is probably closer to any continental European civil law than any of them are to say the U.K.'s or U.S.A.'s common law.

> the U.K.'s [...] common law

You mean England-and-Wales' and Northern Ireland's common law ofc! :)

This is interesting new information to me, but I looked it up and Scotland indeed has what many consider a mixed system, quite unique since the U.K. is a unitary state.

I was already aware that Scotland uniquely for the U.K. has the unus testis, nullus testis principle which is ubiquitous in civil law systems, but I had no idea of how far reaching it was.

https://en.wikipedia.org/wiki/Common_law#Scotland

Regardless, looking at this map of the world here:

https://commons.wikimedia.org/wiki/File:Map_of_the_Legal_sys...

It is quite clear that continental Europe shares more with even China than it does with say, England, Australia, or the U.S.A..

civil law is obviously based on the "code civil" from the french republic and to a lesser extent ancient roman civil law.

The fact that most of the world has this system is quite easily explained. mainly thanks to colonialism and imperialism.

China adopted a similair system after the boxer rebellion.

I think many civilizations independently invented it.

The innovation that law is written down by lawmakers is not that hard to duplicate. If anything, the opposite happened and high frequency of common law in the world is purely because the British Empire was the largest empire to have ever existed and all states that utilize common law descend from the British Empire.

Until the British Empire, common law existed only on the British Isles and the rest of the world had civil law.

But then I assume that I could sue a French alleged-copyright violator in the USA, and if I win, the infringer is banned from selling that product in the USA?

Yes, that would probably work. USA jurisdiction still applies to software sold in the USA, even if it's originally written in France. Things might get more complicated when a French company hires remote workers, though.

The US and France are signatories to a good stack of copyright related treaties:

http://www.copyright.gov/circs/circ38a.pdf

You'd have to dig to find the details of the agreements and what they mean for GPL'd software, but it's not as simple as "France is not part of the USA". I'd be surprised if the authors of the GPL didn't take the differences into account. They aren't fools.

Not sure if naïveté or innocence, but I think you are overestimating the GPL efforts by a lot.

RMS may not be a fool but he is still American (and not a lawyer). I don't think he was focused in solving licensing/copyright issues in the rest of the world in the early 90s - and even late efforts for GPLv3 didn't change it that much.

Eben Moglen (https://softwarefreedom.org/about/team/) is a lawyer.

The GPL has been in existence for more than 30 years and has been enforced in multiple countries that are not the US, including France.

You are talking about GPLv3, of course.

In any case, this was posted along the discussion here, and you might find it interesting: https://www.gnu.org/licenses/translations.en.html

In a global marketplace though, if you can't sell your GPL-derivative in one large market, the fact that you can sell it in a smaller market may not be much comfort.

> GPL was written by lawyers for the American system of law and France is not part of the USA.

The important thing is GPL (and other licenses) are contracts. So violations of there terms is a breach of contract and in turn clearly handled by the court responsible for handling breaches of contract.

The interesting part is why it's not (majorly) handled as a breach of contract in the US?

The trick is following or similar sentence (qouted from GPLv3):

> will automatically terminate your rights under this License

Which means of the breach the license contract you lose the license contract and in turn can be sued for copyright violation.

Now the gotcha is not everything you write in a license agreement is necessary valid in exact the same you write it, so now what happens if automatic contract termination is not valid and contracts in such case e.g. exists in-between valid and invalid until conflicts are resolved or similar?

At this point you would still have a contract, one which grants you the right to use given software "just not in given way" and in turn wouldn't it be a case for courts responsible for contract breaches?

Now I know nothing about French Copyright law, but the idea that laws (against auto termination) which would exists to uphold fair competition and prevent abuse might majorly negative affect copyleft licensed open source makes you think.

>Now the gotcha is not everything you write in a license agreement is necessary valid in exact the same you write it, so now what happens if automatic contract termination is not valid and contracts in such case e.g. exists in-between valid and invalid until conflicts are resolved or similar?

Schrodinger's Cat.

>Assuming you can write a license for use in the USA and enforce it the world around is quite foolish.

In this case there is no reason to suspect it (the GPL or any software license) is not enforceable in France, you just have to go about it the right way and sue in the contract law courts.

As a specific wrinkle of the French system, if you sue under one system (copyright or contract), if you fail you can't sue under the other. In most countries you would not face such a restriction.

It's amazing to me that this has not come up before. Has a software license never been litigated as a copyright violation in France before? Strange.

In a lot of western countries that are not the US, software isn't copyrightable as-is, only as a description of its functionality or something like that (not a lawyer, might have been the patent vs. copyright thing).

This difference means that a lot of what a litigation-centric society would assume about software (perhaps not copyright, but patents) doesn't apply elsewhere.

That said, the goals of the legal systems tend to converge: it was about protecting ideas so the people investing in it stand a chance to see a return. I wouldn't be surprised if there simply is a different system with a different name but with the same goal (yet being incompatible with a global blanket software copyright claim). I have only studied the subject for a small group of western european countries (France not included) and most of them not being based on common law makes quite the difference.

Could someone explain why this was downvoted? I'm not familiar enough with any of this to understand.

Because it’s a horrible mess of confusion between copyrights and patents, and even says so, so it’s hard to see what the author was thinking even posting it.

I can't think of any Western country in which software is not copyrightable as-is. There are some where software is not patentable, but that is a very different thing that is not relevant to this story.

Most countries are a signatury to the Berne Convention, so in general, copyright is pretty similar in those countries. Patents are a different story and different countries have different pantent laws.

This is a travesty. The French law makes no sense.

The victim of a tort is barred from justice simply for not being expert legal scholars or for hiring a bad lawyer.

What's even crazier, The GPL isn't even in French https://www.gnu.org/licenses/translations.en.html , and French courts don't accept English evidence, so the idea that there was a contract is absurd. (and how can France even allows "contracts" where one party doesn't even notify the other that they agree to it?!)

Also, Orange violated the contractual rights of every user of their software, not just the rights of the author. So why can't other users sue Orange? Or maybe they can? They should!

