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

Re: Re: Detect native SVG support

Expand Messages
  • Jim Ley
    Jonathan Watt wrote in message news:f21dbbd705072817234af897e9@mail.gmail.com... ... BUG! It should not return true for either, as it has
    Message 1 of 23 , Aug 2, 2005
    • 0 Attachment
      "Jonathan Watt" <wattie@...> wrote in message
      news:f21dbbd705072817234af897e9@......
      > Hi Mark,
      >
      > You're quite right of course. This would be the *real* way to do it,
      > and Mozilla happily returns true for both tests.

      BUG! It should not return true for either, as it has not implemented
      either, I'll raise a bug when I have some internet access...

      > Opera doesn't support
      > scripting of SVG (yet) I believe, so the test won't work for it, but
      > as far as *script* testing goes this is definately the way to do it.

      Definately not, no user agent should be returning true for DOMImplementation
      tests unless they have actually implemented the specification - ie their
      implementation meets all the conformance requirements, since such a beast is
      unlikely to exist in our lifetimes, we should not bother testing for it, and
      UA's returning true prematurely has already made the tests useless as it
      will be no discriminator even if it does exist one day.

      As you've shown heuristic methods can be used, and these are much more
      useful as they can just test the features of the language that we actually
      use, it doesn't matter to me if you've not authored some feature if I don't
      actually use it.

      Cheers,

      Jim.
    • Mark Birbeck
      Jim, But as you know, the SVG spec defines a large number of possible values to test for with DOMImplementation, not just do you have SVG? . So ideally we
      Message 2 of 23 , Aug 2, 2005
      • 0 Attachment
        Jim,

        But as you know, the SVG spec defines a large number of possible values to
        test for with DOMImplementation, not just "do you have SVG?". So ideally we
        should be able to test for the language features we want, without using
        hacks (sorry, I mean heuristic methods).

        Regards,

        Mark


        Mark Birbeck
        CEO
        x-port.net Ltd.

        e: Mark.Birbeck@...
        t: +44 (0) 20 7689 9232
        w: http://www.formsPlayer.com/
        b: http://internet-apps.blogspot.com/

        Download our XForms processor from
        http://www.formsPlayer.com/

        > -----Original Message-----
        > From: svg-developers@yahoogroups.com
        > [mailto:svg-developers@yahoogroups.com] On Behalf Of Jim Ley
        > Sent: 02 August 2005 11:16
        > To: svg-developers@yahoogroups.com
        > Subject: [svg-developers] Re: Re: Detect native SVG support
        >
        >
        > "Jonathan Watt" <wattie@...> wrote in message
        > news:f21dbbd705072817234af897e9@......
        > > Hi Mark,
        > >
        > > You're quite right of course. This would be the *real* way
        > to do it,
        > > and Mozilla happily returns true for both tests.
        >
        > BUG! It should not return true for either, as it has not
        > implemented either, I'll raise a bug when I have some
        > internet access...
        >
        > > Opera doesn't support
        > > scripting of SVG (yet) I believe, so the test won't work
        > for it, but
        > > as far as *script* testing goes this is definately the way to do it.
        >
        > Definately not, no user agent should be returning true for
        > DOMImplementation tests unless they have actually implemented
        > the specification - ie their implementation meets all the
        > conformance requirements, since such a beast is unlikely to
        > exist in our lifetimes, we should not bother testing for it,
        > and UA's returning true prematurely has already made the
        > tests useless as it will be no discriminator even if it does
        > exist one day.
        >
        > As you've shown heuristic methods can be used, and these are
        > much more useful as they can just test the features of the
        > language that we actually use, it doesn't matter to me if
        > you've not authored some feature if I don't actually use it.
        >
        > Cheers,
        >
        > Jim.
        >
        >
        >
        >
        >
        > ------------------------ Yahoo! Groups Sponsor
        > --------------------~--> <font face=arial size=-1><a
        > href="http://us.ard.yahoo.com/SIG=12hts36l6/M=362131.6882499.7
        > 825260.1510227/D=groups/S=1706030389:TM/Y=YAHOO/EXP=1122992505
        > /A=2889191/R=0/SIG=10r90krvo/*http://www.thebeehive.org
        > ">Get Bzzzy! (real tools to help you find a job) Welcome to
        > the Sweet Life - brought to you by One Economy</a>.</font>
        > --------------------------------------------------------------
        > ------~->
        >
        > -----
        > To unsubscribe send a message to:
        > svg-developers-unsubscribe@yahoogroups.com
        > -or-
        > visit http://groups.yahoo.com/group/svg-developers and click
        > "edit my membership"
        > ----
        > Yahoo! Groups Links
        >
        >
        >
        >
        >
        >
        >
        >
        >
      • Jim Ley
        Mark Birbeck wrote in message news:B841AC96-C2B4-4799-869D-9852793297AE@S009... ... Right, so we agree that Mozilla SVG and
        Message 3 of 23 , Aug 2, 2005
        • 0 Attachment
          "Mark Birbeck" <Mark.Birbeck@...> wrote in message
          news:B841AC96-C2B4-4799-869D-9852793297AE@S009...
          >But as you know, the SVG spec defines a large number of possible values to
          >test for with DOMImplementation, not just "do you have SVG?". So ideally we
          >should be able to test for the language features we want, without using
          >hacks (sorry, I mean heuristic methods).

