Michael Bischoff wrote:
Hello all,
Are you sure Nina? Afaik a WSDL can be defined with a single method that takes a document and define that the document needs to have a NMWG schema?
Yes indeed. Sorry for not being clear, I wanted to point domain specific schema is necessary anyway. Best regards, Nina
Alike: http://www.w3.org/2001/04/wsws-proceedings/uche/wsdl.html
Though it does require a xsd or other supported schema type (eg not relaxng). Now relaxng can be transformed but the transformation isn't formal/unambiguous/two-way, or differently put if you transform from relaxng to xsd and back you are not guaranteed to end up with the same relaxng document. This is because some relaxng concepts can't be mapped completely into xsd. Now afaik we don't make use of such concepts now but it could create problems in the future.
As I frequently hear that there is no objection to particular preference for xsd/relaxng from a theoretical standpoint (I know WG prefers relaxng) Why not switch to xsd for practial reasons?
Considering I haven't seen the detailed report yet, please review the following comments with that in mind;
For all that is wrong about perfsonar(and there are things wrong), certain key area's seem to be overlooked.
Perfsonar is about sharing data to solve problems around services that stretch across multiple administrative domains as opposed to making the data available just within your own domain. For example AA, as far as Authentication is concerned with NETCONF; it is left to the transport layer which basically means that it's up to administrator/deployers to sort it out. Now I'm sure admins/deployers are up to the task but still it would require considerable effort as well as additional specification/conventions. As perfsonar is actually focused on making network measurements avail across multiple administrative domains it actually does solve/handle this. Even currently all it takes is to point a service to an AA-server. As pointed out on the perfsonar-dev list even suppling an AA-server could easily be done automatically.
Perfsonar was not designed to replace current measurement systems the idea was to complement them. As such it shouldn't be competing with netconf; however as there are propriatary solutions for 'local' use. Netconf might find itself caught in the middle.
Regards,
Michael
Nina Jeliazkova wrote:
Roman Lapacz wrote:
Freek Dijkstra wrote:
Hi all,
My apologies for the (very) late reply.
I have a few generic comments and questions on "Extensible Protocol for NMC" for now and will follow up with some more detailed remarks later. (I planned to have the detailed remarks now, but was not able to finish reading all I wanted).
My main point of criticism is that I do not (yet) understand why this protocol was developed; as a naive reader it seems to highly overlap with what has been done with WSDL. Unfortunately, I'm not an expert on either, so I might miss the point. WSDL was developed as a generic request-response protocol framework, and so is "Extensible Protocol for NMC".
NMC schema transformed from Relax NG into XSD format could be used in WSDL file to describe all messages. I didn't try it but basing on my (limited) knowledge of WSDL it's possible. So I don't see the conflict here.
I understand that WSDL was mentioned here as an example for general observation that maybe some existing standard could be used instead of NMC. The same comment was raised in Geant3 by people who were not involved earlier in pS work. The investigation of Netconf in pS is a try to look at this issue and find answers.
WSDL alone is not sufficient for defining working web services, one needs to define namespaces and XML schema, specific for the domain (NMC schema for network monitoring in our case). On the other hand, WSDL can still be used in implementations , together with NMC schema, to allow automatic generation of client classes, but this opportunity has not been explored so far.
Best regards, Nina
Roman
_______________________________________________ Nmc-wg mailing list Nmc-wg@ogf.org http://www.ogf.org/mailman/listinfo/nmc-wg
------------------------------------------------------------------------
_______________________________________________ Nmc-wg mailing list Nmc-wg@ogf.org http://www.ogf.org/mailman/listinfo/nmc-wg