[auscope-geosciml] RE : RE : RE : RE : GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

Boisvert, Eric Eric.Boisvert at RNCan-NRCan.gc.ca
Tue Aug 2 17:49:10 EDT 2011


>Media-types and the model are separate concerns. 
 
yes, I know that.  I was just naively expecting that reference from within a giving serialisation would be consistent.  If within GML/XML, you refer to stuff serialized in the same representation.  What I was against is mixed representations unless specifically stated.  But I see your point.  Web is loosely coupled.  But this it is also fuzzy and this is why complex application are limited (hence, the reason to have RDF as a standard to structure content - ditto for GeoSciML)
 
ok . let's me rephrase this just to check if I got this correctly. (Sorry, I'm the slow one).
 
a) When a xlink:href is dereferenced, it is not garanteed that you'll get GeoSciML even if this is what the model states and it is what you ask.
b) you must write a GIS application that ingest web reference (even if it comes from a WFS) is such a way that it should gracefully handle whatever it does not understand.  Therefore an operation is never garanteed to complete even if the provider claims "compliance" to GeoSciML.
c) you can't know in advance, 
d) UML model documents only GML instance "exactly", might be use to bind OWL or RDF or JSON (because they have a possible mapping to the UML model), but is useless for any other media-type (you can't process them based on knowledge of UML).
 
 
I remember some guy saying that having multiple options was the enemy of interoperability.  ;o)
 
Eric
 

________________________________

De: auscope-geosciml-bounces at lists.arcs.org.au de la part de Simon.Cox at csiro.au
Date: mar. 2011-08-02 17:19
À: auscope-geosciml at lists.arcs.org.au
Objet : Re: [auscope-geosciml] RE : RE : RE : GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]



Eric - 

 

Media-types and the model are separate concerns. 

 

Different clients will prefer the same content (model) in different representations (media-types). 

I agree that most GeoSciML/XML clients would prefer that dereferenced URIs would also be in GeoSciML/XML. 

Thus I would expect clients processing GeoSciML/XML would usually set the HTTP headers to Accept=application/gml+xml. 

This is a client application expressing a preference, and is independent of any information contained in the document. 

But I do not agree that this means that references in GeoSciML/XML documents may only be to sources that can supply this, and consequently the client must be configured not to break if the server sends something different back, as allowed for by the HTTP protocol. 

 

The web, including the XML part of the web, is not a closed world, and you will have to accept some dangles and media-type boundaries. 

 

Simon

 

From: auscope-geosciml-bounces at lists.arcs.org.au [mailto:auscope-geosciml-bounces at lists.arcs.org.au] On Behalf Of Boisvert, Eric
Sent: Wednesday, 3 August 2011 7:11 AM
To: auscope-geosciml at lists.arcs.org.au
Subject: [auscope-geosciml] RE : RE : RE : GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

 

>The web architecture is deliberately forgiving about hypertext links. 

 

Yes, but the possible use cases are reduced accordingly. 

 

>Ultimately it is up to the client application to know what media-types it can handle and to ask for them. 

And the client application will get this information from the model..  otherwise, why bother creating a model if anything goes ?

>This is all part of the HTTP protocol, and this is the place to solve the problem, not by messing around with the conceptual model. 

ok - you lost me. How an application is suppose to know how to process the content if it can even be garantee to get what it expects ?.  This loose architecture just brings us back to a collection of HTML pages linked together.  If you allow a vocab to be return instead of a SamplingFeature, what prevents this to happen to all other xlink:href (byReference) in the model ?

I'm just confused.

Eric

 

________________________________

De: auscope-geosciml-bounces at lists.arcs.org.au de la part de Simon.Cox at csiro.au
Date: mar. 2011-08-02 16:45
À: auscope-geosciml at lists.arcs.org.au
Objet : Re: [auscope-geosciml] RE : RE : GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

The web architecture is deliberately forgiving about hypertext links. 

I agree that content negotiation has an important role here, but that should be done in the HTTP layer, not in the data. 