Interestingly, in another decision, that did go as high as the Cour de Cassation, Microsoft did get a license violation treated as counterfeiting, with higher monetary damages.

  - https://www.legalis.net/jurisprudences/cour-de-cassation-ch-criminelle-arret-du-19-avril-2017/
  - https://www.legalis.net/actualite/contrefacon-de-windows-confirmation-du-mode-de-calcul-de-la-reparation/

Found via https://linuxfr.org/users/spacefox/journaux/une-violation-de...

More here on the Entr'ouvert vs Orange case:

  - https://www.legalis.net/jurisprudences/cour-dappel-de-paris-pole-5-ch-2-arret-du-19-mars-2021/
  - https://www.legalis.net/actualite/pas-daction-en-contrefacon-pour-le-non-respect-dune-licence-de-logiciel/
  - https://www.legalis.net/jurisprudences/tgi-de-paris-3eme-ch-3eme-section-jugement-du-21-juin-2019/
  - https://www.legalis.net/actualite/non-respect-dune-licence-libre-irrecevabilite-de-laction-en-contrefacon/

FYI when you format links that way, they aren’t clickable. I think it’s because you’re indenting them which makes them parse as code.

It’s really a lot of work, but if you select one and right–click on it, then you can choose to open it in a new tab/window/etc.

Um, not reading in a web browser. No mouse.

Interesting case https://www.legalis.net/jurisprudences/cour-de-cassation-ch-...

Looks like it's about someone selling Windows, without paying Microsoft and/or without having a reseller partnership with Microsoft. The article is sadly very short and lacking any details.

There is no contract involved depending on the circumstances.

Found more info:

If I understand alright, it's about somebody selling 13 656 illegitimate copies of Windows.

Microsoft went all the way to cour de cassation because they wanted to get full price for each copy and the court wanted to use bulk price to estimate damages. Microsoft won the case the 3 times, they were just not satisfied with the damages.

The ruling of the cour de cassation is really about denying microsoft the full price and explaining how to calculate damages.

I think the difference is that in the Windows case the "person in question never had a license" (or well enough valid licenses).

On the other hand in this case there was a license given out (to every one) but it's terms where breached. Licenses are contracts so breaches ending up on contract curt isn't absurd. The only reason it doesn't happen in the US is the automatic Termination of the license/contract if it's breached. But I guess this is where the court decided that a automatic full termination in the specific way combined with a lawsuit based on the after effects of the termination isn't valid/doable/...

If the person copied it from a legitimate version of windows that they originally bought, couldn't one then argue that they did have a license that they then violated?

I guess the point I am making is that I don't think that line tracks.

> couldn't one then argue that they did have a license that they then violated?

It's a thin line, but I would say no as they only had a license for a single thing, while GPL is a license for any amount of usage. What was violated was the context of usage, not that illigal copies where made (I mean copies => copyright ;=) ).

But I had similar thoughts, it's a really thin line.

> Orange Business Service used the library when developing a nationwide authentication service for France since 2005 (mon service public).

I was at France Telecom R&D in a team working in authentication, and remember clearly admiring the work of entre'ouvert. At the time, I tested Lasso [0] on my own computer but never used it in real projects. Colleagues also found it very interesting.

I tested many similar open source framework, such as SimpleSamlPhP which we used (and give credit) in a European research project.

There was also a US company making an authentication framework in Java.

> My guess is a government contract like can go for 7 to low 8 figures.

That's as far as possible to the reality. The French state is constantly broke, for example it argued of IT problems to delay military staff pay by several month and delay providers payment by a year (Louvois), while making a law that forbids the private sector to delay payments more than two months!

It asked to Orange to work for free on many topics and it probably did the same at other big companies.

Never do business with the French state!

[0] https://lasso.entrouvert.org/

> The French state is constantly broke

Is this an EU “we can’t set our own monetary policy, we have to ask the ECB nicely to do it for us” thing? Because the idea of a state [of a country whose populace are not, themselves, in an economic crisis] being unable to just make up some money to give people (in the form of e.g. T-bonds), sounds very strange to my ears—almost harkening back to a gold-standard era, where states that couldn’t find precious metals to dig out of the ground could run out of raw materials to mint new coinage out of.

Is it just that French treasury bonds aren’t valued at at-least-par on the market? Is the government of France a bad/risky investment in the eyes of mutual-fund portfolio managers, something not to base the stable part of their portfolios around? (If so, what do French portfolio-managers use as stability? Other governments’ T-bonds?)

(HN is probably not the place to get into an answer of the level of detail appropriate to these questions, so I’d appreciate links to resources answering them!)

> Because the idea of a state [of a country whose populace are not, themselves, in an economic crisis] being unable to just make up some money to give people

but, it's a debt, it increases how broke is the state (in France it's super super badly seen to have any debt, even if financially it is considered an asset, socially the word "debt" is considered as an absolute evil)

as French, I don't really see debt as an evil word (well, depends on the debt, perhaps: debt to buy an asset or realize a project, no, but debt that's taken on to pay for normal operations, like most of the debts of the French state, that's a little more problematic).

> being unable to just make up some money to give people (in the form of e.g. T-bonds), sounds very strange to my ears

Yes, they are free to sell debt. Doesn't mean it's a good idea to just get yourself in ever increasing debt.

> Is it just that French treasury bonds aren’t valued at at-least-par on the market?

That's easy to google. And that is currently:

> The France 10Y Government Bond has a -0.082% yield.

http://www.worldgovernmentbonds.com/country/france/

(Germany is -0.433% and the US is 1.289% for 10Y)

So it's not so much an issue of "being broke" but more of budget priorities

> Yes, they are free to sell debt. Doesn't mean it's a good idea to just get yourself in ever increasing debt.

A state is not a person and can get in debt indefinitely without never reimbursing capital since a state is immortal and can easily reimburse its previous debts with new credit.

The only reason you, as a person, can not have that much debt is because the bank knows that you'll die, but rest assured they would gladly loan you millions if they knew you'd have revenues (taxes) forever.

As a rule of thumb, it's wrong to think that a state should be managed like a home budget or like a company. They are very different (and complementary) economical actors with extremely different properties and goals.

