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

Re: [svg-developers] Siggraph and DENG...

Expand Messages
  • Jerrold Maddox
    Chris How far do you think Deng goes toward filling the browser vacuum you spoke of in your closing keynote? If it goes some of the way, what can we do
    Message 1 of 24 , Aug 1, 2003
    View Source
    • 0 Attachment
      Chris

      How far do you think Deng goes toward filling the "browser vacuum" you
      spoke of in your closing keynote?

      If it goes some of the way, what can we do individually and/or collectively
      to contribute to helping them in filling it?

      Jerry

      At 04:27 PM 8/1/2003 +0200, you wrote:
      >On Friday, August 1, 2003, 3:54:20 PM, Claus wrote:
      >
      > >> This is the viewer that I used as an example in my
      > >> closing keynote at SVG Open 2003.
      >
      >CW> that's pretty cool chris.
      >
      >Well, thanks for mailing me the updated beta in time for my talk.
      >
      >CW> did you get any feedback from the crowd?
      >
      >Well, no-one threw rocks. I did mention that it was all implemented in
      >actionscript on Flash 6MX, which caused a few surprised faces. I
      >demonstrated xframes, xforms, and the css3 ui and css3 selectors being
      >used to style xforms abstract widgets.
      >
      >My keynote was about the need to move out from just SVG into a wider,
      >multi-namespace world to get a more fully featured, standards based and
      >widely interoperable client base for the desktop and mobile web.
      >
      > >>> and I am told that DENG is tested to work with new
      > >>> Flash player 7 coming soon?
      >
      > >> Yes, it does, with a useful but modest speed increase (its
      > >> still very slow, but a useful proof of concept).
      >
      >CW> especially for the DENG SVG module we experienced quite a
      >CW> performance boost with the new flash player beta (up to double
      >CW> speed), without any changes in our code. there is still a lot of
      >CW> room for optimizations, as we have loads of overhead (mandatory
      >CW> for flash player 6) that we can get rid of completely, when
      >CW> targeting the flash player 7.
      >
      >Ah, so a Flash 7 only build will be faster than one that supports both
      >6 and 7?
      >
      >CW> especially CSS (inheritence/cascade) is quite a performance eater.
      >
      >Particularly when the tree can mutate. This is one of those areas
      >where performance optimisation (eg, building dependency graphs,
      >trying to constrain the possible scope of any update, etc) can really
      >pay off.
      >
      >
      >--
      >Chris mailto:chris@...
      >
      >
      >Yahoo! Groups Sponsor
      >ADVERTISEMENT
      ><http://rd.yahoo.com/M=259538.3625325.4914071.1261774/D=egroupweb/S=1706030389:HM/A=1695348/R=0/SIG=11u38u3s2/*http://hits.411web.com/cgi-bin/hit?page=1374-105951838331032>37285c.jpg
      >372908.jpg
      >
      >-----
      >To unsubscribe send a message to: svg-developers-unsubscribe@yahoogroups.com
      >-or-
      >visit
      ><http://groups.yahoo.com/group/svg-developers>http://groups.yahoo.com/group/svg-developers
      >and click "edit my membership"
      >----
      >
      >Your use of Yahoo! Groups is subject to the
      ><http://docs.yahoo.com/info/terms/>Yahoo! Terms of Service.

      ----------



      [Non-text portions of this message have been removed]
    • Ronan Oger
      Hi all, I have not seen much discussion here about Mozilla with native SVG support, which put out their beta release on 20 July. I ve tried it and while it is
      Message 2 of 24 , Aug 1, 2003
      View Source
      • 0 Attachment
        Hi all,

        I have not seen much discussion here about Mozilla with native SVG
        support, which put out their beta release on 20 July.

        I've tried it and while it is not today *quite* SVG-compliant, it is
        quite good. And it's good enough for most of our intenet-limited
        applications anyhow...

        Note that there is an issue with mime type because of our community's
        insistance on using image/svg+xml as the SVG mime type in spite of it
        not being *actually* accredited. The current version of the browser I'm
        using (I am using the Linux version) is 1.5b, and it is excellent.