Ultimately it is up to the client application to know what media-types it can handle and to ask for them. 

Then the server can say 'OK' or maybe just 'Sorry, this is all I've got'. 

This is all part of the HTTP protocol, and this is the place to solve the problem, not by messing around with the conceptual model. 

 

BTW - remember this? - https://www.seegrid.csiro.au/wiki/CGIModel/EarthNaturalSurface

This is not a new problem. 

I'll work up GML and RDF versions of it later. 

 

Simon

 

From: auscope-geosciml-bounces at lists.arcs.org.au [mailto:auscope-geosciml-bounces at lists.arcs.org.au] On Behalf Of Boisvert, Eric
Sent: Friday, 29 July 2011 7:36 PM
To: auscope-geosciml at lists.arcs.org.au
Subject: [auscope-geosciml] RE : RE : GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

 

This does not solve the original issue.  The model (originally) says that the samplingFrame is a SamplingFeature, and when you defer the URI, you end up with a SKOS or OWL representation or whatever.  Unless those thing become valid representations for a SamplingFeature ?

 

If so, should I expect similar behavior from any xlink:href ? 

such as MappedFeature/specification ? - will

 

<gsml:specification xlink:href="http://example.com/surprise" xlink:role="http://sweet.jpl.nasa.gov/2.0/astroPlanet.owl#GeologicUnit"/>

 

be also valid ?  I would expect to be given a change to negociate the content type at least.

 

Eric

 

________________________________

De: auscope-geosciml-bounces at lists.arcs.org.au de la part de Guillaume.Duclaux at csiro.au
Date: jeu. 2011-07-28 23:52
À: auscope-geosciml at lists.arcs.org.au
Objet : Re: [auscope-geosciml] RE : GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

I agree with Simon regarding the fact the conceptual model should be conceptual.
Simon proposed in his first email to make the SamplingFrame byReference. I think it is a good idea.

After some discussion, I believe GeoSciML should use the xlink:role to specify the nature of the representation of the SamplingFrame (defined byReference).

The XLink spec says that: "The value of the role (... )attribute must<http://www.w3.org/TR/xlink/#dt-must> be a URI reference as defined in [IETF RFC 2396]<http://www.w3.org/TR/xlink/#rfc2396> (...). The URI reference identifies some resource that describes the intended property. When no value is supplied, no particular role value is to be inferred. "

A schematron rule should enforce the use of xlink:role for the implementation.
if the SamplingFrame is a SpatialSamplingFeature, the xlink:role URI should be http://www.opengis.net/def/samplingFeatureType/OGC- OM/2.0/SF_SpatialSamplingFeature
if the SamplingFrame is a SamplingFrameTermCode, eg. the earth surface, the xlink:role URI should point to a relevant concept in a vocabulary, eg http://sweet.jpl.nasa.gov/2.0/astroPlanet.owl#Geosphere

Although, could someone confirm that the constraint on the MappedFeature class {self.shape contained in SamplingFrame.shape} is not violated in the case the SamplingFrame is a term, as a term (call it controlled vocab if you want) doesn't have a spatial shape directly encoded in it.

My 2 cents...

Dr Guillaume Duclaux

Research Team Leader
CSIRO Earth Science and Resource Engineering

Phone: +61 8 6436 8728  | Fax: +61 8 6436 8559  | Mobile:  +61 459 835 992

guillaume.duclaux at csiro.au<mailto:guillaume.duclaux at csiro.au> | www.csiro.au<http://www.csiro.au/> |
Address: Australian Resources Research Centre, 26 Dick Perry Avenue, Kensington WA 6151

PLEASE NOTE
The information contained in this email may be confidential or privileged. Any unauthorised use or disclosure is prohibited. If you have received this email in error, please delete it immediately and notify the sender by return email. Thank you. To the extent permitted by law, CSIRO does not represent, warrant and/or guarantee that the integrity of this communication has been maintained or that the communication is free of errors, virus, interception or interference.

