[llvm-foundation] [cfe-dev] Holiday update on LLVM's Relicensing effort

C Bergström via llvm-foundation llvm-foundation at lists.llvm.org
Thu Dec 20 23:30:20 PST 2018


I'm certain you have considered this, but I do wonder if this is possible.
Feel free to ignore this mostly "thinking out loud" reply..
----------------
Person A changes lines 1-10
Person B changes lines 2-25

Who owns the copyright on lines 2-10? They are modifications to Person A
code.

Unless you can get *every* contributor, which I'm a bit pessimistic, then
you're faced with some unknown amount of lost code. I don't see how you can
safely determine every circumstance of "modifications" without running into
some level of risk or obstacles.

If you have decided that all the modifications you can't safely vet get
removed, then what's the plan? Clean-room rewrite? That's a whole other can
of worms since people who have already worked on the project may be
"tainted". This is another risk factor...
---------------
I think what I'm asking myself is who decides how much risk this open
source project is willing to take in order achieve this relicense goal?

Is this council of 7 foundation members vote and decide?
You?
Chris?
Peanut gallery survey...

Lawyers from my experience will advise, but someone else will typically
have their finger on the button to make the bottom line decision.
-----------------
All this aside.. good luck on this project. I personally do think it's a
good idea from a long term standpoint regardless if the big goal is met or
not. If nothing else you move to a direction of having contributor
agreements and stop the bleeding. I guess at some point all code will churn
and eventually it will be 100% coverage.


On Fri, Dec 21, 2018 at 10:43 AM Chandler Carruth <chandlerc at llvm.org>
wrote:

> On Thu, Dec 20, 2018 at 5:49 PM C Bergström <cbergstrom at pathscale.com>
> wrote:
>
>> Is there an elephant in the room? You missed what I'd think is the most
>> important number. How much of the code is or isn't covered? Congrats on
>> almost/maybe/sorta getting all recent work covered, but what about all the
>> work past that?
>>
>
> Not at all being skipped.
>
> The milestone we're targetting is putting the new policy and license in
> place for *new* contributions so we stop accumulating things we need this
> for.
>
> Once we get the new policy and license in place for new code, we'll
> definitely switch to *exactly* the metrics you describe and begin working
> to push that number down.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-foundation/attachments/20181221/1877ecc6/attachment-0001.html>


More information about the llvm-foundation mailing list