          Right, so we agree that Mozilla SVG and SideWinder would be non-conformant
          and incorrect if they return true for http://www.w3.org/TR/SVG11/feature#SVG
          since they don't fulfill any of the sub ones.

          Do they fulfil any of the sub-ones, I believe not, because to claim that
          they do, they have to meet:

          ... the viewer can process and render successfully all of the language
          features ...

          Since they contain bugs and partial implementations, they should not claim
          any of the features. Of course this problem is common to the entire feature
          string concept across all languages, but that's just unfortunate that it
          ever became a specifcation.

          The HTML scripting world has shown how heuristic methods are extremely
          successful in dealing with different implementations, no-one uses the
          DOMImplementation interface, it's simply not acceptable due to the wild
          claims of some browsers (claiming support before there was only partial
          support) and the overly-prudent claims of others (claiming no-support when
          there was pretty good support)

          Cheers,

          Jim.
        • Mark Birbeck
          Jim, Sidewinder wouldn t be conformant or non-conformant. We just pass a call to the methods on a DOMImplementation to the implementations that we ve
          Message 4 of 23 , Aug 2, 2005
          • 0 Attachment
            Jim,

            Sidewinder wouldn't be conformant or non-conformant. We just pass a call to
            the methods on a DOMImplementation to the implementations that we've
            instantiated to create that implementation. In the case of an object that
            'supports' SVG it would most likely be ASV (although it would depend on the
            user's machine). So whether ASV (or whatever SVG plug-in is in use) is
            non-conformant for claiming to support whatever feature strings it claims to
            support, is something that I couldn't answer.

            As to adding more and more hacks, no one doubts that you can keep adding
            another line of script to test for yet another permutation. But as the world
            of compound documents gets closer, don't you think that there needs to be a
            generic solution that allows for testing not just for SVG, but for MathML,
            XForms and so on? I recognise your criticisms of DOMImplementation, but at
            some point we do need it to work...or rather it needs to be made to work.

            All the best,

            Mark


            Mark Birbeck
            CEO
            x-port.net Ltd.

            e: Mark.Birbeck@...
            t: +44 (0) 20 7689 9232
            w: http://www.formsPlayer.com/
            b: http://internet-apps.blogspot.com/

