search

ISO-TC211 / XML

XML schema, transforms, schematron rules, and examples for ISO TC211 Metadata Standards
46 stars 26 forks source link

Incorrect link to "ISO 19136:2005 GML 3.2 XML schemas" #209

Open cportele opened 5 years ago

cportele commented 5 years ago

The XMG section on the homepage https://www.isotc211.org/ has a link to "ISO 19136:2005 GML 3.2 XML schemas". This has several problems:

To avoid confusion, please remove the link and the schema from the isotc211.org site.

ronaldtse commented 5 years ago

@cportele thanks for raising this issue. I think the link was supposed to say ISO 19136:2007 (instead of 2005).

My understanding from https://github.com/ISO-TC211/www.isotc211.org/issues/15 was that the schemas under /2005/ must not be removed; we've had people complain that their systems fail because these schemas were rendered unavailable during the time of migration.

cportele commented 5 years ago

@ronaldtse - thanks for the response.

Changing 19136:2005 to 19136:2007 would still be misleading and incorrect, see the two other issues. The link leads to a version of the old 19139 schemas (which also happens to import some draft of GML, but is not about GML). Please note that this 19139 schema is also not the 19139 schema that was published with ISO/TS 19139:2007.

If the schemas have to stay there, change the link title on the landing page to remove the impression that these have anything to with GML / ISO 19136.

ronaldtse commented 5 years ago

@cportele our reason for having a link on the front page is because the /2005/ path would otherwise be invisible. Do you have any suggestion on what the name of the link could be?

Perhaps @smrgeoinfo @dr-shorthair @PeterParslow @jetgeo can shed light onto this?

cportele commented 5 years ago

I think the correct title would be something like "Unofficial draft schemas of ISO/TS 19139:2007".

Note that https://schemas.isotc211.org/schemas/ does not seem to have the "official schemas of ISO/TS 19139:2007", which is interesting.

Maybe it would be better not to have a (prominent) link to something that is/was not meant to be used and is only there, because some are still using them?

smrgeoinfo commented 5 years ago

there is a gml schema at https://www.isotc211.org/2005/gml/, linked from the https://www.isotc211.org/2005/ page, along with the 2005 versions of implementations of ISO19115, 19110 (gfc), 19135(grg), and the base types in gco, gss, gts, all implemented according to 19139 (at least 19115 is...).

I agree the link shouldn't be very prominent and should have a label like 'Superseded implementations of spatial data XML schema', and a note along the lines of 'These xml schema have been superseded by schema at https://schemas.isotc211.org/schemas/ but are maintained to support legacy applications. '

cportele commented 5 years ago

The point is that it really isn't an official GML schema. It is different from any of the GML schema versions that are approved and published (http://schemas.opengis.net/gml/). Someone just took a draft version, imported it in an (also unpublished) 19139 version and published that draft GML schema as part of it.

In that sense "superseded" would be a very friendly characterization of the status of those schemas. I personally would still hide the link, but I would leave it to the XMG as long as GML/19136/etc are not mentioned or implied in the link title.

ronaldtse commented 5 years ago

I scoured through my email and found these instances:

From @MatsAhlin :

https://www.isotc211.org/2005/resources/ referenced in ISO/TS 19157-2

From @jetgeo :

One more: The directory http://www.isotc211.org/2005/ and the content of it must exist on the new server as well, of historical reasons. Numerous users are using this for the old ISO 19139, which was the XML implementation of the old ISO 19115 standard.

From Yves Amsler:

The link to http://www.isotc211.org/2005/gmd is no longer valid. This is in fact the same for all XML namespaces and prefixes used in the document inspire-tg-metadata-iso19139-2.0.1 @jetgeo: Yes, I think this link is supposed to valid, as it is the namespace for an older and much used XML implementation of ISO 19115 (defined in ISO 19139). I believe I have mentioned it before 😊.

Yves Amsler:

I was referring especially to the official EC document "Technical Guidance for the implementation of INSPIRE dataset and service metadata based on ISO/TS 19139:2007". If you click on the hyperlink before and download the pdf document mentioned on the webpage and then open it and go to page XX1, you find a table as follows:

XML namespaces and prefixes used in this document XML element prefixes are used in this document to refer to the namespaces as follows:

prefix Namespace URI
gmd http://www.isotc211.org/2005/gmd
gco http://www.isotc211.org/2005/gco
gmx http://www.isotc211.org/2005/gmx
srv http://www.isotc211.org/2005/srv
gml http://www.opengis.net/gml/3.2 (for GML 3.2.1) or http://www.opengis.net/gml (for GML 3.2.0)
xsi http://www.w3.org/2001/XMLSchema-instance
xlink http://www.w3.org/1999/xlink

From @MatsAhlin

I followed the conversation with Ted and Knut on the 19139 and the urls: http://www.isotc211.org/2005/gco
http://www.isotc211.org/2005/gmd
http://www.isotc211.org/2005/gmx
http://www.isotc211.org/2005/gss
http://www.isotc211.org/2005/gsr
http://www.isotc211.org/2005/gts

ISO 19110 has the link: http://www.isotc211.org/2005/gfc

ISO 19139-2 has the link: http://www.isotc211.org/2005/gmi It feels related to the 19110 & 19139 mentioned above, so whatever we say we do with them would be for this one as well.

ISO 19145 has the link: http://www.isotc211.org/2005/gplr

From Margarita Arcos:

The namespaces urls are giving us error, we use them at the beginning of our metadata. Have you change the urls? Is it something temporal? xmlns:gco="http://www.isotc211.org/2005/gco" xmlns:gmd="http://www.isotc211.org/2005/gmd" xmlns:gmx="http://www.isotc211.org/2005/gmx" ...