One could even argue that a state with too much spare assets can be seen as morally incorrect as long as unresolved socioeconomic issues exists. And they exists everywhere.

States are no more "immortal" than corporations. Plenty of states have either ceased to exist or repudiated their debts, leaving their creditors without recourse. If a state allows its debt to keep compounding, eventually it will find itself spending every cent of taxes it's capable of raising just to service the interest on that debt—and if anything should happen to shake creditors' confidence in its ability and/or willingness to repay those loans then the government becomes unable to borrow new money to repay its old debts and the entire Ponzi scheme collapses in on itself.

Oh, and issuing new currency to repay the loans is functionally equivalent to defaulting on the debt. Creditors will not lend to you if they expect that you may attempt to pay them back with money which is worth significantly less than what you originally borrowed. A smart creditor will just index the payments to something you can't easily manipulate.

This seems deeply misguided if that’s the reason the are delaying, because owed payroll is also a debt. If you can sell bonds with a negative interet rate, issuing bonds to pay your payroll debt whole makes you richer!

> > The France 10Y Government Bond has a -0.082% yield.

> (Germany is -0.433% and the US is 1.289% for 10Y)

To put this into context: The Euro/ECB's interest rate is -0.50% and the USD/FED's interest rate is 0.25%

> Is this an EU “we can’t set our own monetary policy, we have to ask the ECB nicely to do it for us” thing?

It's more of a "Screw you, we make the rules. What are you going to do, sue us?" thing. The French government is notoriously one of the worst employers in the country. Think unpaid overtime, hiring undocumented workers, low wages, arbitrary requirements, etc.

It's not all bad, but french government agencies notoriously have a bit of a "labor laws don't apply to us" attitude.

> It's more of a "Screw you, we make the rules. What are you going to do, sue us?" thing. The French government is notoriously one of the worst employers in the country. Think unpaid overtime, hiring undocumented workers, low wages, arbitrary requirements, etc.

I think you made the best answer to explain my mind, thanks!

> It's more of a "Screw you, we make the rules.

> It's not all bad, but french government agencies notoriously have a bit of a "labor laws don't apply to us" attitude.

That is sooo true. They give themselves exemptions on so many of their own laws and rules.

Just as an example, junior med students working all mornings in public hospitals are paid below both minimum wage and minimum internship compensation.

In some hospitals, students realised they were paid less than the bonus security guards had for having a guard dog.

They also run residents (7th+ years medical student) on schedules that violate both minimum recovery time and maximum weekly working hours. But they have their own special rule too.

And that’s just the public health sector.

It’s just so wonderful. I too wish I could make the rules for everyone.

Really kind of funny/interesting how the US Federal Gov/US private sector are exactly swapped. The federal civil service (the military/intel agencies are a different story) has a very strong union and gov workers are very well known for doing exactly what is required, but nothing more ("good enough for government work"). Meanwhile, wage theft in the private sector is rampant.

> The federal civil service has a very strong union and gov workers are very well known for doing exactly what is required

That's also something we have in France, with big unions like teachers' and railway workers unions (we've gotten a few nice strikes where all the trains stop for days and I think we're due one soon, I guess once covid's dealt with). And the good enough attitude certainly exists here too!

> make up some money to give people (in the form of e.g. T-bonds),

That's borrowing money.

Making up money to give people is monetizing government finance (print money, spend money, no borrowing), which states with their own fiat currency can also do but traditionally refrain from and pretend to be using a commodity currency that needs to be borrowed because it makes people more comfortable (even when central bank operations end up amounting to just making up money and tossing it into the economy, in a more opaque manner; its kept both opaque and segregated from financing state operations.)

To make this more concrete, the money printer moment in the US is when the federal reserve (central bank) buys the treasury bonds. The cash spent is recorded as a liability, and the bond as an asset in the balance sheet. The cash is created out thin air.

It's all far more complex than needs to be, as PP says.

I like to imagine some Feynman diagram situation where electrons and positrons can spontaneously emerge in pairs. The circulating money and the debt in the central bank balance sheet are the complementary pairs.

The "quantity of money" is a non-concept because only the "net charge"---which is 0 (!) in theory---is preserved.

-----

The older mechanism is the treasury department making coins. The treasury department doesn't have any sort of corresponding "debt" written down when it does so. This is where the "trillion dollar coins" idea comes from.

I suppose the net money is not actually 0, but whatever it was when we came off the gold standard? Who knows!

> The cash is created out thin air.

Isn't that the case with every loan?

Yes! There are soft-money types that criticize charlatanism for characterizing the the government as too distinct from other credit-granting institutions (descriptively, not prescriptively).

That said:

1. I am unaware of any fractional reserve requirement on the Fed, at least one that matters in practice.

2. I am unaware of any time frame in which the outstanding debt needs to be paid back.

All this stuff is quite confusing though, I hope to understand the epicycles here more completely someday, if only to better argue the complexity is accidental, not essential.

Very oversimplified is that the Fed is basically nowadays does "NGDP targeting" with a huge emphasis on as close to full employment as possible.

If the Fed would "pay back" its liabilities (loans), almost all money would disappear. (AFAIK there are some old legal tenders that are not Fed issued.) The central bank is there to supply the money needed to maintain a steady money supply. It's designed to hold that debt.

The basic formula is MV=PQ (money supply * velocity of money = price * quantity). As the economy expands it's natural that it needs more money to maintain the same price level. The central bank then targets a slight increase in prices (to have the steady ~2% inflation) and issues money accordingly.

The Fed transitioned to a zero fraction regime on 2020 March ( https://www.federalreserve.gov/monetarypolicy/reservereq.htm ) It was an unused tool anyway. Now the Fed uses interest on reserves (IOR) and discount rate to guide the fundamental rate. (Discount rate is the rate at which the Fed gives a loan. IOR is the minimum rate, literally free money from the central bank, discount is the maximum of the range. Banks call each other up to get loans, obviously they will ask for rates above the IOR/free-money rate.)

https://research.stlouisfed.org/publications/page1-econ/2020...

https://en.wikipedia.org/wiki/Money_supply#Link_with_inflati...

