new action item needed: DFDL URN specification
Introduction: Using URLs as identifiers has caused no end of problems. E.g., in DFDL we have http://www.ogf.org/dfdl/dfdl-1.0/ as an identifier. W3C has http://www.w3.org/2001/XMLSchema and others. W3C has badly regretted establishing this convention, as they have farms of servers that do nothing but quickly return 404 errors to save network-aware applications the network-timeout delay that would otherwise occur. There are parties interested in exploiting DFDL who want DFDL schemas to NOT contain network URLs because it simply creates a concern about network access whenever a DFDL schema is inspected/used. Pre-Proposal: The new way to do this is with URNs which would look roughly like this: urn:ogf:dfdl:standard:dfdl-1.0. The whole point is that some other mechanism is used to establish correspondences between these and any resources in file systems, networks, or built-in to implementations. One such mechanism is called XML Catalog. The point is that it is a name in a managed namespace which cannot be confused with a network protocol URL. OGF is already establishing urn:ogf, and an ogf subgroup has already proposed urn:ogf:network for network resources. DFDL schemas aren't network resources so we don't want to be a substructure underneath network. Some other mechanism is used to establish correspondences between these and any resources in file systems, networks, or built-in to implementations. One such mechanism is called XML Catalog. Summary: An action item should be to specify DFDL urn, submit to OGF as a proposed namespace, and then produce errata/spec changes to specify its use. This requires a small design activity to specify a scheme for the sub-structure of the DFDL URNs (i.e., scheme for the stuff after urn:ogf:dfdl:...) where we want standard identifiers for versions of the standard, but we probably also want a few other things (e.g., I would like a space for implementations to identify themselves, i.e., an implementation-specific sub-area within our URNs.) Our existing URLs can be compatible (deprecated) practice vs the preferred URNs. -- Mike Beckerle | OGF DFDL WG Co-Chair Tel: 781-330-0412
I wanted to add to this discussion the w3c blog page that highlights the problem of the http-based naming scheme. There are many articles about this, this is just one of them. *http://www.w3.org/blog/systeam/2008/02/08/w3c_s_excessive_dtd_traffic/ * On Wed, Oct 24, 2012 at 10:35 AM, Mike Beckerle <mbeckerle.dfdl@gmail.com>wrote:
Introduction:
Using URLs as identifiers has caused no end of problems. E.g., in DFDL we have http://www.ogf.org/dfdl/dfdl-1.0/ as an identifier. W3C has http://www.w3.org/2001/XMLSchema and others.
W3C has badly regretted establishing this convention, as they have farms of servers that do nothing but quickly return 404 errors to save network-aware applications the network-timeout delay that would otherwise occur.
There are parties interested in exploiting DFDL who want DFDL schemas to NOT contain network URLs because it simply creates a concern about network access whenever a DFDL schema is inspected/used.
Pre-Proposal:
The new way to do this is with URNs which would look roughly like this: urn:ogf:dfdl:standard:dfdl-1.0. The whole point is that some other mechanism is used to establish correspondences between these and any resources in file systems, networks, or built-in to implementations. One such mechanism is called XML Catalog.
The point is that it is a name in a managed namespace which cannot be confused with a network protocol URL.
OGF is already establishing urn:ogf, and an ogf subgroup has already proposed urn:ogf:network for network resources. DFDL schemas aren't network resources so we don't want to be a substructure underneath network.
Some other mechanism is used to establish correspondences between these and any resources in file systems, networks, or built-in to implementations. One such mechanism is called XML Catalog.
Summary:
An action item should be to specify DFDL urn, submit to OGF as a proposed namespace, and then produce errata/spec changes to specify its use.
This requires a small design activity to specify a scheme for the sub-structure of the DFDL URNs (i.e., scheme for the stuff after urn:ogf:dfdl:...) where we want standard identifiers for versions of the standard, but we probably also want a few other things (e.g., I would like a space for implementations to identify themselves, i.e., an implementation-specific sub-area within our URNs.)
