[dpdk-dev] Beyond DPDK 2.0

O'Driscoll, Tim tim.o'driscoll at intel.com
Thu Apr 23 13:36:03 CEST 2015


> From: Stephen Hemminger [mailto:stephen at networkplumber.org] 
> 
> This does a good job of stating the need for action without getting into the details.
> Perhaps this would be better resolved by some more interactive discussion.
> I know it is hard to all get together, but there needs to be more some more creative and focused
> thought on this. A phone conference is just not enough.

I agree. It's difficult to know the best way to progress it. We were hoping that there would be some debate on the mailing list first, and then we'd have one or more calls to discuss it. We can try holding a call on this anyway, but the risk is that nobody has anything to say. Maybe that's something we should try anyway though. I can set something up.

> Alternatively, propose some options and vote, but I don't think we have things defined
> enough for that yet.

We tried to keep the initial communication neutral and avoid suggesting solutions to give others a chance to comment. At a very high level, there seem to be 3 possible approaches though:

1. Do nothing. The project is increasing in size, and the releases are getting delivered according to the roadmap, so one option is to continue as we are.

2. Add a more formal governance structure to dpdk.org. This might involve putting in place a Technical Steering Committee to give long-term technical direction to the project, and to resolve any differences of opinion that don't reach a conclusion on the mailing list.

3. Transition the project to an organization such as the Linux Foundation, and use their help to implement a more formal governance structure. This would probably involve a TSC, and possibly also a governing board and the creation of some form of centralized marketing/branding/promotional budget for the project.


More information about the dev mailing list