There is no required reserve for any bank as of 2020 ;D

> That's borrowing money.

I get that's it's a bit different that what GP said, but isn't that also a viable solution for the problem of "not enough money right now to pay wages but in a few months there will be so wait for then." ? Borrow money, pay wages on time, when that lump of money comes in that would allow the paying of back pay, use it to pay off loan (or set it aside to pay off portions as they come due).

The EU has terrible economics baked into their constitution about governments taking on debt too -- fiscal and monetary policy are both constrained.

This should be close enough to chump-change that I am skeptical that's what's going on, but I cannot rule it out either.

> The french state is constantly broke

I'm curious as to why you'd say that and generalize for contractor work. Just as an example, Thales has landed multiple (and sometimes in a shady manner it seems) juicy contracts with the french government. Just looking it up quickly, you can find a 263 millions euros contract for radios for example. [0]

Mismanaged, I wouldn't argue, but it sure spends a lot and is definitely not broke.

[0]. (in French) https://www.tradingsat.com/thales-FR0000121329/actualites/th...

Yeah, the idea that the French government can't pay for IT seems really odd to me. They can definitely pull out the big contracts. It's just a sector with a lot of entrenched interests and regulatory capture.

I personally think on the contrary that the French overspends too much and get kinda milked by big IT companies such as Thales that bank on the fact that no-one on the administration side can evaluate how much the work they're requesting actually costs.

Yeah, mismanagement of these large projects is more of an issue than a lack of budget.

FYI / "FHNI": The GPL also has issues in Taiwan, apparently because it's illegal there to enforce its constraints on anything but the first user / immediate level. Or something like that — not a lawyer here.

Found this paper on the topic: http://www.law.nagoya-u.ac.jp/ls/review/_userdata/10-04.pdf But honestly I don't understand too much of it.

Either way if France joins Taiwan, that sure isn't a good thing for people relying on the GPL, but it won't break its back either. Some "industry of opportunity" may spring up (it definitely has in Taiwan), but it's not the end of the world.

If the enforcement body is corrupt, no licence will stand.

You can still pursue GPL violations in France, it just has to be framed as contract violations rather than copyright violations.

Depending on the context you probably still could frame it as a copyright violations.

My guess is that the problems lies in the aspects of the automatic termination of the contract not being valid in France law in exactly that way. I.e. it being "in between" valid and invalid until the court case around the termination is settled and in turn it means there is still a (breached) contract or similar.

If that is the case you might be able to avoid this problem by making sure the contract is properly and explicitly terminated before you start suing, but it also requires the offender to continue using your software after you explicitly terminated the contract. I.e. you couldn't do it in retrospective, and it probably also would be a problem if there is a ongoing curt case for damages due to the breach of contract...

> The GPL also has issues in Taiwan, apparently because it's illegal there to enforce its constraints on anything but the first user / immediate level.

NEW! License restriction free Linux source code for only $ 9.95 ! Only used once! ORDER NOW!

Um, no. The enforceability about third parties discussed in this paper is about following situation:

1. A released the work under CC 2. A then transfer his economic rights of the work to B 3. then C used the CC licensed work in 1.

Under Taiwan's law, if we treat CC as traditional license then C has the right to use the work due to 1., but if we treat CC as contract then C can not invoke CC in 1. to claim usage rights from B. The author then proceed to argue that in the contract case, base on (1) If the type of use is unspecified, we should consider original intention (2) Principle of abuse of rights. Either B can only seek compensation from A by (1) or the transfer is invalid by (2).

I'm just curious how anyone figures this example is a GPL violation? The GPL states (and Stallman has discussed and stated) that if an organization uses GPL code internally and never redistributes the code outside the organization, it's perfectly fine. Which seems to be this case. So why does anyone think it's a violation and what's the point of this article?

Edit - also, the GPL is clearly a contract. Stallman has touched on copyright issues too, saying (for example) that a game can be open source but the assets still copyrighted. They're different concepts altogether.

I thought Orange Business Services sold it to the French gov, did they not?

Did they sell the software to the French gov, or deploy a managed service that only used this software under the hood (thus avoiding even the "public performance" part of copyright, that might otherwise apply to a SaaS deployment)? The distinction is quite critical to this case, if they did the latter it's no different than Amazon doing the same thing.

Using it for a service still isn't 'redistribution' as per the GPL. It still wouldn't be considered to have left the organization. It's perfectly fine use as per the GPL.

And if it's the French government who contracted the whole thing, they're the original owner, not Orange, but either way it's fine (an owner giving access to a contractor to work on a program containing GPL code isn't considered redistribution either, nor is a contractor using open-source code in a project for a client).

Edit - also even if Orange used it internally, then made the French government a service which uses it, as long as Orange simply re-sourced the code for the government it's fine. It's like Amazon selling hosted Linux images.

think: Orange makes software for the government, that the government pays for, that actually turns out to be existing software owned by someone else and under GPL license (which extends to the whole project and has other implications).

That's very much the definition of counterfeiting.

There's also workarounds and it isn't necessarily 'counterfeiting'. Think Amazon selling hosted Linux OS images. Think commercial software that needs GPL software to run so the user downloads the GPL software themselves instead of it being bundled.

Did the French government ask for source code or seek to do anything else the GPL allows only to have Orange refuse or block the move?

That doesn't seem to be asserted in the article.

Hi, current maintainer of Lasso here; Orange bundled Lasso (a GPL library) with their internal authentication system at the time (called IDMP, itself made as an Apache module) to give it support for SAML then sold a license of the bundled software to the French state; they never distributed the source of the whole package, and they did not distribute the whole package under a GPL compatible license. Clear violation.

Of course IANAL (je ne suis pas avocat). But in the end it may not matter. While GPL2 says "You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works", if a French court decides that if someone modifies and distributes a GPL covered work an enforceable contract is formed, it seems that the effect is the same: they have to comply with the terms.

It does matter: you can no longer sue for copyright violation, you can only sue for contract violation.