            Download our XForms processor from
            http://www.formsPlayer.com/

            > -----Original Message-----
            > From: svg-developers@yahoogroups.com
            > [mailto:svg-developers@yahoogroups.com] On Behalf Of Jim Ley
            > Sent: 02 August 2005 18:30
            > To: svg-developers@yahoogroups.com
            > Subject: [svg-developers] Re: Detect native SVG support
            >
            >
            > "Mark Birbeck" <Mark.Birbeck@...> wrote in message
            > news:B841AC96-C2B4-4799-869D-9852793297AE@S009...
            > >But as you know, the SVG spec defines a large number of
            > possible values
            > >to test for with DOMImplementation, not just "do you have SVG?". So
            > >ideally we should be able to test for the language features we want,
            > >without using hacks (sorry, I mean heuristic methods).
            >
            > Right, so we agree that Mozilla SVG and SideWinder would be
            > non-conformant and incorrect if they return true for
            > http://www.w3.org/TR/SVG11/feature#SVG
            > since they don't fulfill any of the sub ones.
            >
            > Do they fulfil any of the sub-ones, I believe not, because to
            > claim that they do, they have to meet:
            >
            > ... the viewer can process and render successfully all of the
            > language features ...
            >
            > Since they contain bugs and partial implementations, they
            > should not claim any of the features. Of course this problem
            > is common to the entire feature string concept across all
            > languages, but that's just unfortunate that it ever became a
            > specifcation.
            >
            > The HTML scripting world has shown how heuristic methods are
            > extremely successful in dealing with different
            > implementations, no-one uses the DOMImplementation interface,
            > it's simply not acceptable due to the wild claims of some
            > browsers (claiming support before there was only partial
            > support) and the overly-prudent claims of others (claiming
            > no-support when there was pretty good support)
            >
            > Cheers,
            >
            > Jim.
            >
            >
            >
            >
            >
            > ------------------------ Yahoo! Groups Sponsor
            > --------------------~--> <font face=arial size=-1><a
            > href="http://us.ard.yahoo.com/SIG=12hlnqdd7/M=362131.6882499.7
            > 825260.1510227/D=groups/S=1706030389:TM/Y=YAHOO/EXP=1123011190
            > /A=2889191/R=0/SIG=10r90krvo/*http://www.thebeehive.org
            > ">Get Bzzzy! (real tools to help you find a job) Welcome to
            > the Sweet Life - brought to you by One Economy</a>.</font>
            > --------------------------------------------------------------
            > ------~->
            >
            > -----
            > To unsubscribe send a message to:
            > svg-developers-unsubscribe@yahoogroups.com
            > -or-
            > visit http://groups.yahoo.com/group/svg-developers and click
            > "edit my membership"
            > ----
            > Yahoo! Groups Links
            >
            >
            >
            >
            >
            >
            >
            >
            >
          • Jim Ley
            Mark Birbeck wrote in message news:126DE715-5318-467A-A2D7-79A6CE931AE1@S009... ... I would still call SideWinder non-conformant,if
            Message 5 of 23 , Aug 2, 2005
            • 0 Attachment
              "Mark Birbeck" <Mark.Birbeck@...> wrote in message
              news:126DE715-5318-467A-A2D7-79A6CE931AE1@S009...

              >Sidewinder wouldn't be conformant or non-conformant. We just pass a call to
              >the methods on a DOMImplementation to the implementations that we've
              >instantiated to create that implementation.

              I would still call SideWinder non-conformant,if its underlying
              implementation was non-conformant, just like I call IE non-HTTP conformant
              when it's simply using an underlying HTTP implementation, and of course it
              certainly wouldn't be your fault, you're absolutely right to be re-using
              components.

              > But as the world
              >of compound documents gets closer, don't you think that there needs to be a
              >generic solution that allows for testing not just for SVG, but for MathML,
              >XForms and so on?

              Absolutely, but DOMImplementation is not it, I'm not sure what it would look
              like, indeed I've tried thinking of something before, as the criticisms of
              DOM Implementation are not new, but I've never come up with anything, 'nor
              heard of anything.

              Cheers,

              Jim.
            • Bjoern Hoehrmann
              ... I certainly agree. Such facilities aren t cheap though, you need * Very clear well-defined conformance requirements * Extensive test suites for these
              Message 6 of 23 , Aug 2, 2005
              • 0 Attachment
                * Mark Birbeck wrote:
                >As to adding more and more hacks, no one doubts that you can keep adding
                >another line of script to test for yet another permutation. But as the world
                >of compound documents gets closer, don't you think that there needs to be a
                >generic solution that allows for testing not just for SVG, but for MathML,
                >XForms and so on? I recognise your criticisms of DOMImplementation, but at
                >some point we do need it to work...or rather it needs to be made to work.

                I certainly agree. Such facilities aren't cheap though, you need

                * Very clear well-defined conformance requirements
                * Extensive test suites for these facilities
                * Strong commitment to 100% compliance
                * Strong and reliable means to enforce compliance

                At the moment, for SVG feature strings, we don't have any of that. In
                fact, it's probably worse, there are still many authors who do not care
                at all even about most basic conformance requirements like using well-
                formed XML; many authors still think "We usually find checking with
                [IE, Opera and Firefox] is enough for our work". If we can't convince
                authors to ensure their content is conforming, there is little hope to
                convince implementers to code correctly.
                --
                Björn Höhrmann · mailto:bjoern@... · http://bjoern.hoehrmann.de
                Weinh. Str. 22 · Telefon: +49(0)621/4309674 · http://www.bjoernsworld.de
                68309 Mannheim · PGP Pub. KeyID: 0xA4357E78 · http://www.websitedev.de/
              Your message has been successfully submitted and would be delivered to recipients shortly.