u1db-discuss team mailing list archive

[John Arbash Meinel <john@xxxxxxxxxxxxxxxxx>]

This came up on an IRC discussion, so I figured I'd start a conversation 
here.

When writing up the prototype code, I just made it GPLv3 on the grounds that

1) It is a Canonical approved license, so we don't have to get explicit 
approval for it, or discuss it.

2) If we want to go LGPL/MIT/BSD/ALGPL/... we can always do so 
afterwards, and people feel like we're giving them more access to the 
code, rather than realizing we wanted to be GPL from the beginning and 
'taking away' access.

3) There's at least a discussion of what LGPL/GPL means in the case of 
Python code, that doesn't really have a binary/compile step. (Robert 
Collins has discussed that quite a bit in various forums.)

I personally don't have a big stake in it. LGPL/MIT seems a reasonable 
fit for a library that we want to give to application developers to make 
our platform desirable to program for.

Something explicitly PSF compatible seems reasonable. I certainly don't 
see something like this as ever being in upstream python, but if you're 
writing in python, a compatible license seems fitting.

I don't think there is an urgent need to decide, though I think we'll 
want to have a decision by the P release as well, and doing it earlier 
is just fewer lines of text to replace :).

John
=:->