If you violate a contract, the court will order compensation calculated to "make you whole" again - to restore things to how they would have been before the contract was executed

The author seems to think there may be some issue for a writer of free software to demonstrate that he had been harmed by the violation; which wouldn't be an issue in a copyright suit. Well, at least here, copyright violation carries statutory damages.

IANAL.

One of the strengths of the GPL is that companies don't like it for some reason. Since they don't like it, you can offer to sell them different licensing terms. That is what the damages are. It's not $0, it's all the money you missed out on from not being able to license your project.

> they have to comply with the terms.

They do but not doing so is a breach of contract not a copyright violation.

What (at least in the US) makes it a copyright violation is the clause in the GPL contract which causes a automatic termination, and in turn makes you end up without a (previously existing but breached) contract granting you usage rights. Which then can be used to sue you based on copyright law.

But everything which causes the then not to be reached causes you not to be able to sue based on copyright but only based on breach of contract.

Through in practice this kinda could make a lot of copyright/licensing a joke (not just for software, but also e.g. art, movies etc.).

So in France, GPL violations are not considered a copyright issue, they're considered a breach of contract. And since the court looks at the damage to the victim in breach of contract cases, which is hard to realistically imagine when the software is given away for free, this is some bad news indeed. Someone uses your GPL code, gets caught, pays a minimal amount in damages, and continues using it.

I'm not a lawyer and stuff.

> So in France, GPL violations are not considered a copyright issue, they're considered a breach of contract.

They are breach of contract in probably any country, but due to a automatic termination clause the breach of contract immediately leads to a copyright violation as you no-longer have usage rights as they where terminated.

So it's always a breach of contract but in the US that breach of contract triggers a copyright violation where in France in that specific case it did not.

> to realistically imagine when the software is given away for free, this is some bad news indeed.

The software is free under certain terms, so you would need to find a realistic price for software without such terms. Which might be hard for a small ad-hoc library, but should be very doable for any larger software project.

> , and continues using it.

This wouldn't work they would pay some amount of damages (see above) for their usage until then, but you also can terminate their contract and in turn they would need to stop using it or (as far as I understand) they then (and not before) could be sued for copyright violation.

Aha. Okay. So (even if you're super pro Free Software like me) this doesn't seem all that bad, I think -- the error was the small guys "going for the money," but the GPL was never much about money, it's about compliance with the license.

Does this seem right? (Yeah, I am a lawyer, but also not French)

I agree that the goal is to get people (actually mostly businesses) to comply with the license, but that is sadly tied to what the consequences are if you are found to be in violation.

The requirement go through contract violation instead of counterfeit will in practice drastically reduce these costs, and thus the incentives for compliance, while at the same time reducing the incentive of the authors to even try defending their rights.

It can be, but as I recall, the crux of the whole router WRT story was the FSF or whoever called up Linksys and said, "You're in violation, but calm down, we don't want money. Let's talk." -- and that is why we have e.g. DD-WRT, Tomato et al today.

If you want to get compliance and legal cases take a decade, then damages need to cover a decade of lawyers, plus however much internal time/cost you spend on it over that decade.

https://en.m.wikipedia.org/wiki/CeCILL CeCILL (from CEA CNRS INRIA Logiciel Libre) is a free software license adapted to both international and French legal matters, in the spirit of and retaining compatibility with the GNU General Public License (GPL).

In my opinion, the EU Public License is the most carefully thought out FLOSS license for European jurisdictions.

https://en.m.wikipedia.org/wiki/European_Union_Public_Licenc...

Enforcing CeCILL in France on the same case would give the same result, if the court use the same interpretation of the law. That's not a problem with the license, it's a problem of the interpretation of copyright law (or "Droit d'auteur") in France with respect to copyleft license (any copyleft license with serious "obligations" to honor, so not BSD and alike, would give the same result).

Would this be fixable by adjusting, somewhat, the terms of the GPL? If the issue is that this is a contractual violation, would it be possible to put preconditions before granting a license in the first place -- such as requiring that legal claims for both copyright violation _and_ contractual violation are explicitly allowed before the license is granted?

This case has both parties in France so the courts will probably rule that the license was made in bad faith since it isn't valid in France, I have no idea what the court will do after that. (in the US the courts will look down on a license made in bad faith, I have no idea about France)

The license already has preconditions. That's part of the problem.

To determine whether there is a violation (copyright or anything), it first requires to determine whether the conditions are respected, which first requires to determine what are the conditions, which first requires to interpret the entire contract to determine the obligations and conditions and everything on both parties.

In software terms, you've got a deadlock, can't consider copyright without considering the contract. (This plays really poorly in practice with these being different claims going to different courts).

It seems to me that if, e.g., mandatory arbitration clauses can be binding, then so could a precondition clause allowing particular types of torts (or disallowing particular types of defenses). I imagine it wouldn't be too hard to convince a judge that engaging in a defense explicitly disallowed by the contract in dispute isn't fair play.

I guess the question is whether you can have preconditions at all on blanket licenses?

For example, if I say "anyone who sends me, by registered mail, a check for 50 EUR will be granted a license", and someone just uses my code without even attempting to send the check, is that copyright violation or contractual violation? It seems to me that would be a copyright violation, or else all copyright violations are just latent contract violations, in their most absurd extents ("I didn't sign the licensing contract they offered, but because they offered one at all, it's a contractual violation" seems to be a particularly weak defense).

Arbitration clauses aren't binding. It's a perfect example of something that is very common and abused in the US, yet has zero bearing in the rest of the world. ^^

A clause that tries to define particular types of torts/defenses is a prime example of clause that is immediately void across the world, because it tries to redefine the law. Contrary to what you think, the judge cares little what is written in the contract, the judge mainly applies the law.

P.S. The GPL is not a blanket license. It's a contract in the US too. I shall make a follow up article on contract vs license because there is a lot more to cover about that.

In my opinion no. France doesn't recognize copyright for software the same way the US does. And since an end user agreement is essentially a generic one-sided contract, the court can reject any clauses that it feels is over reaching.

