Talk:Open Data License/Timeline
Users who have not edited ever
In "Technical work" it says:
Move all users who have not edited ever to the new license automatically
Is this a good idea? After all, those users that have signed up while the old license was in force would expect their contributions to be under that license, not the new one. Probably better to force them to accept/decline the license like all other users too. --Schuetzm 12:07, 20 September 2008 (UTC)
- Surely this is not a separate step, it just happens naturally. All users need to agree to have their data shared under the new license the next time they log on after the license is implemented. If they don't agree, their account is suspended, and edits they made in the past are purged. If they haven't edited anything yet, it just makes the purging part easier :-). --Rjmunro 22:41, 20 October 2008 (UTC)
How much data do we have to purge
make clear that all their edits will be removed if they do not respond or vote no
be implemented for objects with edit history of various users? Let's say XY declined to move to the new licence. Now a few cases:
1) An object was created by A, then edited by none or several others, the last edit by XY => Simple fix, remove edit of XY
2) An object was created by A, then edited by XY, then edited by several others => What shall we do?
3) idem 2), but with a last edit again by XY => Just remove the last edit or else what shall we do?
4) An object was created by XY, then edited by several others => what shall we do?
There are a number of further cases possible. I believe we first need a decision on what "remove all edits" translates to in reality before implementing the procedure. --Notaris 07:32, 23 September 2008 (UTC)
- We have some degree of leeway here. An "insubstantial" contribution does not merit copyright/database right protection. Any user who has only made an insubstantial contribution does not need to have their edits removed. --Richard 09:32, 17 October 2008 (UTC)
- If their contribution is insubstantial, it doesn't matter if we remove it. :-) --Rjmunro 22:41, 20 October 2008 (UTC)
We will have to get users to agree to a dual license arrangement to begin with, something like "Data is licensed CC-BY-SA, but do you agree to allow it to be additionally licensed ODBL?" Then we have an overlap period of a few months where we get users to agree to the license switch, during which time all downloads are CC-BY-SA, but all users agree to both licenses.
How long do we keep CC-BY-SA as a license option? --Rjmunro 22:41, 20 October 2008 (UTC)
"remove past edits" = suicide editors base?
I am a bit upset by the apparently already-made decision to kill all previous edits from those that won't say "yes".
Indeed, I found a much more balanced proposal (consider maintaining present of ree access license in //) in the recent paper there: 
Isn't there a real risk that if OSM starts killing an unknown percentage of its own base, just for the purpose of clarifying the license, the first result will be a poorer database?
Don't we expect a dangerous lobbying from all private map sellers, and dozens of papers/articles saying "OSM is currently deleting a huge part of its own data... so come back later"?
To be clear, I'll click "yes" on whichever license change I'll receive.
My question is more about the ones that won't answer, our unability to predict which percentage of the common work will be trashed, and consequences including press comments and general public impact (not just tech. database size).
In my country, France, we have a saying to describe my admittedly worst-case scenario: "the surgical operation was a success (though the patient is dead)".
I fully agree on working to clarify OSM database open status, but I don't want to risk crashing it for this: I believe being careful on this specific issue is of the essence here.