License clarification of stavekontrolden.dk downloads

[Freso]

I assume the triple GPL/LGPL/MPL for {,th_}da_DK.{dic,aff,dat,idx} is an OR license ((GPL-2.0-only OR LGPL-2.1-only OR MPL-1.1) using SPDX license expression), ie. that one is free to use those under any one of those licenses without adhering to the other two?

For the hyphen files (is that any files other than hyph_da.dic?), is “License: GNU LGPL license.” the same as LGPL-2.0-or-later or which LGPL is it?

Are the rest of the files[1] under the same license as this repository? (Ie., GPL-3.0-only – I don’t see any mentions of “or any later version” in the source tree, so I assume it’s -only at least.)

[1] the rest of the files

• HYPH_da_DK_README.txt
• README_da_DK.txt
• desc_da_DK.txt
• desc_en_US.txt
• description.xml
• help.tree
• manifest.xml
• page1.xhp
• page2.xhp
• Trold_42x42.png
• Trold.png

[jeppebundsgaard]

I have to admit that I am not very well-versed regarding licenses.

The Danish dictionary files are the result of many peoples work over a long time period. At the moment only 1½ person is active (I am the ½, the other person is doing all the work with the dictionary - I support by developing stavekontrolden to his needs).

I think that your interpretation is reasonable - it is one of the three licenses (OR), not all of them (AND). The hyphen files were created many years ago (before I was part of the project). So I guess there was no versions of GPL at the time? If it is acceptable legally to update the licenses, I am sure it would be fine to update all of them to latest and add "or later".

Stavekontrolden itself is my work alone, so there I have control of the licensing - and I guess it would be a good idea to uodate to AGPL 3.0 or later?

Thank you for your help.

[Freso]

Well, first of all, I’m not a lawyer, so I will not make any comments with regards to what is or isn’t legally acceptable… :)

The hyphen files were created many years ago (before I was part of the project). So I guess there was no versions of GPL at the time?

LGPL 2.0 was released in 1991, LGPL 2.1 in 1999. The hyphen readme says it was first made in ~1988, so predating the LGPL, but then with changes made in 1994 which might indicate the LGPL 2.0, possibly? Not sure when the "conversation author" contributed and whether that contribution affects the copyright ownership in any way. The only license in this repository seems to be LGPL 2.1 (seemingly "or later", but it seems a bit tangled indeed and I don’t really know enough about the structure/history of the project to make any educated guesses).

Stavekontrolden itself is my work alone, so there I have control of the licensing - and I guess it would be a good idea to uodate to AGPL 3.0 or later?

AGPL vs. GPL is mostly a question of preference. If this includes the website code, then AGPL would improve the protection against someone running a fork of the website on their own server without sharing their modifications.

Regarding licenses, one thing I would suggest (if you are able/where possible) would be to release “pure data” under CC0. Given that this is how the data from “Det Centrale Ordregister” have been released data and that it’s mostly (if not solely) a collection of facts (thus at best possibly protected under EU “sweat-of-the-brow”/database copyright, but possibly not at all – and e.g. US copyright law doesn’t have any “sweat-of-the-brow” doctrine at all), it might be good to just (where possible) put it in the public domain so there’s no confusion across different jurisdictions and would also make it interoperable with, say, Wikidata’s lexicographical data as well as the COR dataset without having to consider a lot of legal “what about”s.