My understanding of the case as a legal noob:

1) The defendant was not willing to redistribute the source-code it and hence they were sued for violation of GPL from the owner of the software under copyright law.

2) French court did not recognize the validity of software licenses to be enforced under French copyright law.

Does that mean GPL and all similar free licenses that rely on copyright law cannot be enforced in similar jurisdictions? Does anyone have a list of countries/jurisdictions where a similar ruling is likely to happen? What alternatives exist for free software in those countries?

> Does that mean GPL and all similar free licenses that rely on copyright law cannot be enforced in similar jurisdictions

You missed one important point

3) The French court said take this is for the contracts courts, take it to them - something that hasn't happened yet.

The contract courts have not ruled (this hasn't even been brought to them from what I can tell) on this case at all, but probably would find in favor of the GPL. The GPL probably can be enforced in France, but it is a contract violation not a copyright violation which apparently the FSF doesn't like. (I'm not clear on why - possibly because the court award would be much less)

The FSF is not mentioned in this article, nor are they the author of the article. I was surprised when the author wrote "Remember that it’s all about money!" because that is never what the FSF says - in fact the FSF whenever they sue they always say "We don't care about money, we want you to comply with the terms of the GPL."

Good point. Though the courts might disagree and decide it is about money anyway. I have no idea how France will handle this - nor any of the other countries in the world.

> validity of software licenses to be enforced under French copyright law.

Neither does the US (as far as I know)

What happens in the us is that a termination clause in the GPL contract leads to a termination of the contract and only then you can sue for copyright violations.

So my guess is that the difference of law or the interpretation/application of it lies in how/when the termination applies.

> cannot be enforced in similar jurisdictions

They still can be enforced it's just based on a different legal basis, which in general is much less costly for the offender and might lead to less "damage payment" for the offended.

"That is to conclude, by having a contract (a license is a contract), you lose your rights to pursue any IP/copyright claims (the two are exclusive in French law)."

So hold on..

Virtually all software comes with a license of some sort. This seems to be required in EU since otherwise you don't have the right to use software.

And now courts in France think that if there's a license (=contract) involved, then copyright cases are dismissed.

Does that mean I can wipe my butt with the French equivalent of copyright law (Droit d'auteur), as far as software is concerned?

Only if the software does not have a license. It does not mean that copyright does not exist; it’s just one or the other.

This is an interesting yet very frustrating read. Every few weeks we see a post on HN about an MIT-licensed project that gets picked up by AWS or Azure and the general response is "should've been GPL". This case right here highlights that in France that doesn't work either.

Their equivalent of AT&T used GPL software (admitted to it) but claims that it's a contract issue and not a copyright one so the case gets thrown out.

GPL doesn't force you to release your changes to the outside world if you don't distribute it.

Indeed, that's what the AGPL is for.

AGPL only changes things if you give access to your software over the internet. If you're just using the software internally (e.g. it's used in the corporate SSO solution for access to an internal wiki) than no OSS license will force you to publish the code.

Furthermore, if I contract a company to build a service which I will run internally, and they use (A)GPL software in this service that they are building for my internal use, neither I nor the company probably have any requirement to publish the source code to anyone, perhaps except that the contractor may have to provide me with the source code if this was not otherwise stipulated. But you don't have any right to see the code that they modified for me, even if you are the initial author of that code.

AGPL does not seem to apply only over the internet. It is a network. That could be a USB cable or WiFi or a serial cable, possibly even a SATA port.

I don't believe this has everything been tested in court, but I fail to imagine how a USB cable or SATA constitutes a network. Wifi, sure. If you have an IPX or TokenRing or an ISO/OSI network, then sure.

Either way, the more relevant part here is the use inside an internal organization.

It's about intent. If the customer is communicating with you then it's "supposed" to count. This is important, because it covers devices that you talk to to use.

Per the literal text of the AGPL, it's about providing a service over a network via a protocol. So if it's a SATA device the network is peer to peer and the protocol is SATA. I have no reason to believe a real test of the AGPL would be so restrictive.

True. Though I am very curios if there is anyone offering services to 3rd parties over SATA or USB.

The service you are offering is SATA compliance for storage, or USB (sub)protocol compliance for one of myriad reasons. E.g. I have projects where the main interface is USB or CAN bus. I expect the AGPL to apply. I mention this explicitly when discussing the license as well just to be airtight, but it's my belief it is already.

> perhaps except that the contractor may have to provide me with the source code if this was not otherwise stipulated.

They wouldn't have the right to otherwise stipulate. The contractor can't change the license to the software without losing the license to the software. Even if they were distributing a modified version to a single customer, that customer would have the right to the source. That customer, using it internally, wouldn't have the obligation to distribute any of the source or changes but they would have the right to.

Otherwise, you could just sell software as a "contractor" in name only to any number of customers, and effectively relicense it with "stipulations."

Yes, I agree with everything you say. What I meant with that phrase was the opposite - I was thinking that for software subcontracting it is somewhat common to explicitly stipulate that the source code of what is being built belongs to the client. I meant that the (A)GPL would give you the right to the source even if the there were no other stipulations to give you that right to other source code.

Did Orange then "distribute" the library to users as part of deploying this authentication service, or was this purely internal, SaaS use? If the latter, it makes sense that the counterfeiting claims were dismissed, since these only arise when the software is "distributed".

Slight correction: it is usually "should've been AGPL".

Not that it usually actually matters, because if you are a smallish company making most of your money by selling support for your FOSS-licensed software the big cloud providers are going to eat your lunch regardless of what FOSS software license you use.

They will be using your software at sufficient scale that it is more efficient for them to handle support for it for their own use anyway, and by extending that support to their paying support customers that support they were already going to have to do for their own use becomes an advantage for their customers.

The original company's advantage over everyone else providing support and maintenance is that the original company knows the direction of future development. That's a great advantage if the software is the kind for which people really want to rapidly upgrade to the latest version.

The kind of cloud infrastructure services that are usually involved in the "cloud company X killed my business by selling support for my project" stories are the kind of services that people don't rush to upgrade.

> Every few weeks we see a post on HN about an MIT-licensed project that gets picked up by AWS or Azure and the general response is "should've been GPL".

