Re: State of ZFS in 9.0_BETA

[Greg Troxel <gdt%lexort.com@localhost>]

Chavdar Ivanov <ci4ic4%gmail.com@localhost> writes:

> On Sat, 10 Aug 2019 at 17:19, Martin Husemann <martin%duskware.de@localhost> wrote:
>>
>> On Sat, Aug 10, 2019 at 06:17:50PM +0200, Hauke Fath wrote:
>> > On Thu, 8 Aug 2019 16:46:46 +0200, Marc Baudoin wrote:
>> > > ZFS has been updated for 9.0_BETA.
>> >
>> > On the same topic, is there a perspective for
>> >
>> > file-system ZFS             # Solaris ZFS
>> >
>> > (i.e. a monolithic, non-module kernel)?
>>
>> I think this is not possible due to licensing issues.
>
> IANAL, but as far as I understand it, nothing in the CDDL licence
> precludes the static inclusion of ZFS into the NetBSD kernel.
>
> If someone develops a product on the basis of the NetBSD kernel AND
> this product includes modifications of the CDDL-licensed parts of the
> kernel THEN he will be obliged to publish the source code of only the
> files under CDDL (a file-based license) which have been modified. So
> it introduces an element of obligation above and beyond one expects to
> have when using a BSD-licensed software, but does not taint the rest
> of the system in any way.
>
> I personally don't see a problem in having GENERIC+ZFS kernel
> configuration, if this were technically feasible. But, e.g., if a
> commercial entity decides to produce a SAN appliance based on NetBSD
> and using ZFS, and if any CDDL-licensed files have been modified,
> their source code will have to be produced.

Perhaps true, but if the module scheme works for normal cases, little is
gained for NetBDS to distribute a kernel like that, and it adds
complexity thinking about licenses.