[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [nm-wg] Relax NG Request Schema

Hey all,

Some quick comments from my end:

- First of all, looks great, I think the relax-ng schemas are easier to read for everybody, and certainly are easier to modify. I haven't run any of it through xml-spy, because I just use the XSD to validate my little sample documents.

- Date/time will probably need to be more complex in the request than the response schema, since they're not exactly the same thing -- in the request, you're providing a query and/or (I think) implying things about how to do active tests, in the response you're just saying when the thing actually occurred.

- The response (report) requirements doc DOES imply changes in the report schema and, to a lesser extent, the request schema.

- Dan

Paul D Mealor wrote:
On Fri, 27 Feb 2004, Eric Boyd wrote:
Paul,

Looks very interesting. I've just skimmed it, but had a couple of quick questions.

1) Does this schema still meet the requirements for the request schema, or have you had to alter/not meet any of the requirements?

I think it still meets all of the requirements. I will make sure however
(especially as I've just noticed I haven't put in a startTime and endTime
version in the time section; and I'm not yet supporting all forms of
date/time that we had before).
2) Does this schema suggest any new requirements of the request or response schema that we hadn't thought of before?

I do not believe so.
3) Does this schema require any changes of Dan's relaxed version of the response schema?

Probably... if both schemas are to use the same parameters section, then I
think something must be added, as there are extra attributes in the
request schema (required/range/report). It may be easier than I think to
make this work with RelaxNG, however.

The support for different types of date/time may also require changes to the report schema, if we actually decide to maintain that (which we may).

Sorry to be so brief. Gotta go :)

Paul


--Eric

At 06:46 PM 2/27/2004 +0000, Paul D Mealor wrote:


Hi all,

The request schema can be found relaxing here:

http://www/~pdm/nmwg/

I've tried to make it as similar to the newest report schema as I can, and
it does look quite similar.

The most notable changes in the request schema are:

* time information has been moved within the methodology, a la response
 schema;

* tool information is also in the methodology... I'm not sure if it's
 entirely sensible there, though - you can only specify one tool to use
 if you want to specify a tool, and there isn't yet a way of marking it
 optional/required etc; also the command lines are specified here;

* the parameters now have the same names as in the response schema with
 the extra "required", "range" and "report" attributes;

* an allowNewMeasurements attribute and the maxResults attribute have been
 added to the "Request" element.

No examples yet, and no XMLSpy "documentation", because XMLSpy crashes
when I try to load Trang-generated XSDs. Did yours work straight off, Dan?

Both schemas need support for nodes that have only a name and no IP (I
think). Otherwise, nm_subject are all but identical (the request schema
has a lot taken out, you might notice)

Both nm_tools are very close, except for the added command-line options.
Possibly boolean required and report attributes need to be added, however.

nm_timeinfo (represents a time period) and nm_statistics (represents
requests for particular statistics) are unique to the request schema, and
nm_time and nm_values are unique to the report schema.

I think the next step is to try to use the same nm_parameters files...

Paul
--
Paul Mealor
 --+->    pdm@hep.ucl.ac.uk    <-+--
 <-+-- www.hep.ucl.ac.uk/~pdm/ --+->
   |     +44 (0)20 7679 3044   <-+--
 --+->         x33044            |

Eric L. Boyd
eboyd@internet2.edu
734-352-7032 (work)
734-834-6708 (cell)

Office 230
Internet2
3025 Boardwalk
Ann Arbor, MI 48108