How would using GPL help you in that situation?

OP means AGPL - https://www.gnu.org/licenses/why-affero-gpl.html

> Every few weeks we see a post on HN about an MIT-licensed project that gets picked up by AWS or Azure and the general response is "should've been GPL".

If you are distributing GPL software, the GPL says that you are obliged to provide the source code of the software on request. Amazon or Microsoft use this as a loophole to claim they have no obligation to release the source code of any GPL software they use on AWS or Azure. They claim they are not distributing the GPL software but only executing the softwares on their servers and providing it to the users as a service offering.

This is why the Free Software Foundation created the Affero General Public License (AGPL).

The AGPL includes all the provisions of the GPL but also has extra clauses that makes it obligatory to provide the source code even if the software isn't distributed but made available as a "service" through some server.

"The GNU Affero General Public License is a modified version of the ordinary GNU GPL version 3. It has one added requirement: if you run a modified program on a server and let other users communicate with it there, your server must also allow them to download the source code corresponding to the modified version running there."

(Why the Affero GPL - https://www.gnu.org/licenses/why-affero-gpl.html ). This is why everyone now recommends that you should use the AGPL (instead of the GPL / MIT / BSD license etc) for open source projects.

> This case right here highlights that in France that doesn't work either ... it's a contract issue and not a copyright one so the case gets thrown out.

The French court did conclude that there is no case to be made under copyright law, and that it considered the GPL as a contract. And so any dispute between parties on the GPL can only be judged under France's "contract laws". This doesn't mean that the GPL is no longer valid in France. France is just asserting that the GPL is a contract. You can still take anyone who violates the GPL to court. You just have to file the case under the right law. When you claim in court that someone in breaking the law, you have to point out the right laws that you claim are being violated, or your case will be thrown out.

(It's like adultery laws - adultery was once a criminal offence. It is now treated as a civil offence in many countries. Meaning, you can no longer file a complain with the police if your spouse cheats on you, and have them investigated, arrested and tried in court as a criminal. But you can still sue your spouse in court and get damages.)

Just to clarify even further, it's not only AGPL instead of the GPL (some times it's also GPL instead of MIT), but the cloud providers would be perfectly conformant with the AGPL too.

Usually the HN people that are complaining are talking about market concentration.

Hi, I'm one of the author of Lasso and the current maintainer, still working at Entr'ouvert. Ask me anything :)

Do you know if the company appealed to the cour de cassation?

Yes, we started the process.

Couldn't it be argued that it is unfair for that to be required to be qualified as a breach of contract rather than "copyright" (droit d'auteur) violation if anybody can "contract" with you without even you being notified about it?

Or would that qualification actually be not that much interesting and it is "just" a matter of knowing that in France you must sue for contract violation in case of GPL violation?

> Couldn't it be argued that it is unfair for that to be required to be qualified as a breach of contract rather than "copyright" (droit d'auteur) violation if anybody can "contract" with you without even you being notified about it?

I don't know.

> Or would that qualification actually be not that much interesting and it is "just" a matter of knowing that in France you must sue for contract violation in case of GPL violation?

I prefer this question. As stated by the blog author, counterfeiting gives more protection and ensure greater reparation than the contractual way. That's were the "cour d'appel" did not respect the opinion of the CJUE in our opinion. The contractual way does not give the same protections as the counterfeiting way, contrary to what the UE directive asks of the transposition into national laws.

The very important point that a jurist friend noticed: there were past or current contracts between Orange and Entr'ouvert.

Extract from the appeal judgment: "However, it appears from the elements submitted to the proceedings that as early as 2004, the parties maintained business relations, during which the Orange company very quickly showed its interest in LASSO by requesting various information/trainings and services on this software."

Training + Services = contracts. So the judgement actually makes complete sense.

(Original text: “Or, il ressort des éléments versés à la procédure que dès 2004, les parties ont entretenu des relations d’affaires, à l’occasion desquelles la société Orange a très vite montré son intérêt pour LASSO en sollicitant divers renseignements/formations et prestations sur ce logiciel.”)

This article has me confused about the legal status of licenses in US copyright law. Are they not contracts? I mean, the GPL promises the recipient something (the right to make copies of a software) in return for something else (the source code for any modifications). That smells like a contract to me.

It seems that unlike in civil law countries, in common law countries such as the UK and US, a license is a separate legal category from a contract. A license can be interpreted as a contract as well, but you can make a "bare license" that is simply a grant of permission and not a contract all. In civil law countries such as continental Europe, all copyright licenses are interpreted as contracts.

My understanding is based on this LWN article summarizing the opinions of several legal experts on the subject: https://lwn.net/Articles/747563/

How about I make a second article to explain licenses vs contracts in US and the rest of the world?

Your question went to US court in 2017. I wanted to cover that briefly but the blog is already quite long without getting into more specifics about licenses in the US.

A (incomplete) Tl;DR:

The strategy that is well tried in the US is to go for counterfeiting/copyright, to seek the most damages. Copyright/IP law can carry significantly more damages, both in France and in the US, compared to a contractual dispute. (Remember that it’s all about money!)

For an IP violation, the court typically looks at the gain of the counterfeiter. That may be winning and executing the government contracts for Orange. My guess is a government contract like that can go for 7 to low 8 figures.

For a contractual violation, the court typically looks at the loss of the victim. That may be some licensing fees and contracting fees for Entr’Ouvert. My guess is a library like that can go for 6 figures, while the bigger player selling OIDC/SAML solutions (Sun, ForgeRock, Oracle, Microsoft) go for an amount per user per year, easily adding up to 7 figures (but I don’t think that one standalone library from a small vendor in 2003 can do that).

Entr’Ouvert followed the counterfeiting strategy, they sued for counterfeiting and got dismissed. (I should stress out that it was a very reasonable strategy, similar law exist, it should have worked, it just didn’t).

And these people will go around asking why there is so little good software built in EU and why is funding for startups so scarce.

France is not the EU.

France law applies in France, not any other EU country.

Rulings likely would have been very different in many of the other EU countries.