        I'd say that within 1 year this distro could well be the standard to
        work against if they keep picking up momentum.

        http://www.croczilla.com/svg/

        There's more info

        On Fri, 2003-08-01 at 17:36, Jerrold Maddox wrote:
        > Chris
        >
        > How far do you think Deng goes toward filling the "browser vacuum" you
        > spoke of in your closing keynote?
        >
        > If it goes some of the way, what can we do individually and/or collectively
        > to contribute to helping them in filling it?
        >
        > Jerry
        >
        > At 04:27 PM 8/1/2003 +0200, you wrote:
        > >On Friday, August 1, 2003, 3:54:20 PM, Claus wrote:
        > >
        > > >> This is the viewer that I used as an example in my
        > > >> closing keynote at SVG Open 2003.
        > >
        > >CW> that's pretty cool chris.
        > >
        > >Well, thanks for mailing me the updated beta in time for my talk.
        > >
        > >CW> did you get any feedback from the crowd?
        > >
        > >Well, no-one threw rocks. I did mention that it was all implemented in
        > >actionscript on Flash 6MX, which caused a few surprised faces. I
        > >demonstrated xframes, xforms, and the css3 ui and css3 selectors being
        > >used to style xforms abstract widgets.
        > >
        > >My keynote was about the need to move out from just SVG into a wider,
        > >multi-namespace world to get a more fully featured, standards based and
        > >widely interoperable client base for the desktop and mobile web.
        > >
        > > >>> and I am told that DENG is tested to work with new
        > > >>> Flash player 7 coming soon?
        > >
        > > >> Yes, it does, with a useful but modest speed increase (its
        > > >> still very slow, but a useful proof of concept).
        > >
        > >CW> especially for the DENG SVG module we experienced quite a
        > >CW> performance boost with the new flash player beta (up to double
        > >CW> speed), without any changes in our code. there is still a lot of
        > >CW> room for optimizations, as we have loads of overhead (mandatory
        > >CW> for flash player 6) that we can get rid of completely, when
        > >CW> targeting the flash player 7.
        > >
        > >Ah, so a Flash 7 only build will be faster than one that supports both
        > >6 and 7?
        > >
        > >CW> especially CSS (inheritence/cascade) is quite a performance eater.
        > >
        > >Particularly when the tree can mutate. This is one of those areas
        > >where performance optimisation (eg, building dependency graphs,
        > >trying to constrain the possible scope of any update, etc) can really
        > >pay off.
        > >
        > >
        > >--
        > >Chris mailto:chris@...
        > >
        > >
        > >Yahoo! Groups Sponsor
        > >ADVERTISEMENT
        > ><http://rd.yahoo.com/M=259538.3625325.4914071.1261774/D=egroupweb/S=1706030389:HM/A=1695348/R=0/SIG=11u38u3s2/*http://hits.411web.com/cgi-bin/hit?page=1374-105951838331032>37285c.jpg
        > >372908.jpg
        > >
        > >-----
        > >To unsubscribe send a message to: svg-developers-unsubscribe@yahoogroups.com
        > >-or-
        > >visit
        > ><http://groups.yahoo.com/group/svg-developers>http://groups.yahoo.com/group/svg-developers
        > >and click "edit my membership"
        > >----
        > >
        > >Your use of Yahoo! Groups is subject to the
        > ><http://docs.yahoo.com/info/terms/>Yahoo! Terms of Service.
        >
        > ----------
        >
        >
        >
        > [Non-text portions of this message have been removed]
        >
        >
        >
        > -----
        > 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"
        > ----
        >
        > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
        >
        >
        --
        Ronan Oger <ronan@...>
        ROASP - the serverside SVG portal
      • Jim Ley
        Ronan Oger wrote in message news:1059754964.9908.606.camel@dev.roitsystems.com... ... Yeah, it surprised me since I couldn t see much
        Message 3 of 24 , Aug 1, 2003
        View Source
        • 0 Attachment
          "Ronan Oger" <ronan@...> wrote in message
          news:1059754964.9908.606.camel@......
          > I have not seen much discussion here about Mozilla with native SVG
          > support, which put out their beta release on 20 July.

          Yeah, it surprised me since I couldn't see much movement since jan/feb time,
          so wondered how it suddenly got into a public release.

          > Note that there is an issue with mime type because of our community's
          > insistance on using image/svg+xml as the SVG mime type in spite of it
          > not being *actually* accredited.

          text/xml while correct for labelling SVG, is completely useless for doing
          so, since there's no reason for UA's to render it, and certainly anything
          but these lets handle everything UA's will not be able to on current
          internet architecture. I'm not happy that image/svg+xml exists, but we need
          something for SVG. Mozilla has some very strange ideas on mime-types
          (prefering Joe's Random XML format which it's not got a clue how to render
          to HTML for example)

          > I'd say that within 1 year this distro could well be the standard to
          > work against if they keep picking up momentum.

          I think they need a lot of developers getting involved for that to happen.