Our existing URLs can be compatible (deprecated) practice vs the preferred URNs.
-- Mike Beckerle | OGF DFDL WG Co-Chair Tel: 781-330-0412
-- Mike Beckerle | OGF DFDL WG Co-Chair Tel: 781-330-0412
I tend to agree that specifying urn for namespaces is better choice and URL should be used for schema locations; however convention of specifying URL for namespaces is long established some implementations use the namespace URL to return the actual schema. When I try to access the namespace URL for XML schema , it gives me reference to the document but not schema.. Suman Kalia IBM Canada Lab WMB Toolkit Architect and Development Lead Tel: 905-413-3923 T/L 313-3923 Email: kalia@ca.ibm.com For info on Message broker http://www.ibm.com/developerworks/websphere/zones/businessintegration/wmb.ht... From: Mike Beckerle <mbeckerle.dfdl@gmail.com> To: dfdl-wg@ogf.org, Date: 10/25/2012 08:18 AM Subject: Re: [DFDL-WG] new action item needed: DFDL URN specification Sent by: dfdl-wg-bounces@ogf.org I wanted to add to this discussion the w3c blog page that highlights the problem of the http-based naming scheme. There are many articles about this, this is just one of them. http://www.w3.org/blog/systeam/2008/02/08/w3c_s_excessive_dtd_traffic/ On Wed, Oct 24, 2012 at 10:35 AM, Mike Beckerle <mbeckerle.dfdl@gmail.com> wrote: Introduction: Using URLs as identifiers has caused no end of problems. E.g., in DFDL we have http://www.ogf.org/dfdl/dfdl-1.0/ as an identifier. W3C has http://www.w3.org/2001/XMLSchema and others. W3C has badly regretted establishing this convention, as they have farms of servers that do nothing but quickly return 404 errors to save network-aware applications the network-timeout delay that would otherwise occur. There are parties interested in exploiting DFDL who want DFDL schemas to NOT contain network URLs because it simply creates a concern about network access whenever a DFDL schema is inspected/used. Pre-Proposal: The new way to do this is with URNs which would look roughly like this: urn:ogf:dfdl:standard:dfdl-1.0. The whole point is that some other mechanism is used to establish correspondences between these and any resources in file systems, networks, or built-in to implementations. One such mechanism is called XML Catalog. The point is that it is a name in a managed namespace which cannot be confused with a network protocol URL. OGF is already establishing urn:ogf, and an ogf subgroup has already proposed urn:ogf:network for network resources. DFDL schemas aren't network resources so we don't want to be a substructure underneath network. Some other mechanism is used to establish correspondences between these and any resources in file systems, networks, or built-in to implementations. One such mechanism is called XML Catalog. Summary: An action item should be to specify DFDL urn, submit to OGF as a proposed namespace, and then produce errata/spec changes to specify its use. This requires a small design activity to specify a scheme for the sub-structure of the DFDL URNs (i.e., scheme for the stuff after urn:ogf:dfdl:...) where we want standard identifiers for versions of the standard, but we probably also want a few other things (e.g., I would like a space for implementations to identify themselves, i.e., an implementation-specific sub-area within our URNs.) Our existing URLs can be compatible (deprecated) practice vs the preferred URNs. -- Mike Beckerle | OGF DFDL WG Co-Chair Tel: 781-330-0412 -- Mike Beckerle | OGF DFDL WG Co-Chair Tel: 781-330-0412 -- dfdl-wg mailing list dfdl-wg@ogf.org https://www.ogf.org/mailman/listinfo/dfdl-wg
