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

Re: [svg-developers] Preview on KDE 3.1 SVG-Icons

Expand Messages
  • Rick Bullotta
    Actually, using SVG for monitoring of industrial processes is a great example of where real-time rescaling is of value. Often the user will need to open
    Message 1 of 20 , Sep 1, 2002
    • 0 Attachment
      Actually, using SVG for monitoring of industrial
      processes is a great example of where real-time
      rescaling is of value. Often the user will need to
      open multiple windows, requiring rescaling to
      accomodate the precious limited screen "real estate"
      available.

      In the ideal situation, the user interfaces can
      support dynamic "decluttering" based on the zoom
      level...at the "50,000 foot level", detailed data
      that might be unreadable or
      unnecessarily "cluttering" would be automatically
      rendered invisible, whereas as the user zoomed in,
      new detail would reveal itself.

      While this could be done with tons of ugly
      scripting, it sure would be a nice feature in core
      SVG - threshold visibility, based on the zoom level,
      which could be done declaratively rather than
      through voluminous scripts.

      I'm starting to think that there's a need for a
      parallel spec to SVG to focus on application-level
      functionality (SVG programmable objects, non-
      scripting dynamic behaviors that go beyond SMIL
      stuff, contextual behaviors, prebuilt XForms-like UI
      components, etc.). But since I can't afford to join
      W3C (;<), who wants to take the lead?

      Cheers...

      - Rick
    • Max Dunn
      ... Sounds good, but it also sounds mainly like a natural consequence of XForms, SMIL, and SVG evolving, I don t imagine another effort outside of these. If it
      Message 2 of 20 , Sep 1, 2002
      • 0 Attachment
        > I'm starting to think that there's a need
        > for a parallel spec to SVG to focus on
        > application-level functionality (SVG
        > programmable objects, non-scripting dynamic
        > behaviors that go beyond SMIL
        > stuff, contextual behaviors, prebuilt
        > XForms-like UI components, etc.). But since
        > I can't afford to join W3C (;<), who wants
        > to take the lead?

        Sounds good, but it also sounds mainly like a natural consequence of
        XForms, SMIL, and SVG evolving, I don't imagine another effort outside
        of these. If it is another effort you'd have to clearly draw the line as
        to where it differed from these specs.

        The example case of "decluttering" for example, is great, but should be
        part of SVG/SMIL, not something else.

        I would think trying to get major companies/open source efforts to
        support existing W3C efforts (Microsoft and Macromedia apparently think
        they can ignore SVG, for example, will they ignore XForms as well? Is
        this all theory?) would go a long way towards eliminating the need for
        proliferating new specs.

        Have you provided feedback on the XForms working draft? Review closes on
        9/4.
        http://www.w3.org/TR/2002/WD-xforms-20020821/



        Max
      • josh
        Sunday, September 01, 2002, 4:14:21 PM, Rick wrote: RB In the ideal situation, the user interfaces can RB support dynamic decluttering based on the zoom
        Message 3 of 20 , Sep 1, 2002
        • 0 Attachment
          Sunday, September 01, 2002, 4:14:21 PM, Rick wrote:
          RB> In the ideal situation, the user interfaces can
          RB> support dynamic "decluttering" based on the zoom
          RB> level...at the "50,000 foot level", detailed data
          RB> that might be unreadable or
          RB> unnecessarily "cluttering" would be automatically
          RB> rendered invisible, whereas as the user zoomed in,
          RB> new detail would reveal itself.

          RB> While this could be done with tons of ugly
          RB> scripting, it sure would be a nice feature in core
          RB> SVG - threshold visibility, based on the zoom level,
          RB> which could be done declaratively rather than
          RB> through voluminous scripts.

          RB> - Rick

          I'm sure everyone already knows this, but just in case people have
          forgotten, VRML has a notion of this, usually called LOD
          (level of detail) and it was always a very important part of making
          reasonably optimized scenes.

          - josh
        • neumannandreas
          ... VRML also had the possibility of creating prototypes which worked pretty much like symbols but could contain dynamic code (scripts and variables) - I
          Message 4 of 20 , Sep 2, 2002
          • 0 Attachment
            >
            > I'm sure everyone already knows this, but just in case people have
            > forgotten, VRML has a notion of this, usually called LOD
            > (level of detail) and it was always a very important part of making
            > reasonably optimized scenes.
            >
            > - josh

            VRML also had the possibility of creating "prototypes" which worked
            pretty much like symbols but could contain dynamic code (scripts and
            variables) - I liked VRML prototypes quite much - and think it would
            be a cool feature for SVG as well ...

