Since I was reminded (by you! :)) that I never commented - here are my comments: all of this is fine with me. Proceed at full throttle. -jason On Nov 6, 2012, at 8:42 AM, Jeroen van der Ham <vdham@uva.nl> wrote:
Hi,
I'm trying to go through all of your comments. Most are very sensible and I'll apply the changes as suggested. The new version is at: https://forge.ogf.org/sf/docman/do/downloadDocument/projects.nml-wg/docman.r...
For other changes that I have not (yet) included, comments are below:
- Abstract "computer networks" -- the idea of Paola's GreenSonar talk (and other things we are trying to do) is that we combine descriptions using different ontologies. So yes, it makes sense that we broaden the scope, but I feel that NML should focus on computer networks.
- Introduction "sub-classed" -- in some way it does map to an OO context, but slightly different, which is to be expected in this different context. It is explained in more detail in the doc and in the examples, so I'm proposing to keep it like this.
- Schema unidirectionality -- the Port object itself does not define direction, but the relations to that Port (hasInboundPort / hasOutboundPort) do define a direction.
- Schema "Label" -- the Label is an object which is described later (just like other objects referred to), so I'm proposing to keep it like this.
- Schema "isSerialCompoundLink" -- New blurb added to the Link text: A \emph{Link} object can have a \emph{isSerialCompoundLink} relation to a \emph{List} of \emph{Links}. This describes that the \emph{Link} represents a path through the network implemented by that (ordered) \emph{List} of \emph{Links}.
- Schema Service -- We don't explicitly mention extension possibilities for other objects, should we really mention it explicitly for the Service object?
- Schema Relation -- The table is a nice idea, but in practice I think it would take up too much space, and would double much information. You would have four columns: the name, the domain, range and definition of the relation: for example:
existsDuring | Network Object | LifeTime | relates a LifeTime object to a Network Object
It really becomes quite wide, and for most would not fit on a single line anymore.
-- Schema Extension Paragraph-- Very good suggestion, I'll add something on this soon.
-- Identifiers interpreting -- I would love to include something to describe why you should not be interpreting identifiers, but I don't know a succint way of doing that. Does anyone have a good idea?
-- Informative References -- I looked up the article you suggested to add. It mostly describes the NMWG format. Is that still relevant? If so, a paper on NDL would similarly be relevant, and I could add some other papers from other projects too.
-- Acknowledgements -- I think it's proper to add financial support from different projects, should we do that on a personal or general level? (I have to add grant numbers also).
Jeroen.
On 4 Nov 2012, at 21:09, Jason Zurawski <zurawski@internet2.edu> wrote:
Hi Jeroen/All;
Please find attached my review/comments on the NML base document. All in all an excellent effort and we are nearly there.
Thanks;
-jason
<20121104-nml-base-jz.pdf> On Oct 11, 2012, at 3:25 PM, Jeroen van der Ham <vdham@uva.nl> wrote:
Hello all,
It is with great trepidation that I present to you a draft of the "Network Markup Language Base Schema version 1".
Please have a look through it and let us/me know what you think about this document.
With the feedback we get, we will improve the document and we plan to finally start the process on this document next month.
Thanks! Jeroen.
<nml-base.pdf>_______________________________________________ nml-wg mailing list nml-wg@ogf.org https://www.ogf.org/mailman/listinfo/nml-wg
----- Jason Zurawski, Senior Research Engineer Internet2 zurawski@internet2.edu office: [+1-202-331-5354] mobile: [+1-703-981-2494] fax: [+1-202-872-4318] TIP2013, University of Hawaii Mānoa January 13 - January 17, 2013, Honolulu, HI http://events.internet2.edu/2013/tip/