Using the OGF's URN registration document [GFD.191] as a reference, I suggest: Type URL Proposed URN common root urn:ogf:dfdl xmlns attribute http://www.ogf.org/dfdl/dfdl-1.0/ urn:ogf:dfdl:2011 source attribute http://www.ogf.org/dfdl/ urn:ogf:dfdl [GFD.191] http://www.ogf.org/documents/GFD.191.pdf Regards Steve Hanson Architect, Data Format Description Language (DFDL) Co-Chair, OGF DFDL Working Group IBM SWG, Hursley, UK smh@uk.ibm.com tel:+44-1962-815848 From: Suman Kalia <kalia@ca.ibm.com> To: Mike Beckerle <mbeckerle.dfdl@gmail.com>, Cc: dfdl-wg@ogf.org, dfdl-wg-bounces@ogf.org Date: 25/10/2012 17:15 Subject: Re: [DFDL-WG] new action item needed: DFDL URN specification Sent by: dfdl-wg-bounces@ogf.org I tend to agree that specifying urn for namespaces is better choice and URL should be used for schema locations; however convention of specifying URL for namespaces is long established some implementations use the namespace URL to return the actual schema. When I try to access the namespace URL for XML schema , it gives me reference to the document but not schema.. Suman Kalia IBM Canada Lab WMB Toolkit Architect and Development Lead Tel: 905-413-3923 T/L 313-3923 Email: kalia@ca.ibm.com For info on Message broker http://www.ibm.com/developerworks/websphere/zones/businessintegration/wmb.ht... From: Mike Beckerle <mbeckerle.dfdl@gmail.com> To: dfdl-wg@ogf.org, Date: 10/25/2012 08:18 AM Subject: Re: [DFDL-WG] new action item needed: DFDL URN specification Sent by: dfdl-wg-bounces@ogf.org I wanted to add to this discussion the w3c blog page that highlights the problem of the http-based naming scheme. There are many articles about this, this is just one of them. http://www.w3.org/blog/systeam/2008/02/08/w3c_s_excessive_dtd_traffic/ On Wed, Oct 24, 2012 at 10:35 AM, Mike Beckerle <mbeckerle.dfdl@gmail.com> wrote: Introduction: Using URLs as identifiers has caused no end of problems. E.g., in DFDL we have http://www.ogf.org/dfdl/dfdl-1.0/ as an identifier. W3C has http://www.w3.org/2001/XMLSchema and others. W3C has badly regretted establishing this convention, as they have farms of servers that do nothing but quickly return 404 errors to save network-aware applications the network-timeout delay that would otherwise occur. There are parties interested in exploiting DFDL who want DFDL schemas to NOT contain network URLs because it simply creates a concern about network access whenever a DFDL schema is inspected/used. Pre-Proposal: The new way to do this is with URNs which would look roughly like this: urn:ogf:dfdl:standard:dfdl-1.0. The whole point is that some other mechanism is used to establish correspondences between these and any resources in file systems, networks, or built-in to implementations. One such mechanism is called XML Catalog. The point is that it is a name in a managed namespace which cannot be confused with a network protocol URL. OGF is already establishing urn:ogf, and an ogf subgroup has already proposed urn:ogf:network for network resources. DFDL schemas aren't network resources so we don't want to be a substructure underneath network. Some other mechanism is used to establish correspondences between these and any resources in file systems, networks, or built-in to implementations. One such mechanism is called XML Catalog. Summary: An action item should be to specify DFDL urn, submit to OGF as a proposed namespace, and then produce errata/spec changes to specify its use. This requires a small design activity to specify a scheme for the sub-structure of the DFDL URNs (i.e., scheme for the stuff after urn:ogf:dfdl:...) where we want standard identifiers for versions of the standard, but we probably also want a few other things (e.g., I would like a space for implementations to identify themselves, i.e., an implementation-specific sub-area within our URNs.) Our existing URLs can be compatible (deprecated) practice vs the preferred URNs. -- Mike Beckerle | OGF DFDL WG Co-Chair Tel: 781-330-0412 -- Mike Beckerle | OGF DFDL WG Co-Chair Tel: 781-330-0412 -- dfdl-wg mailing list dfdl-wg@ogf.org https://www.ogf.org/mailman/listinfo/dfdl-wg -- dfdl-wg mailing list dfdl-wg@ogf.org https://www.ogf.org/mailman/listinfo/dfdl-wg Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
participants (3)
-
Mike Beckerle -
Steve Hanson -
Suman Kalia