On the contrary, the majority of FLOSS is developed in EU.

This might be a very big thing for other reasons too

...What does this imply for projects like Github's Copilot?

https://en.m.wikipedia.org/wiki/Vernor_v._Autodesk,_Inc.

Has some intersection with this case

It's too bad that AGPL wouldn't have been more effective; this seems like a dead end for copyright-based copyleft in France. We'll have to make extra effort to phrase things in terms of contract-based copyleft.

Although, doesn't this suggest a meeting of the minds? Free Software is usually proffered and consumed without such a meeting. So was a contract actually formed? I'm skeptical.

W.R.T. Copyright assignment the FSF constantly complains "we've hired the lawyers and done the work, you all need to tread more carefully", but situations like this do seem to undermine that.

As an American reading through this, this has a distinct smell of French courts ruling in ways favorable to the national ISP because it might cost the government money to do otherwise.

Basically - this looks pretty bad from my perspective.

Anyone with more context who can explain how copyright could exist in France if all licenses are considered contracts instead?

The other side (Entr’Ouvert) is also French. I don't see how you can think this is a case of France against not France.

One is a national asset and one isn’t? (Perhaps I’m wrong)

Nobody thinks it's France v !France. Its big national corp (Orange) vs a little guy. The court sided with the BigCo. That's the issue.

Yeah, if Orange was willing to do this once, the chances are that their tech stack would be swiss cheese if they had to excise all the GPL-licensed tech they are using without rights. The precedent is more important to them than the $3mm.

> to the national ISP

It is the previous national ISP. I'm not even sure it has ever been a private ISP. It's completely private now [0] (but still owned at 30% by the government) and they have three big competitors: SFR, Iliad, Bouygues

[0] https://www.google.com/search?client=firefox-b-d&q=EPA:+ORA

Or it's French courts making a ruling which is compatible with French law and incompatible with an American's understanding of what French law should be. You can read about it here [i].

The point is that behavior contrary to the terms of a license is considered a contract breach and not a copyright violation, so Orange didn't violate copyright. It's yet to be determined whether they violated the terms of a contract (the court in question can't make that determination.)

[i]: https://en.wikipedia.org/wiki/Copyright_law_of_France

So if you're aware of the specifics here, help me out with the question I asked.

How can you possibly have a valid copyright for software in French law if all license terms are considered contract breaches?

From your own link... - >>The legal position was resolved by the transposition of May 14, 1991 EU Directive into French law: computer programs and any associated preparatory works qualify for copyright protection in France as in other European Union jurisdictions.

Once a work has been published, the author cannot prevent:

1. Private family performances.

2. Copies for the private and personal use of the copier. This provision does not apply to works of art, computer programs (where a single safeguard copy is allowed, Art. L122-6-1-II) and databases.

3. In cases where the name of the author and the source are clearly indicated,

a) Analyses and short citations justified by the critical, polemical, scientific or pedagogical nature of the work.

b) Press reviews.

c) Diffusion of public speeches as current news.

d) Reproductions of works of art in catalogues for auctions in France (subject to regulatory restrictions).

4. Parody, pastiche and caricature, "taking into account the usage of the genre".

5. Acts necessary to access a database within the limits of the agreed use.

But how can this be true? Orange clearly does not fall into any of the excluded uses, and I must release my software with some sort of licensing statement (even the lack of a statement is a statement in my jurisdiction). But it appears this ruling would mean all breaches could only ever be contract breaches.

So again, how do I have copyright for software in France? I'm genuinely trying to understand if there's some nuance here that I'm missing.

> How can you possibly have a valid copyright for software in French law if all license terms are considered contract breaches?

Perhaps you are confusing the copyright with the licence.

The copyright adheres strictly to the author. He can sell it, in which case he sells the right to licence and enforce.

If he licences, then he retains the copyright, and grants the licensee specific rights in exchange for specific actions. That's a licence agreement; to my mind, the word "agreement" signifies a contract. I find it a stretch to interpret a licence violation as a copyright violation.

Ok, so in the context of this case:

The legal system of France lays out two different penalties for different behaviors

1. Violation of copyright/IP, which is punished based on how much gain the counterfeiter makes.

2. Violation of contract, which is punished solely on how much money the party violating the contract caused the other party to lose.

So my take is still pretty firm that you have no copyright for software in France - Or (and maybe we're getting closer here) to restate: Copyright for software in France is not equal to Copyright for other intellectual property, such as art/song/performance. Since this ruling would prohibit anyone from ever getting the much more meaningful ruling that the other party is a counterfeiter, not simply a party in breach of contract, if the IP in discussion is software.

In which case we could call that copyright, but it feels like a misnomer.

That’s not quite right. As I understand it (IANAL etc), the same rules apply to art/song/performance as to software. The key difference is whether the misbehaviour is carried out by a licensee. Once you license a creative work to someone, whether art or software, breaches are violations of the contact, not the copyright. In contrast, if the copyright holder hasn’t granted a license, then it’s a copyright violation, not a contract breach.

So software does have copyright in France, but it can’t be used for copyleft licenses in the same way as in the US. That doesn’t mean it isn’t copyright.

You say you must release your software under a license. But you don’t have to make everyone in the world a potential licensee. No one is compelled to grant universal license to distribute their software. They can just decline to do so and rely on copyright laws.

You’re defining copyright narrowly as “that which allows open source copyleft licenses like the GPL to work.” But that’s based on a US-centric understanding of copyright that simply doesn’t apply in France.

And it’s not surprising. The GPL is a hack of the copyright system. It uses copyright to do something it wasn’t originally intended to do and is only capable of doing because of the peculiarities of the US (and other) legal systems. Hacks often don’t work outside of their original context.

Heh, and I get down voted here when I comment that the GPL is probably a contract...

Oh well, I don't have a Harvard law degree so what do I know

You got down voted for sarcasm in a way people don't like and missing context, through yes GPL is a contract and you don't sue for GPL violations in front of the copyright curt but for people not having a license as the GPL contract was automatically terminated due to violations. (As far as I know).

Reply on news.ycombinator.com