          KSVG is coming along nicely too, and IMO Konq's a nicer UA than Mozilla.

          Jim.
        • Ronan Oger
          ... I can t say I agree with this. text/xml instructs the UA that it is dealing with proper xml and tells it to go find out if it knows what to do with the
          Message 4 of 24 , Aug 1, 2003
          View Source
          • 0 Attachment
            On Fri, 2003-08-01 at 18:32, Jim Ley wrote:
            > "Ronan Oger" <ronan@...> wrote in message
            > news:1059754964.9908.606.camel@......
            > > I have not seen much discussion here about Mozilla with native SVG
            > > support, which put out their beta release on 20 July.
            >
            > Yeah, it surprised me since I couldn't see much movement since jan/feb time,
            > so wondered how it suddenly got into a public release.
            >
            > > Note that there is an issue with mime type because of our community's
            > > insistance on using image/svg+xml as the SVG mime type in spite of it
            > > not being *actually* accredited.
            >
            > text/xml while correct for labelling SVG, is completely useless for doing
            > so, since there's no reason for UA's to render it, and certainly anything
            > but these lets handle everything UA's will not be able to on current
            > internet architecture. I'm not happy that image/svg+xml exists, but we need
            > something for SVG. Mozilla has some very strange ideas on mime-types
            > (prefering Joe's Random XML format which it's not got a clue how to render
            > to HTML for example)
            >

            I can't say I agree with this. text/xml instructs the UA that it is
            dealing with proper xml and tells it to go find out if it knows what to
            do with the dtd/schema at hand.

            case in point: moz can handle numerous types of xml, and renders them
            appropriately - svg is drawn as svg, and xhtml is drawn as that, and
            mathml as that too. We don't need mathml to have mime type
            poetry/structre+process to know it's mathml. All we need is its dtd and
            a hook into the app to let us process it.

            Isn't that the idea of xml, after all?

            > > I'd say that within 1 year this distro could well be the standard to
            > > work against if they keep picking up momentum.
            >
            > I think they need a lot of developers getting involved for that to happen.
            >
            > KSVG is coming along nicely too, and IMO Konq's a nicer UA than Mozilla.
            >

            agreed.


            ronan
          • Jim Ley
            Claus Wahlers wrote in message news:MGEIIMPHDDELBHEBDJKHKEHIENAA.claus@ego7.com... ... So moving away from the ECMAScript standard again,
            Message 5 of 24 , Aug 2, 2003
            View Source
            • 0 Attachment
              "Claus Wahlers" <claus@...> wrote in message
              news:MGEIIMPHDDELBHEBDJKHKEHIENAA.claus@......

              > and i hear people talking about real oop

              So moving away from the ECMAScript standard again, seems strange after they
              moved towards it.

              > String.charAt for example got proportionally slower, the longer the string
              > got (a desaster when parsing long strings,

              Were the RegExp methods significantly slower too?

              Jim.
            • Jim Ley
              Chris Lilley wrote in message news:1278609031.20030801162759@w3.org... ... We d asked the hotel for rotten fruit, unfortunately they weren t
              Message 6 of 24 , Aug 2, 2003
              View Source
              • 0 Attachment
                "Chris Lilley" <chris@...> wrote in message
                news:1278609031.20030801162759@......
                > On Friday, August 1, 2003, 3:54:20 PM, Claus wrote:
                > CW> did you get any feedback from the crowd?
                >
                > Well, no-one threw rocks.

                We'd asked the hotel for rotten fruit, unfortunately they weren't able to
                supply us in time.

                Jim.
              • Robin Debreuil
                ... they ... No, ecmaScript has moved further towards true oo. I am very tired of pretending classes are functions and fiddling with the prototype, I hope
                Message 7 of 24 , Aug 2, 2003
                View Source
                • 0 Attachment
                  > So moving away from the ECMAScript standard again, seems strange after
                  they
                  > moved towards it.

                  No, ecmaScript has moved further towards true oo. I am very tired of
                  pretending classes are functions and fiddling with the prototype, I hope
                  actionscript (and javascript) get up to speed this year...

                  Cheers,
                  Robin

                  PS I hadn't check out DENG in a while Claus, really fabulous stuff!!
                • Jim Ley
                  Robin Debreuil wrote in message news:002e01c3596f$8e34c210$0700a8c0@xpbaby... ... No it hasn t... or do you have an Ed. 4 draft to hand?
                  Message 8 of 24 , Aug 2, 2003
                  View Source
                  • 0 Attachment
                    "Robin Debreuil" <robin@...> wrote in message
                    news:002e01c3596f$8e34c210$0700a8c0@xpbaby...
                    > > So moving away from the ECMAScript standard again, seems strange after
                    > they
                    > > moved towards it.
                    >
                    > No, ecmaScript has moved further towards true oo.

