[ovs-dev] [PATCH 1/4] datapath-windows: Base code for developing the Hyper-V switch entension.

Ben Pfaff blp at nicira.com
Wed Jun 25 21:57:34 UTC 2014


On Mon, Jun 23, 2014 at 05:34:56PM -0700, Saurabh Shah wrote:
> This is the "Hyper-V Extensible Switch extension filter driver" sample code
> available at:
> http://code.msdn.microsoft.com/windowshardware/Hyper-V-Extensible-Virtual-e4b31fbb
> 
> The sample code is licensed under Microsoft Limited Public License version
> 1.1. The license is available here -
> http://msdn.microsoft.com/en-us/cc300389.aspx (Exhibit B).
> 
> Our core Hyper-V switch extension builds on top of this sample code to develop
> the forwarding engine.
> 
> Signed-off-by: Eitan Eliahu <eliahue at vmware.com>
> Signed-off-by: Guolin Yang <gyang at vmware.com>
> Signed-off-by: Linda Sun <lsun at vmware.com>
> Signed-off-by: Nithin Raju <nithin at vmware.com>
> Signed-off-by: Saurabh Shah <ssaurabh at vmware.com>

The authorship on this patch is a bit murky.  It appears that
everything added in the patch was written by Microsoft (did VMware
modify it?), so why is it signed off by five VMware people?  One, I
could understand, since someone has to compose the patch; whoever did
that should be the author of the patch.

Unless Eitan or Nithin wrote up the patch, it's not time to add them
to AUTHORS.

The license, not just a URL to it, should be included in the source
tree.  I think that that is always a good idea, but 3(d) in the
license says so explicitly.

I don't think that the URL you provide for the license is correct.
When I go to the URL for the code, then click on the MS-LPL link
there, I get slightly different text (e.g. the wording of 3(f) is
different there).

I question whether we can include this in the source tree at all.
License 3(f) only allows distributing code "that run[s] directly on a
Microsoft Windows operating system product, ..."  Source code doesn't
"run directly" on anything at all, so this might forbid distributing
source.  I think we'd need to get a lawyer's opinion.  Have you asked
a lawyer about that?

3(f) definitely violates DFSG #6 (see
https://www.debian.org/social_contract), so if this code goes in then
it'll have to be removed during packaging for Debian upload.  (I don't
know whether Fedora etc. have a similar policy.)



More information about the dev mailing list