Minutes of IPET-MDI-2

[Steve Foreman]

Please let me have comments on these draft minutes of IPET-MDI-2 by 1st June.

Thanks

Steve

[TOYODA Eizi]

Dear Steve,

 

Thanks a lot for great work.  I think the minutes would better reflect the discussion there with attached changes.

 

 - Example of major difference between GRIB and TC211 is handling of time(s).  It's clear and worth noting to delineate in what domain an approach based on common data-model would be beneficial.

 - I tried to reproduce what Jeremy explained on versioning.  I hope it's not a problem.

 

Best Regards,

Eizi

[TOYODA Eizi]

So sorry I forgot to attach the file (again).

[Siegfried Fechner]

Dear all,

I discussed the results of our meeting with Heinrich Knottenberg and send a modified version as attachment (only some typos and one proposal/comment - topic 3.21)

Best regards
Siegfried

[Steve Foreman]

Eizi

I am a little confused by your text for the "middle" level of versioning:

b: Extension to structure (in the unlikely case that we give up to maintain WMO Core Profile to be “type-1 profile”) or addition of business rule against which compliance tests are run.  This update is caused by guidances raised to compliance tests.  A metadata record once compliant to older profile may no longer be compliant to newer profile, due to new test.  Any metadata record compliant to newer profile are compliant to older profile.  That means there is no need for parallel operation at WIS centres.

A WIS centre that is running the validation against the higher level of "b" will find that older metadata is not compliant - but one running the lower version will have no problems.

So, are you saying that the GISCs and other centres should not update the validation/conformance tests at every change of "b" version?

Or are you saying that the software should detect the version of the metadata and not bother running the validation for items that were not a conformance issue in the earlier "b" version of the metadata?

Steve

On Thursday, May 31, 2012 2:44:49 PM UTC+2, TOYODA Eizi wrote:



[TOYODA Eizi]

Dear Steve,

 

First of all, please note that two "validations" for XSD and Schematron has different nature.  XSD is enforced before inter-GISC exchange, but Schematron is not: it's only evaluation after exchange.

 

First meeting of IPET-MDI agreed that WIS data catalogue must accept metadata records that is compliant only to ISO 19139 XSD and not to WMO profile.  That was deemed essential to have good coverage over all WMO activities, which may include organizations under various requirements.  Please find Recommendation 19 of http://www.wmo.int/pages/prog/www/WDM/IPET-MDI-I/report/Report_IPETMDI_I_20100503.htm for reference.  It is okay to change that, but we must be clearly explicit to rule out all possibilities we have once granted to the entire world.

 

IPET-MDI-I also agreed that GISCs do XSD validation and reject records non-compliant to ISO 19139.   Personally it's okay for me that single catalogue accomodate multiple schemas or namespaces, but I also respect the sense of our German colleagues that such freedom sounds like insecure garbage box.

 

So, what will happen when we increment middle level are:

 

(1) Initially all GISCs do the same XSD and schematron validation.

  Suppose all metadata records are compliant to both.

 

(2) IPET-MDI releasesnew schematron validation with more restriction.

 

(3) Some advanced originating centres (I mean NC/DCPC) may follow that, before being warned by their GISC. 

 

(4) GISCs updates schematron validation.  That will cause bunch of alert of noncompliance.  But it's okay since that cannot reject records from catalogue.

 

(5) Non-advanced originating centres receive alert of noncompliance from their GISCs.  It would be nicer that GISC warns before promotion of guidance to rules, but it's another story.

 

(6) When all old records are superseded by originating centre, the migration is done.

 

Please see GISCs can update schematron, and validation software doesn't need "if-branching" for version at any time.

Some may feel unhappy seeing WMO profile is not enforced, but I still believe this is the only workable way in practical timescale.

[Steve Foreman]

Hi Everyone,

I apologise that this has taken so long, but here is what I hope is the last draft of the IPET-MDI minutes.

I have tried to include all the comments I received on the last draft.

Steve

[Steve Foreman]

Hi everyone, I will take the silence to mean that the report has been accepted.

I will take the "draft" off and post it to the WMO web site.

Thanks

Steve

[Jeremy Tandy]

Steve. The minutes are an accurate reflection of our discussion. I spotted a few formatting inconsistencies but nothing requiring an amendment to the draft.

Thanks for your efforts. Jeremy