                    No it hasn't... or do you have an Ed. 4 draft to hand?

                    > I am very tired of
                    > pretending classes are functions and fiddling with the prototype, I hope
                    > actionscript (and javascript) get up to speed this year...

                    Ed. 4 will move towards OO (if it ever arrives, and I'm beginning to suspect
                    it won't)

                    There's nothing out there yet though.

                    Jim.
                  • Robin Debreuil
                    ... A draft? Sure: http://www.mozilla.org/js/language/es4/index.html ... So everyone should just live with crappy support for oo, frozen in time? Maybe in a
                    Message 9 of 24 , Aug 3, 2003
                    View Source
                    • 0 Attachment
                      > > No, ecmaScript has moved further towards true oo.
                      >
                      > No it hasn't... or do you have an Ed. 4 draft to hand?

                      A draft? Sure:
                      http://www.mozilla.org/js/language/es4/index.html

                      > Ed. 4 will move towards OO (if it ever arrives, and I'm
                      > beginning to suspect it won't)
                      So everyone should just live with crappy support for oo, frozen in time?
                      Maybe in a perfect world (where no one has to work for a living that is).

                      I do agree with you about inline comments though, maybe we can build on that
                      : ).

                      Cheers,
                      Robin
                    • Jim Ley
                      Robin Debreuil wrote in message news:000f01c359e0$5c9ac8a0$0700a8c0@xpbaby... ... That was Netscapes proposal, and is very old now,
                      Message 10 of 24 , Aug 3, 2003
                      View Source
                      • 0 Attachment
                        "Robin Debreuil" <robin@...> wrote in message
                        news:000f01c359e0$5c9ac8a0$0700a8c0@xpbaby...
                        > > > No, ecmaScript has moved further towards true oo.
                        > >
                        > > No it hasn't... or do you have an Ed. 4 draft to hand?
                        >
                        > A draft? Sure:
                        > http://www.mozilla.org/js/language/es4/index.html

                        That was Netscapes proposal, and is very old now, there were publicly
                        available (by accident I believe) drafts later than that, I still have one
                        on back up systems somewhere. The very fact that nothing has happened in
                        over 2 years, and the silence when I ask TC39 folks is what's making me
                        think there's not really anything there. The JScript.NET implementation
                        also contains much that was in the Ed. 4 thinking, but we've not got an Ed.4
                        and as I say, I don't think it's likely.

                        > > Ed. 4 will move towards OO (if it ever arrives, and I'm
                        > > beginning to suspect it won't)
                        > So everyone should just live with crappy support for oo, frozen in time?

                        Oh no, I think it would be good, I was just asking if Flash was moving away
                        from standards to proprietary again, that was all.

                        Jim.
                      • evolgrafix_aadam
                        hey Jim, ... publicly ... have one ... happened in ... making me ... implementation ... got an Ed.4 ... How can it be old when the last change on it was on
                        Message 11 of 24 , Aug 3, 2003
                        View Source
                        • 0 Attachment
                          hey Jim,

                          > That was Netscapes proposal, and is very old now, there were
                          publicly
                          > available (by accident I believe) drafts later than that, I still
                          have one
                          > on back up systems somewhere. The very fact that nothing has
                          happened in
                          > over 2 years, and the silence when I ask TC39 folks is what's
                          making me
                          > think there's not really anything there. The JScript.NET
                          implementation
                          > also contains much that was in the Ed. 4 thinking, but we've not
                          got an Ed.4
                          > and as I say, I don't think it's likely.

                          How can it be old when the last change on it was on 30th June 03 ?
                          I read shortly through it and mostly like it, wondering why these
                          guys are so slow to give something out? I mean, when Ed. 4 should
                          *ever* be out, then it also takes a longer time for implementations.

                          Alexander
                        • Jim Ley
                          evolgrafix_aadam wrote in message news:bgjque+kttd@eGroups.com... ... Ah, sorry, yes they are still updating it, you ll note it
                          Message 12 of 24 , Aug 3, 2003
                          View Source
                          • 0 Attachment
                            "evolgrafix_aadam" <evolgrafix_aadam@...> wrote in message
                            news:bgjque+kttd@......
                            > How can it be old when the last change on it was on 30th June 03 ?
                            > I read shortly through it and mostly like it, wondering why these
                            > guys are so slow to give something out? I mean, when Ed. 4 should
                            > *ever* be out, then it also takes a longer time for implementations.