Please consider the environment before printing this email.





On 29/07/2011, at 11:14 AM, <Oliver.Raymond at ga.gov.au<mailto:Oliver.Raymond at ga.gov.au>> <Oliver.Raymond at ga.gov.au<mailto:Oliver.Raymond at ga.gov.au>> wrote:

And the â*?termâ*? is not just a free text field.  It is a href to a controlled vocabulary concept, the same as all the other href-to-controlled-vocabulary-terms that we have implemented throughout the model (at Simonâ*?s suggestion).


________________________________
From: auscope-geosciml-bounces at lists.arcs.org.au<mailto:auscope-geosciml-bounces at lists.arcs.org.au> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] On Behalf Of Bruce.Simons at dpi.vic.gov.au<mailto:Bruce.Simons at dpi.vic.gov.au>
Sent: Friday, 29 July 2011 12:46 PM
To: auscope-geosciml at lists.arcs.org.au
Subject: Re: [auscope-geosciml] RE : GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

In Edinburgh Clemens argued that with just an xlink the client would expect a spatial frame of reference somewhere in the resolved xlink.
By explicitly indicating whether it is a sampling feature or a vocabulary term this expectation is removed.

----------------------------------------------------
Senior Information Systems Analyst
Prospectivity & Exploration, Earth Resources Development Division
IUGS-Commission for Geoscience Information Oceania Councillor
Level 9, 55 Collins St
PO Box 4440
Melbourne, Victoria, 3001
Australia

Ph: +61-3-9658 4502
Fax: +61-3-9658 4555
Mobile: +61 429 177155



From:        <Simon.Cox at csiro.au<mailto:Simon.Cox at csiro.au>>
To:        <auscope-geosciml at lists.arcs.org.au<mailto:auscope-geosciml at lists.arcs.org.au>>
Date:        29/07/2011 12:39 PM
Subject:        Re: [auscope-geosciml] RE :  GeoSciML ready to do for        FullMoon [SEC=UNCLASSIFIED]
Sent by:        auscope-geosciml-bounces at lists.arcs.org.au<mailto:auscope-geosciml-bounces at lists.arcs.org.au>
________________________________



We also should not be using â*?termsâ*? if they are just text fields.
Not reliable, not interoperable.
But if it is a URI, then there is no need for special treatment â*" just use the xlinks.

The encoding rule is fine, _providing_ it enforces the use of URIs rather than text for â*?termsâ*?.
The conceptual model should say what it means.

Simon


From: auscope-geosciml-bounces at lists.arcs.org.au<mailto:auscope-geosciml-bounces at lists.arcs.org.au> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] On Behalf Of Boisvert, Eric
Sent: Friday, 29 July 2011 10:29 AM
To: auscope-geosciml at lists.arcs.org.au<mailto:auscope-geosciml at lists.arcs.org.au>
Subject: [auscope-geosciml] RE : GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

Since Fullmoon will generate an XSD that will constrain us to either provide a real SamplingFeature or a term, this is the way to encode it.

do you disagree with the way XSD are generated or the conceptual model itself ?

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 at csiro.au<mailto:Simon.Cox at csiro.au>
Date: jeu. 2011-07-28 21:45
Ã*: auscope-geosciml at lists.arcs.org.au<mailto:auscope-geosciml at lists.arcs.org.au>
Objet : Re: [auscope-geosciml] GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]
Tags and encoding rule are for implementation detail.
Classes are for the conceptual model.
The class diagram should show the conceptual logic.
The last time I looked term-code was not a geologic concept.

I strongly disagree with the logic and decision here.

Simon

From: auscope-geosciml-bounces at lists.arcs.org.au<mailto: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<mailto:Oliver.Raymond at ga.gov.au>
Sent: Friday, 29 July 2011 9:44 AM
To: auscope-geosciml at lists.arcs.org.au
Subject: Re: [auscope-geosciml] GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

Conceptually, I agree.  But implementation-wise, a value from a vocabulary (ie, â*?EarthSurfaceâ*?) is not a sams:SpatialSamplingFeature.

