On Wed, 20 Jun 2012 13:09:13 +0900 Simon Horman <[email protected]> wrote:
>> > In general the policy for other differences, in particular different >> > wildcard rules, is to adopt the most general rule for both OXM and NXM. >> > >> > Personally I might have chosen to implement stricter enforcement of OXM and >> > NXM parsing. However, it does simplify the code and I believe that is why >> > Ben Pfaff prefers the approach taken by Open vSwtich. And I'm not entirely >> > convinced that stricter parsing would be a win in practice. >> >> What the strict enforcement means? > > This is not an example based in fact, but for the sake of discussion > suppose that NXM allows masking of port numbers and OXM does not. In > that case the Open vSwtich decoder would actually allow masking > of port numbers for both OXM and NXM. Thus the parser does not > strictly enforce the Open Flow 1.2 specification of OXM. That is what I > meant by strict. I see. btw, about Ryu, I have no intention to share the parser between NXM and OXM. NXM will be extended from now on. 1.1, 1.2, 1.3 OXM will not. I don't want the possibility of the regression to the existing version of OXM when I extend NXM. I have no problem about sharing the code that is unlikely to be changed in the future between NXM and OXM however I don't think that it's a good idea that OXM uses NXM's ever-evolving code. ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Ryu-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/ryu-devel
