[Auscope-geosciml] Alignment of UML and XML schema

Stephen M Richard smrtucson at gmail.com
Wed Oct 14 11:29:26 EDT 2009


I suggest we make a practice of putting notes in the UML models where 
there are mapping schemes that result in discrepancies between the UML 
model and the XML schema that implement the model. This will alert users 
not to expect the XML to exactly match the UML...

steve

Simon Cox wrote:
> Ollie - you are reading very grumpy about this.
> What do you propose instead?
>  
> In this case there are indeed various 'anomalies' caused by the fact 
> that the development of the ISO UML for 19103, 19115, and 19107 in 
> particular were not synchronized with GML or indeed with the web in 
> general. 
> And there are other even more appropriate technologeis that have come 
> along subsequently, like SKOS.  
> I'm afraid various people and groups were involved, bringing various 
> requirements and experience to the table, and it is quite a triumph of 
> cooperation that it is as aligned as well as it is.
> Future specs will have less anomalies because we understand the system 
> as a whole better now.
>  
> I've tried to explain a way to understand it.
> If that is not good enough for you, then you are quite welcome to go 
> off and reinvent the whole thing so that it is consistent.
> Eric may 'have it', but I'm afraid that the real world includes 
> compromises.
> So you will just have to learn to live with it. 
>  
> Thats the last I'm going to say on the topic.
>  
> Simon
>  
>
> --------------------------------------------------------
> *Simon Cox
> *
> European Commission, Joint Research Centre
> Institute for Environment and Sustainability
> Spatial Data Infrastructures Unit, TP 262
> Via E. Fermi, 2749, I-21027 Ispra (VA), Italy
> Tel: +39 0332 78 3652
> Fax: +39 0332 78 6325
> mailto:simon.cox at jrc.ec.europa.eu
> http://ies.jrc.ec.europa.eu/simon-cox
>
> SDI Unit: http://sdi.jrc.ec.europa.eu/
> IES Institute: http://ies.jrc.ec.europa.eu/
> JRC: http://www.jrc.ec.europa.eu/
>
> --------------------------------------------------------
>
>  
>
> ------------------------------------------------------------------------
> *From:* auscope-geosciml-bounces at lists.arcs.org.au 
> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] *On Behalf Of 
> *Oliver.Raymond at ga.gov.au
> *Sent:* Wednesday, 14 October 2009 00:37
> *To:* auscope-geosciml at lists.arcs.org.au
> *Cc:* John.Hockaday at ga.gov.au
> *Subject:* Re: [Auscope-geosciml] An issue withISODataQualityand 
> UnitsOfMeasure schema [SEC=UNCLASSIFIED]
>
> Eric has it.  Now we realise that ISO schemas do not necessarily 
> reflect the ISO UML.  But unless we construct instance docs as we 
> model, or carry the ISO documentation with us, we are unlikely to 
> realise that our model which uses ISO elements will not be delivered 
> as we expect.  In the case of UnitOfMeasure, the difference between 
> UML and schema is not insignificant.
>
>  
>
> Data modelling teams like us construct models using UML, not xsd, and 
> until now (as Eric points out) we have expected the xsd to replicate 
> the UML (eg; GeoSciML UML goes into FullMoon, isomorphic GeoSciML xsd 
> comes out the other end).  I suppose now that we are going to use 
> SKOS/RDF mappings for encoding GSML vocabularies, the assumption of 
> UML/XSD isomorphism no longer holds for GeoSciML either.
>
>  
>
> Cheers,
>
> Ollie
>
>  
>
> /------------------------------------------------------------------------------------------------/
>
> /Ollie Raymond
> National Advice,  Maps and Standards Project/
>
> */Geoscience /**/Australia/*
>
> *//* 
>
> *Address:* GPO Box 378, Canberra, ACT, 2601, Australia *|* *ABN:* 80 
> 091 799 039
>
> *Ph:* (02) 62499575 *|* *Fax:* (02) 62499992 *|* *Email: 
> *Oliver.Raymond at ga.gov.au
>
> *Web:*  
> http://www.ga.gov.au/minerals/research/national/nat_maps/nat_geol_maps.jsp
>
> *Google Map 
> <http://maps.google.com/maps?f=q&hl=en&geocode=&q=canberra+australia&ie=UTF8&ll=-35.344028,149.158362&spn=0.007684,0.016404&t=h&z=17&iwloc=addr&om=1> 
> *
>
>  
>
> -- This message was created with 100% recycled electrons --
>
>  
>
>  
>
> -----Original Message-----
> *From:* auscope-geosciml-bounces at lists.arcs.org.au 
> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] *On Behalf Of 
> *Simon Cox
> *Sent:* Wednesday, 14 October 2009 2:03 AM
> *To:* auscope-geosciml at lists.arcs.org.au
> *Cc:* Hockaday John
> *Subject:* Re: [Auscope-geosciml] An issue withISODataQualityand 
> UnitsOfMeasure schema [SEC=UNCLASSIFIED]
>
>  
>
> >90% of the time the assumption is correct.
>
> But in some places the UML is clumsier than an available 
> implementation (e.g. compare Measure with gml:MeasureType, GenericName 
> with gml:CodeType, CharacterString with xs:string) or there is another 
> reason why the implementation does not follow the encoding rule (e.g. 
> most of the basic geometry classes are implemented by something that 
> is slightly different in GML). In these cases the encoding rule is 
> overridden by a mapping.
>
>  
>
> It is mostly in basic datatypes where you wouldn't notice or care.
>
> I'm a little unclear why Ollie does in this case.
>
>  
>
