3 ms·
That will lead to even more confusion in the minds of the hundreds of policy makers and courts worldwide, who already have plenty of opportunities to be confuse
by reedciccio 3y ago
That will lead to even more confusion in the minds of the hundreds of policy makers and courts worldwide, who already have plenty of opportunities to be confused. More: https://blog.opensource.org/court-affirms-its-false-advertising-to-claim-software-is-open-source-when-its-not/ https://blog.opensource.org/court-affirms-its-false-advertis...
Use other terms, there are plenty of other options
- Pannoniae 3y agoI agree with what you say, but the blog post you linked is hugely misleading. (remember, it's the OSI blog, of course they are going to twist it for their goals) The court did not affirm that "calling your non-OSI-approved licence open source is false advertising". As in the judgement: "It is undisputed that Defendants made the statements at issue to convince customers to adopt ONgDB over Neo4j EE. Because Defendants misrepresented ONgDB as a free version of Neo4j EE licensed under the APGL, there is no doubt that this price differential (free versus paid) was likely to influence customers purchasing decisions. See Hinojos v. Kohl’s Corp., 718 F.3d 1098, 1106–1107 (9th Cir. 2013), as amended on denial of reh’g and reh’g en banc (July 8, 2013) (recognizing under the UCL that price is material to purchasing decisions). Thus, the Court finds that Defendants’ statements suggesting that customers could obtain a “free and open source drop in replacement” for Neo4j EE were material." What the court did affirm is that the upstream licence was a modified AGPL with additional restrictions. The defendant has stripped those restrictions (claiming the AGPL "you may remove" clause allowed them to) and distributed the product as "free and open source". This was false advertising as they were not capable of stripping the Commons Clause and the copyright identification. Notably, this does not mean that it's false advertising to claim your software is "open source" when it is not OSI-approved, it means that if you offer a modified version of a piece of software, you can't advertise it as "free and open source" when the forked version is still bound by the Commons Clause prohibiting selling it in any way.
- shkkmo 3y ago> means that if you offer a modified version of a piece of software, you can't advertise it as "free and open source" when the forked version is still bound by the Commons Clause prohibiting selling it in any way. This argument makes no sense. If the original modified AGPL license was "open source", then it wouldn't have been false advertising to make that claim since that was the license that actually applies. From the original (non-appealate) decision: >As discussed above, Defendants’ claim that ONgDB is free and open source Neo4j is false because it relies on an interpretation of the Neo4j Sweden Software License that this Court has rejected. The actual argument here would be that the usage of "open source" only constitutes false advertising because it was being used in the context of claiming to be different from the original license, thus implying a meaning in-line with the one used by OSI. It is quite possible that in other context (i.e. not part of a direct comparison to a less open license) that courts will decide that does not constitute false advertising.