GeoSciML is more than just a conceptual model â*" it is very much a logic implementation model too.  The UML model to XML schema is pretty much a 1-to-1 match, and we have here specifically stated how we want to implement the conceptual model.



Cheers,
Ollie



________________________________

From: auscope-geosciml-bounces at lists.arcs.org.au<mailto:auscope-geosciml-bounces at lists.arcs.org.au> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] On Behalf Of Simon.Cox at csiro.au<mailto:Simon.Cox at csiro.au>
Sent: Friday, 29 July 2011 11:28 AM
To: auscope-geosciml at lists.arcs.org.au
Subject: Re: [auscope-geosciml] GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

Yeah â*" but it is a vocabulary-term that denotes a sampling-feature.
So semantically it is a sampling feature.
It is just a different representation (a rather limited one).

This is quite important. The model should show â*?whatâ*? in terms of the model, not in terms of some implementation artefact like a vocabulary term.
â*?Vocabulary termâ*? has meaning in lexicography.
â*?SpatialSamplingFeatureâ*? has meaning in mapping.
Is this a model of a vocabulary, or of geological information?

Simon

From: auscope-geosciml-bounces at lists.arcs.org.au<mailto: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<mailto:Oliver.Raymond at ga.gov.au>
Sent: Friday, 29 July 2011 9:21 AM
To: auscope-geosciml at lists.arcs.org.au
Subject: Re: [auscope-geosciml] GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

Hi Simon,

The model was byReference to a SpatialSamplingFeature, but in practice most instances were using a byReference to a vocabulary term, like â*?EarthSurfaceâ*?.  Considerable discussion at Edinburgh around what should be expected at the end of an xlink:href decided that if we model the href as a sampling feature, then the link should resolve to a sampling feature, not a vocabulary term.  The model now explicitly states that the link can be either to a sampling feature or to a vocabulary term.

Cheers,
Ollie

_______________________________________________________________________

Ollie Raymond

Project Leader
National Geological Maps and Data Standards Project<http://www.ga.gov.au/minerals/projects/current-projects/geological-maps-standards.html>
Geoscience Australia

Interoperability Working Group<https://www.seegrid.csiro.au/wiki/bin/view/CGIModel/InteroperabilityWG>
IUGS Commission for the Management and Application of Geoscience Information

Address: GPO Box 378, Canberra, ACT, 2601, Australia | ABN: 80 091 799 039
Ph: +61 2 62499575  |  Fax: +61 2 62479992  |  Email: oliver.raymond at ga.gov.au<mailto:oliver.raymond at ga.gov.au>  |  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 ---


________________________________

From: auscope-geosciml-bounces at lists.arcs.org.au<mailto:auscope-geosciml-bounces at lists.arcs.org.au> [mailto:auscope-geosciml-bounces at lists.arcs.org.au] On Behalf Of Simon.Cox at csiro.au<mailto:Simon.Cox at csiro.au>
Sent: Friday, 29 July 2011 11:04 AM
To: auscope-geosciml at lists.arcs.org.au; Francois.Letourneau at RNCan-NRCan.gc.ca
Subject: Re: [auscope-geosciml] GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

â*?For samplingFrame, we have created a union class which explicitly states that the samplingFrame association can be either a vocabulary term or a real SpatialSamplingFeature?<https://www.seegrid.csiro.au/wiki/bin/edit/CGIModel/SpatialSamplingFeature?topicparent=CGIModel.EdinburghModelAgendaAndNotes>. (DONE - OllieRaymond?<https://www.seegrid.csiro.au/wiki/bin/edit/CGIModel/OllieRaymond?topicparent=CGIModel.EdinburghModelAgendaAndNotes>, 22-7-2011) [Union class documentation on the TWiki at UmL2GMLAS<https://www.seegrid.csiro.au/wiki/bin/view/AppSchemas/UmL2GMLAS>]. Needs to be tested in Testbed instance docs.â*?