> --------------------------------------------------------
> *Simon Cox
> *
> European Commission, Joint Research Centre
> Institute for Environment and Sustainability
> Spatial Data Infrastructures Unit, TP 262
> Via E. Fermi, 2749, I-21027 Ispra (VA), Italy
> Tel: +39 0332 78 3652
> Fax: +39 0332 78 6325
> mailto:simon.cox at jrc.ec.europa.eu
> http://ies.jrc.ec.europa.eu/simon-cox
>
> SDI Unit: http://sdi.jrc.ec.europa.eu/
> IES Institute: http://ies.jrc.ec.europa.eu/
> JRC: http://www.jrc.ec.europa.eu/
>
> --------------------------------------------------------
>
>  
>
>  
>
> ------------------------------------------------------------------------
>
> *From:* auscope-geosciml-bounces at lists.arcs.org.au 
> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] *On Behalf Of 
> *Boisvert, Eric
> *Sent:* Tuesday, 13 October 2009 13:27
> *To:* auscope-geosciml at lists.arcs.org.au
> *Cc:* John.Hockaday at ga.gov.au
> *Subject:* Re: [Auscope-geosciml] An issue withISODataQualityand 
> UnitsOfMeasure schema [SEC=UNCLASSIFIED]
>
> I can understand Ollie's comment because we have been working for some 
> times now with the assumption that the UML was the only representation 
> of the model, XSD was just a Fullmoon click away.  So our model is 
> isomorphic (is it ?) and as Ollie, I expected other UML of the ISO/OGC 
> familly would also be.  
>
>  
>
> Is your comment saying "//those other UML are not ismorphic and it is 
> sad//" or "//UML models, including GeoSciML, are not required to be 
> isomorphic//"..  This has an impact because the trend is to include 
> other models (Metadata, SKOS, O&M, Sampling) and we do this through UML.
>
>  
>
> Eric
>
>  
>
> ------------------------------------------------------------------------
>
> *De :* auscope-geosciml-bounces at lists.arcs.org.au 
> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] *De la part de* 
> Simon Cox
> *Envoyé :* 13 octobre 2009 04:15
> *À :* auscope-geosciml at lists.arcs.org.au
> *Cc :* John.Hockaday at ga.gov.au
> *Objet :* Re: [Auscope-geosciml] An issue with ISODataQualityand 
> UnitsOfMeasure schema [SEC=UNCLASSIFIED]
>
> The UML model is a conceptual schema, not a graphical view of the XML 
> schema.
>
> The use of a regular encoding rule makes them generally isomorphic, 
> but not fully.
>
> You just have to trust the encoder, which uses mappings as well as rules.
>
>  
>
> If you want a view of the schema, then use an XML tool, not UML.
>
>  
>
> --------------------------------------------------------
> *Simon Cox
> *
> European Commission, Joint Research Centre
> Institute for Environment and Sustainability
> Spatial Data Infrastructures Unit, TP 262
> Via E. Fermi, 2749, I-21027 Ispra (VA), Italy
> Tel: +39 0332 78 3652
> Fax: +39 0332 78 6325
> mailto:simon.cox at jrc.ec.europa.eu
> http://ies.jrc.ec.europa.eu/simon-cox
>
> SDI Unit: http://sdi.jrc.ec.europa.eu/
> IES Institute: http://ies.jrc.ec.europa.eu/
> JRC: http://www.jrc.ec.europa.eu/
>
> --------------------------------------------------------
>
>  
>
>  
>
> ------------------------------------------------------------------------
>
> *From:* auscope-geosciml-bounces at lists.arcs.org.au 
> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] *On Behalf Of 
> *Oliver.Raymond at ga.gov.au
> *Sent:* Monday, 12 October 2009 23:41
> *To:* auscope-geosciml at lists.arcs.org.au
> *Cc:* John.Hockaday at ga.gov.au
> *Subject:* Re: [Auscope-geosciml] An issue with ISO DataQualityand 
> UnitsOfMeasure schema [SEC=UNCLASSIFIED]
>
> Hmm, it just makes it difficult for us poor souls who are trying to 
> use the ISO UML to model our data when it doesn't accurately reflect 
> what will be delivered by the xml schema....
>
>  
>
>  
>
>  
>
> ------------------------------------------------------------------------
>
> *From:* auscope-geosciml-bounces at lists.arcs.org.au 
> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] *On Behalf Of 
> *Simon Cox
> *Sent:* Monday, 12 October 2009 7:50 PM
> *To:* auscope-geosciml at lists.arcs.org.au
> *Cc:* Hockaday John
> *Subject:* Re: [Auscope-geosciml] An issue with ISO DataQuality and 
> UnitsOfMeasure schema [SEC=UNCLASSIFIED]
>
> Ollie - Its all OK.
>
>  
>
> You have to distinguish between the abstract model and the 
> implementation.
>
>  
>
> Sometimes they are related by a rule (e.g. the 19136 Annex E, or 19139 
> encoding rule), but sometimes they are just related by a 'mapping', 
> particularly if there is a pre-existing implementation that captures 
> the semantics.
>
> Thats what has happened in this case.
>
>  
>
> It is also explained in Annex D of ISO 19136 (GML) where the GML 
> implementations of classes from ISO 19103 are explained.
>
>  
>
> Simon
>
>  
>
> --------------------------------------------------------
> *Simon Cox
> *
>
>

-- 
Stephen M. Richard
Section Chief, Geoinformatics
Arizona Geological Survey
416 W. Congress St., #100
Tucson, Arizona, 85701 USA

Phone: 
Office: (520) 209-4127
Reception: (520) 770-3500 
FAX: (520) 770-3505

email: steve.richard at azgs.az.gov

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opengeospatial.org/pipermail/geosciml/attachments/20091014/386250a6/attachment.htm>


More information about the GeoSciML mailing list