                            Ah, sorry, yes they are still updating it, you'll note it says things like:

                            "The TC39TG1 working group's current schedule calls for a release of a
                            ECMAScript Edition 4 standard sometime in 2002."

                            Nothing I can get out of TC39 suggests development is ongoing. I'd like to
                            see it though.

                            Jim.
                          • evolgrafix_aadam
                            hi, ... of a ... I d like to ... huh? what about some lines after the index: The following are recent major changes in this document: Date Revisions Jun 30,
                            Message 13 of 24 , Aug 3, 2003
                            View Source
                            • 0 Attachment
                              hi,

                              > Ah, sorry, yes they are still updating it, you'll note it says
                              things like:
                              >
                              > "The TC39TG1 working group's current schedule calls for a release
                              of a
                              > ECMAScript Edition 4 standard sometime in 2002."
                              >
                              > Nothing I can get out of TC39 suggests development is ongoing.
                              I'd like to
                              > see it though.

                              huh? what about some lines after the index:

                              The following are recent major changes in this document:

                              Date Revisions
                              Jun 30, 2003 Defined «Udddddddd» notation for supplementary
                              characters.
                              Defined code points, code units, supplementary characters, and
                              surrogates.
                              Defined the SupplementaryChar and Char21 classes and corresponding
                              utility functions. Revised the description of strings to clarify
                              that they are composed of Char16 code units and sort as such.
                              Added «u0085» to the LineTerminator productions.
                              Added numerous missing functions and classes to the semantics. Fixed
                              bugs in the semantics.

                              Alexander
                            • Ahmet Zorlu
                              afaik, ECMAScript 4th Edition is scheduled for publication for the first quarter of year 2004
                              Message 14 of 24 , Aug 4, 2003
                              View Source
                              • 0 Attachment
                                afaik, ECMAScript 4th Edition is scheduled for publication for the first quarter of year 2004

                                <quote source="http://xml.coverpages.org/ECMAScript-XML.html">
                                ECMA is in the process of harmonizing the various diverging extensions of ECMAScript. The full second version of the language is scheduled for publication as ECMA-262
                                Edition 4 in Q1 2004. This will update the standard with respect to the language and the various differing implementations.
                                </quote>

                                JScript .NET is a fully functional implementation of the upcoming standard.
                                I would like to see that JavaScript and ActionScript go in the same direction.
                                As (imnsho) we need a truly OOP language for writing robust library packages and highly effective code.

                                // A. Zorlu
                                // Web developer
                              • Claus Wahlers
                                ... did flash ever move towards standards? ;) cheers, claus. http://mozquito.com http://w3blog.com/news
                                Message 15 of 24 , Aug 4, 2003
                                View Source
                                • 0 Attachment
                                  > Oh no, I think it would be good, I was just asking if
                                  > Flash was moving away from standards to proprietary
                                  > again, that was all.

                                  did flash ever move towards standards? ;)
                                  cheers,
                                  claus.
                                  http://mozquito.com
                                  http://w3blog.com/news
                                • Alexander Adam
                                  ... No and that is good how it is. We have a great standard, why are the peoples always bothering with that .. ahem.. Flash? Alexander Adam Privately
                                  Message 16 of 24 , Aug 4, 2003
                                  View Source
                                  • 0 Attachment
                                    --- In svg-developers@yahoogroups.com, "Claus Wahlers" <claus@e...>
                                    wrote:
                                    > > Oh no, I think it would be good, I was just asking if
                                    > > Flash was moving away from standards to proprietary
                                    > > again, that was all.
                                    >
                                    > did flash ever move towards standards? ;)


                                    No and that is good how it is. We have a great standard, why are the
                                    peoples always bothering with that .. ahem.. Flash?

                                    Alexander Adam
                                    Privately
                                  • Chris Lilley
                                    On Friday, August 1, 2003, 9:48:37 PM, Ronan wrote: ... In accordance with the SVg specification ... There is no accreditation process. There is a registration
                                    Message 17 of 24 , Aug 4, 2003
                                    View Source
                                    • 0 Attachment
                                      On Friday, August 1, 2003, 9:48:37 PM, Ronan wrote:

                                      RO> On Fri, 2003-08-01 at 18:32, Jim Ley wrote:
                                      >> "Ronan Oger" <ronan@...> wrote in message
                                      >> news:1059754964.9908.606.camel@......
                                      >> > I have not seen much discussion here about Mozilla with native SVG
                                      >> > support, which put out their beta release on 20 July.
                                      >>
                                      >> Yeah, it surprised me since I couldn't see much movement since jan/feb time,
                                      >> so wondered how it suddenly got into a public release.
                                      >>
                                      >> > Note that there is an issue with mime type because of our community's
                                      >> > insistance on using image/svg+xml as the SVG mime type

                                      In accordance with the SVg specification

                                      >> in spite of it
                                      >> > not being *actually* accredited.