Why not just make it â*?byReferenceâ*?.
Then the semantics are preserved even if there is not an instance in the correct format.

From: auscope-geosciml-bounces at lists.arcs.org.au<mailto: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<mailto:Oliver.Raymond at ga.gov.au>
Sent: Friday, 29 July 2011 8:49 AM
To: Francois.Letourneau at RNCan-NRCan.gc.ca
Cc: auscope-geosciml at lists.arcs.org.au
Subject: [auscope-geosciml] GeoSciML ready to do for FullMoon [SEC=UNCLASSIFIED]

Hi Francois and all,

The trunk GeoSciML model and classmaps are now ready for FullMoon processing.  (Thanks to Gilly for helping with the classmaps)

This will be the final release candidate schemas for GeoSciML v3.  All the edits from Edinburgh have been done and we are on track for our release schedule (see https://www.seegrid.csiro.au/wiki/CGIModel/EdinburghModelAgendaAndNotes).

Cheers,
Ollie

_______________________________________________________________________

Ollie Raymond

Project Leader
National Geological Maps and Data Standards Project<http://www.ga.gov.au/minerals/projects/current-projects/geological-maps-standards.html>
Geoscience Australia

Interoperability Working Group<https://www.seegrid.csiro.au/wiki/bin/view/CGIModel/InteroperabilityWG>
IUGS Commission for the Management and Application of Geoscience Information

Address: GPO Box 378, Canberra, ACT, 2601, Australia | ABN: 80 091 799 039
Ph: +61 2 62499575  |  Fax: +61 2 62479992  |  Email: oliver.raymond at ga.gov.au<mailto:oliver.raymond at ga.gov.au>  |  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 ---
 _______________________________________________
auscope-geosciml mailing list
auscope-geosciml at lists.arcs.org.au<mailto:auscope-geosciml at lists.arcs.org.au>
http://lists.arcs.org.au/cgi-bin/mailman/listinfo/auscope-geosciml

ã"?â*OE[1]È"{ch'â*OESLSNË"9CCâ??í?OEí"YGñ?µ±N5;"Í-8ÔfiÇ*&Nzf?ª|Ö?gÉ?É...'wè®oeí?«bÚ?ʧ~'^Ø?ez*kzjw(*â??ã"?ã"?â*OE[1]Ú©jh~+luz趧â*OEuZ??(kÆ-yß?8Ô?8ÔfiÇ*&«aë-?ê«...>®«-zw(ǧí?§(*â??h·SOñ?µ¿Ï?SNÈ"{aN57Ú±à«?H+-êi"'&â^«razEUR¨r+jwkzj/zǬSOñ?µ¿Ï?SMí¸"í¸...âª?"Í-*.?-çi...í?·í¶?zf)?®+W騶'ò¶~¬zw^zEUR«é?...W^ëi...?«l2?oejw]zË«&É...)ë?¢æ«ºz-jë¡¢yEUR¨Ç?i'ê«-é"¢{az)ߢ*'ríª©?¶)í?«â*OESOó??£S}â*OE[1]Ê¥4óY´",??^jÇ?{"uê-?ç§?*èi§)í...¸-+â*OESOó??£S}â*OE[1]È"{oŸԧnÊ¿
_______________________________________________
auscope-geosciml mailing list
auscope-geosciml at lists.arcs.org.au<mailto:auscope-geosciml at lists.arcs.org.au>
http://lists.arcs.org.au/cgi-bin/mailman/listinfo/auscope-geosciml

_______________________________________________
auscope-geosciml mailing list
auscope-geosciml at lists.arcs.org.au
http://lists.arcs.org.au/cgi-bin/mailman/listinfo/auscope-geosciml

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/ms-tnef
Size: 27435 bytes
Desc: not available
URL: <http://lists.opengeospatial.org/pipermail/geosciml/attachments/20110802/8961282e/attachment.bin>


More information about the GeoSciML mailing list