Loading ...
Sorry, an error occurred while loading the content.

Re: [soapbuilders] WCF Plug-fest endpoints

Expand Messages
  • Paul Downey
    ... sounds like a good plan, though I m personally open to ideas for more lax validation such as Henry Thompson s Validate Twice idea to ignore unknowns
    Message 1 of 8 , Mar 6, 2006
    • 0 Attachment
      On 25 Feb 2006, at 22:50, Steve Loughran wrote:
      >
      > Most SOAP stacks dont validate incoming against the schema, relying on
      > the O/X mapping engine to do the right thing. I normally set up my
      > debug builds of an application to check both incoming and outgoing
      > data against the schema, because its saves recriminations later, and
      > because I prefer correct over fast.

      sounds like a good plan, though I'm personally open to ideas for more
      lax validation such as Henry Thompson's "Validate Twice" idea to
      ignore unknowns using the PSVI.

      >
      > Perhaps if we used a type language for XML that was faster to validate
      > than XSD, running with validation turned on would be the rule, rather
      > than the exception.

      That seems to be the main reason why people throw away their schema
      processor after baking it into code.

      For my money, there is room for some serious innovation around XML
      Schema
      - I'd love to describe my messages as a bunch of absolute XPaths and
      build mappers and parsers based upon that, generating XML Schema for
      those who really need such a thing.

      Paul
      --
      http://blog.whatfettle.com
    • noah_mendelsohn@us.ibm.com
      ... You may be interested to know that my group has had a paper on high performance XML Schema validation accepted at the upcoming WWW 2006 conference. The
      Message 2 of 8 , Mar 6, 2006
      • 0 Attachment
        Paul Downey wrote:

        > On 25 Feb 2006, at 22:50, Steve Loughran wrote:

        > >
        > > Perhaps if we used a type language for XML that was faster to validate
        > > than XSD, running with validation turned on would be the rule, rather
        > > than the exception.
        >
        > That seems to be the main reason why people throw away their schema
        > processor after baking it into code.

        You may be interested to know that my group has had a paper on high
        performance XML Schema validation accepted at the upcoming WWW 2006
        conference. The results suggest that schema validation can be done much
        more efficiently than is typically assumed. The conference is at the end
        of May in Edinburgh, and all papers will be posted by the conference
        organizers on the web, presumably around the time of the conference. One
        of our goals was indeed to prove that Schema validation could be deployed
        in production systems, or at least that performance need not be the
        impediment to doing so. If you're interested, look for the paper, or
        better yet come to the conference.

        (Note: the system described in the paper is a Research prototype; nothing
        should be inferred from this note or from the paper as to what IBM will or
        won't offer in its products.)

        Noah

        --------------------------------------
        Noah Mendelsohn
        IBM Corporation
        One Rogers Street
        Cambridge, MA 02142
        1-617-693-4036
        --------------------------------------








        Paul Downey <paul.downey@...>
        Sent by: soapbuilders@yahoogroups.com
        03/06/2006 06:44 AM
        Please respond to soapbuilders

        To: soapbuilders@yahoogroups.com
        cc: (bcc: Noah Mendelsohn/Cambridge/IBM)
        Subject: Re: [soapbuilders] WCF Plug-fest endpoints



        On 25 Feb 2006, at 22:50, Steve Loughran wrote:
        >
        > Most SOAP stacks dont validate incoming against the schema, relying on
        > the O/X mapping engine to do the right thing. I normally set up my
        > debug builds of an application to check both incoming and outgoing
        > data against the schema, because its saves recriminations later, and
        > because I prefer correct over fast.

        sounds like a good plan, though I'm personally open to ideas for more
        lax validation such as Henry Thompson's "Validate Twice" idea to
        ignore unknowns using the PSVI.

        >
        > Perhaps if we used a type language for XML that was faster to validate
        > than XSD, running with validation turned on would be the rule, rather
        > than the exception.

        That seems to be the main reason why people throw away their schema
        processor after baking it into code.

        For my money, there is room for some serious innovation around XML
        Schema
        - I'd love to describe my messages as a bunch of absolute XPaths and
        build mappers and parsers based upon that, generating XML Schema for
        those who really need such a thing.

        Paul
        --
        http://blog.whatfettle.com



        -----------------------------------------------------------------
        This group is a forum for builders of SOAP implementations to discuss
        implementation and interoperability issues. Please stay on-topic.
        Yahoo! Groups Links
      Your message has been successfully submitted and would be delivered to recipients shortly.