                                      There is no accreditation process. There is a registration process,
                                      which has known problems and is being revised to alow standfards
                                      organisations other than IETF (such as W3C) to register media types.
                                      W3C is tracking these changes.

                                      >>
                                      >> text/xml while correct for labelling SVG,


                                      No, its not correct.

                                      Text/xml has a known flaw that the character encoding has to be
                                      US-ASDCII. UTF-16 is not allowed and UTF-8 can only be used if the
                                      ASCII-subset of it is used. This makes text/xml essentially useless,
                                      and it should be deprecated.

                                      >> is completely useless for doing
                                      >> so, since there's no reason for UA's to render it, and certainly anything
                                      >> but these lets handle everything UA's will not be able to on current
                                      >> internet architecture. I'm not happy that image/svg+xml exists,

                                      Why, it sexactly the right type for an image format that is in xml.
                                      Thats what the +xml is *for*.

                                      >> but we need
                                      >> something for SVG. Mozilla has some very strange ideas on mime-types
                                      >> (prefering Joe's Random XML format which it's not got a clue how to render
                                      >> to HTML for example)

                                      RO> I can't say I agree with this. text/xml instructs the UA that it is
                                      RO> dealing with proper xml


                                      As does application/xml and anything/anything+xml

                                      RO> and tells it to go find out if it knows what to
                                      RO> do with the dtd/schema at hand.

                                      Please cite some evidence to support that assertion about "go find out
                                      if it knows what to do with the dtd/schema at hand"

                                      RO> case in point: moz can handle numerous types of xml, and renders them
                                      RO> appropriately - svg is drawn as svg,

                                      because its special cased on the namespace URI.

                                      RO> and xhtml is drawn as that, and
                                      RO> mathml as that too. We don't need mathml to have mime type
                                      RO> poetry/structre+process to know it's mathml. All we need is its dtd

                                      This is incorrect. Try changing the DTD and see what happens.

                                      RO> and a hook into the app to let us process it.

                                      RO> Isn't that the idea of xml, after all?

                                      >> > I'd say that within 1 year this distro could well be the standard to
                                      >> > work against if they keep picking up momentum.
                                      >>
                                      >> I think they need a lot of developers getting involved for that to happen.
                                      >>
                                      >> KSVG is coming along nicely too, and IMO Konq's a nicer UA than Mozilla.

                                      RO> agreed.

                                      Is ksvg still linux-only?


                                      --
                                      Chris mailto:chris@...
                                    • jdowdell@macromedia.com
                                      ... Sorry, I dozed off while you were arguing when an ECMAScript draft was published... what was the question again...? ;-) (NB: Title trimmed in accordance
                                      Message 18 of 24 , Aug 4, 2003
                                      View Source
                                      • 0 Attachment
                                        >>> Oh no, I think it would be good, I was just asking if
                                        >>> Flash was moving away from standards to proprietary
                                        >>> again, that was all.
                                        >>
                                        >> did flash ever move towards standards? ;)
                                        >
                                        > No and that is good how it is. We have a great standard, why
                                        > are the peoples always bothering with that .. ahem.. Flash?


                                        Sorry, I dozed off while you were arguing when an ECMAScript draft was
                                        published... what was the question again...? ;-)

                                        (NB: Title trimmed in accordance with RFC1855, but not yet renamed because
                                        drift is not confirmed.)

                                        jd










                                        John Dowdell, Macromedia Developer Support, San Francisco
                                        (Best to reply on-list, to avoid my mighty spam filters!)
                                        Technotes: http://www.macromedia.com/support/search/
                                        Column: http://www.macromedia.com/desdev/jd_forum/
                                        Technical daily diary: http://www.macromedia.com/go/blog_jd
                                      • Ronan Oger
                                        This is getting confusing. Pls bear with me... First of all, pleeeease don t overly nitpick exerpts assuming that they are on their own icons of opinion. they
                                        Message 19 of 24 , Aug 5, 2003
                                        View Source
                                        • 0 Attachment
                                          This is getting confusing. Pls bear with me...

