> On Apr 28, 2015, at 3:07 PM, Ben Pfaff <b...@nicira.com> wrote: > > On Mon, Apr 27, 2015 at 10:14:50PM -0700, Justin Pettit wrote: >> Limit the supported tunnel types to Geneve and STT on hypervisors, and >> VXLAN, Geneve, and STT on gateways. This commit adds "DESIGN.md" that >> describes the reasoning. >> >> Signed-off-by: Justin Pettit <jpet...@nicira.com> > > I like the content, but did you consider adding this content to > ovn-architecture? I think that's a good place to document design > decisions.
I did, but I think of the architecture document as describing how the system works; whereas this is really justifying why we made certain decisions. I'm afraid adding too many details about the "why" in the architecture document makes it less focused and relevant to its intended audience. For example, the architecture document should definitely describe how the bits in the tunnel metadata are laid out. However, I feel that a discussion of why we chose that approach would be an uninteresting digression to most people. As you and I discussed off-line, we could just add these design decisions as a sort of appendix to the architecture document. Since you feel more strongly about this than I do, I moved the content to a new "Design Decisions" section header within the architecture document. Let me know if you like that approach. > Acked-by: Ben Pfaff <b...@nicira.com> Since I moved things around and I want to resend the series anyway, I'll wait for your Acked-by on the next round. --Justin _______________________________________________ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev
