1 |
Apparently, though unproven, at 20:49 on Wednesday 18 August 2010, Joerg |
2 |
Schilling did opine thusly: |
3 |
|
4 |
> > remember it was running on one of the BSD's, too, since it's a matter of |
5 |
> > licensing that is the hurdle of greatest height. I've only played with |
6 |
> > BTRFS on my dev box and the simple workout I gave it did not tax it in |
7 |
> > any way--it worked okay. |
8 |
> |
9 |
> ZFS has a very free license. This was the reason, why it could be ported to |
10 |
> the BSDs. So why do you believe there is a "license hurdle"? |
11 |
|
12 |
You appear to not fully understand the licenses. |
13 |
|
14 |
Remember that the Linux kernel is GPL-2 and it's modules are considered |
15 |
derivative works. The GPL-2 license demands that all derivative works be |
16 |
either GPL-2 licensed or 100% compatible with the GPL-2. |
17 |
|
18 |
ZFS is licensed CCDL which although free and liberal, is not GPL-2 compatible. |
19 |
It is BSD-compatible which is why the BSDs can (and some do) ship it. |
20 |
|
21 |
The ZFS license is thus not a mere hurdle, it is an un-overcomeable barrier in |
22 |
it's current form. If Oracle were to re-license it then the problem could be |
23 |
solved, but few in this game hold any hope of that ever happening. |
24 |
|
25 |
But all of this has been hashed to death many many many times here and in |
26 |
other places - to the point where it is now conclusive. Google will reveal the |
27 |
entire discussion in all it's painful detail. Start with lkml. |
28 |
|
29 |
Let's not rehash it here again. Please, I beg of you. Let us not do that. |
30 |
|
31 |
-- |
32 |
alan dot mckinnon at gmail dot com |