                                          First of all, pleeeease don't overly nitpick exerpts assuming that they
                                          are on their own icons of opinion. they fit within the context as much
                                          as they stand up on their own. most of what i say makes sense (to me at
                                          least) in context but falls apart on its own.... ;-)


                                          On Mon, 2003-08-04 at 18:32, Chris Lilley wrote:
                                          > On Friday, August 1, 2003, 9:48:37 PM, Ronan wrote:
                                          >
                                          > RO> On Fri, 2003-08-01 at 18:32, Jim Ley wrote:
                                          > >> "Ronan Oger" <ronan@...> wrote in message
                                          > >> news:1059754964.9908.606.camel@......
                                          > >> > I have not seen much discussion here about Mozilla with native SVG
                                          > >> > support, which put out their beta release on 20 July.
                                          > >>
                                          > >> Yeah, it surprised me since I couldn't see much movement since jan/feb time,
                                          > >> so wondered how it suddenly got into a public release.
                                          > >>
                                          > >> > Note that there is an issue with mime type because of our community's
                                          > >> > insistance on using image/svg+xml as the SVG mime type
                                          >
                                          > In accordance with the SVg specification

                                          right, and I think everyone respects and accepts this.


                                          >
                                          > >> in spite of it
                                          > >> > not being *actually* accredited.
                                          >
                                          > There is no accreditation process. There is a registration process,
                                          > which has known problems and is being revised to alow standfards
                                          > organisations other than IETF (such as W3C) to register media types.
                                          > W3C is tracking these changes.
                                          >

                                          right. wrong vocabulary, right meaning. blame it on my mother toungue
                                          :-)
                                          image/svg+xml is not registered. It would certainly solidify my
                                          argument with 'purist' client implementation people if the mime type I
                                          proposed for them to use to recognize svg was registered by Those Who
                                          Know. (maybe something ronam/likes+the+name+xml?)

                                          So the Big Question for everyone is, of course, When will it be
                                          registered so we can put this issue to bed? (mind you, Mozilla has built
                                          a branch that recognises image/svg+xml as a synonym for text/xml.
                                          Hopefully this will solve rather than complicate the issue).

                                          > >>
                                          > >> text/xml while correct for labelling SVG,
                                          >
                                          >
                                          > No, its not correct.

                                          Why is it not correct? Other than the w3c wg specifying that a new
                                          non-registered mime type is the new mime type for our particular flavour
                                          of xml, why is the umbrella mime type unacceptable?

                                          I do realize that this question is blasphemous and immoral. but the real
                                          question I am asking is 'is it really important (or desirable) to have
                                          OurOwnMimeType for every dialect of XML that comes over the horizon?

                                          And if that's the case, why did the wg bother using xml at all, with all
                                          that overhead that the user agents obviously don't need like the xml
                                          declaration, the dtd resource, etc? If you already know what it is, why
                                          bother with structure like xml?

                                          hey... i have an idea... let's start again from scratch.... Clearly my
                                          question is almost pointless at this stage given that we are using and
                                          are mostly (I am completely) satisfied with svg. But the question is
                                          relevant in the communication and handshaking part of my context (The
                                          part where the server says 'here is your document, and here's what you
                                          do with it').

                                          IS SVG not xml? Is it not enough for a UA to know that it is dealing
                                          with text? My understanding of XML is that it is intended to be
                                          self-describing if it is correctly written. Could you pls. explain?

                                          I would be very dissatisfied with the mimetype people if in 50 years I
                                          have to have a separate hard drive on my PC to store the possibly huge
                                          mime-type mapping matrix that would grow from a desire to support 'all'
                                          useful mime types.

                                          > Text/xml has a known flaw that the character encoding has to be
                                          > US-ASDCII. UTF-16 is not allowed and UTF-8 can only be used if the
                                          > ASCII-subset of it is used. This makes text/xml essentially useless,
                                          > and it should be deprecated.
                                          >

                                          Ah. interesting. UTF-16 problems are an issue. Need to look further
                                          into it.
                                          ...But I don't get it. You are clearly associating text/xml with a
                                          particular encoding rather than a content (read mime) type. Why is it
                                          incorrect to say that SVG is xml ,and thereform falls under the text/xml
                                          mime type?

                                          > >> is completely useless for doing
                                          > >> so, since there's no reason for UA's to render it, and certainly anything
                                          > >> but these lets handle everything UA's will not be able to on current
                                          > >> internet architecture. I'm not happy that image/svg+xml exists,
                                          >
                                          > Why, it sexactly the right type for an image format that is in xml.
                                          > Thats what the +xml is *for*.

                                          right. no argument from me against *having* image/svg+xml if it adds
                                          knowledge to my UA when it fetches a document. But if all my UA needs is
                                          *either* the mime type or the xml document header to correctly render a
                                          page, then I don't see why we are using mime type distinction. But
                                          that's just my simpleton opinion, and I'm sure You Lerned Men have a
                                          good reason that was not explored here so far.

                                          >
                                          > >> but we need
                                          > >> something for SVG. Mozilla has some very strange ideas on mime-types
                                          > >> (prefering Joe's Random XML format which it's not got a clue how to render
                                          > >> to HTML for example)
                                          >
                                          > RO> I can't say I agree with this. text/xml instructs the UA that it is
                                          > RO> dealing with proper xml
                                          > As does application/xml and anything/anything+xml
                                          >

                                          Right. But is that a problem? Do we need to add mime types? If there is
                                          a need to, such my guessed pressing need for the plugins to interpret
                                          it, then obviously a practical solution is better than a theoretically
                                          pure one. But otherwise, it might appear that someone just wanted a cool
                                          new mime type.

                                          > RO> and tells it to go find out if it knows what to
                                          > RO> do with the dtd/schema at hand.
                                          >
                                          > Please cite some evidence to support that assertion about "go find out
                                          > if it knows what to do with the dtd/schema at hand"
                                          >

                                          cite? i think you misunderstand me. The UA will decide what to do with a
                                          document based on the mime type and based on the dtd (and the other
                                          parts of the header). The mime type tells us how to interpret the body
                                          of the message, and the header (with dtd, schema, root element,
                                          resources, etc.) tells us the context of the body and provides the hooks
                                          into the UA for interpretation/rendering/invocation.

                                          Either soft coded or hard coded, but when a UA is passed a url, it first
                                          interprets the mime type to decide how to parse the document, then
                                          parses the document, then interprets the parsed information.

                                          A primitive UA will take the mime type and apply a fixed rule set. A
                                          more advanced UA will parse the content and use information there (for
                                          example in the HTML header for html).


                                          > RO> case in point: moz can handle numerous types of xml, and renders them
                                          > RO> appropriately - svg is drawn as svg,
                                          >
                                          > because its special cased on the namespace URI.
                                          >

                                          Right (*nodding*)

                                          > RO> and xhtml is drawn as that, and
                                          > RO> mathml as that too. We don't need mathml to have mime type
                                          > RO> poetry/structre+process to know it's mathml. All we need is its dtd
                                          >
                                          > This is incorrect. Try changing the DTD and see what happens.
                                          >

                                          Right. in this case isn't it because the UA is making assumptions about
                                          the dtd to use? The header is supposed to tell us what to do with a
                                          document content and too many UAs cheat and guess what to do without
                                          checking that they have the right definitions.

                                          With xhtml we use the parsers to verify we have the right dtd. Does the
                                          same not apply to svg and the other xmls? Are we not simply being sloppy
                                          in our implementations? It seems to me that we are. We've already been
                                          living in a 1-UA world (svg-wise) for 2 years and I find myself getting
                                          irrationally irritated at squiggle when it refuses to render my pages
                                          which render in ASV or CSV. But whose fault is that? Probably my own
                                          fault for not making sure I am using strictly-legal svg.

                                          Wouldn't strict dtd enforcement deal with this?

                                          > RO> and a hook into the app to let us process it.
                                          >
                                          > RO> Isn't that the idea of xml, after all?
                                          >
                                          > >> > I'd say that within 1 year this distro could well be the standard to
                                          > >> > work against if they keep picking up momentum.
                                          > >>
                                          > >> I think they need a lot of developers getting involved for that to happen.
                                          > >>
                                          > >> KSVG is coming along nicely too, and IMO Konq's a nicer UA than Mozilla.
                                          >
                                          > RO> agreed.
                                          >
                                          > Is ksvg still linux-only?

                                          it's kde only. kde also works on unix/FreeBSD afakik (I have used kde on
                                          unix about 2 years ago). The problem will of course be with the
                                          libraries.

                                          I see no reason why it would not work on OSX. But then again, not having
                                          to implement it myself provides me the luxury of making vage assertion.

                                          Ronan
                                          --
                                          Ronan Oger <ronan@...>
                                          ROASP - the serverside SVG portal
                                        Your message has been successfully submitted and would be delivered to recipients shortly.