            Andreas
          • Andreas Streichardt
            ... no...that would really confuse the user;) ... nope...the icons will have the size you specify and won t scale;) ... Andreas Streichardt
            Message 5 of 20 , Sep 2, 2002
            • 0 Attachment
              Tobias Reif wrote:
              > Andreas Streichardt wrote:
              >
              >
              >
              >>>In which situations do the SVGs scale? OnResize? Will there be apps
              >>>
              >>
              >>i am not quite sure what you mean.
              >
              >
              >
              > Given a window containing SVG icons, eg Konqueror showing SVG file
              > preview icons. When I resize the windows to be much smaller; will the
              > icons scale to be smaller as well?

              no...that would really confuse the user;)

              > I'm not sure if I would want them to scale in this particular situation;
              > I'm just curious if there are situations where the scalability of SVG is
              > used in realtime.

              nope...the icons will have the size you specify and won't scale;)

              >
              > Tobi

              Andreas Streichardt
            • Tobias Reif
              ... I don t think so, but like I said, in this situation I m not sure what s better anyways. ... So SVG will be used for icons only, and there is no situation
              Message 6 of 20 , Sep 2, 2002
              • 0 Attachment
                Andreas Streichardt wrote:


                >>Given a window containing SVG icons, eg Konqueror showing SVG file
                >>preview icons. When I resize the windows to be much smaller; will the
                >>icons scale to be smaller as well?
                >>
                > no...that would really confuse the user;)


                I don't think so, but like I said, in this situation I'm not sure what's
                better anyways.


                >>I'm not sure if I would want them to scale in this particular situation;
                >>I'm just curious if there are situations where the scalability of SVG is
                >>used in realtime.
                >>
                > nope...the icons will have the size you specify and won't scale;)


                So SVG will be used for icons only, and there is no situation where they
                scale? Hmm... When I place some on the quick start bar (is that what
                they call kicker?), and I make it double as high, then the SVG icons
                won't be bigger?

                Tobi

                --
                http://www.pinkjuice.com/
              • jabberwock_mo
                ... I agree. Having done some VRML stuff in the past, I was very disappointed in SVG s reuse capabilities. IIRC, VRML prototypes were also parameterizable.
                Message 7 of 20 , Sep 2, 2002
                • 0 Attachment
                  --- In svg-developers@y..., "neumannandreas" <neumann@k...> wrote:
                  > VRML also had the possibility of creating "prototypes" which worked
                  > pretty much like symbols but could contain dynamic code (scripts and
                  > variables) - I liked VRML prototypes quite much - and think it would
                  > be a cool feature for SVG as well ...

                  I agree. Having done some VRML stuff in the past, I was very
                  disappointed in SVG's "reuse" capabilities. IIRC, VRML prototypes
                  were also parameterizable.
                  --
                  John Breen
                • Andreas Streichardt
                  ... of course they will get bigger in that special case;) Andreas Streichardt
                  Message 8 of 20 , Sep 2, 2002
                  • 0 Attachment
                    Tobias Reif wrote:
                    > Andreas Streichardt wrote:
                    >
                    > So SVG will be used for icons only, and there is no situation where they
                    > scale? Hmm... When I place some on the quick start bar (is that what
                    > they call kicker?), and I make it double as high, then the SVG icons
                    > won't be bigger?

                    of course they will get bigger in that special case;)

                    Andreas Streichardt
                  • josh
                    Monday, September 02, 2002, 10:57:43 AM, you wrote: ... j I agree. Having done some VRML stuff in the past, I was very j disappointed in SVG s reuse
                    Message 9 of 20 , Sep 2, 2002
                    • 0 Attachment
                      Monday, September 02, 2002, 10:57:43 AM, you wrote:

                      j> --- In svg-developers@y..., "neumannandreas" <neumann@k...> wrote:
                      >> VRML also had the possibility of creating "prototypes" which worked
                      >> pretty much like symbols but could contain dynamic code (scripts and
                      >> variables) - I liked VRML prototypes quite much - and think it would
                      >> be a cool feature for SVG as well ...

                      j> I agree. Having done some VRML stuff in the past, I was very
                      j> disappointed in SVG's "reuse" capabilities. IIRC, VRML prototypes
                      j> were also parameterizable.
                      j> --
                      j> John Breen


                      You do remember correctly, but it's not too cumbersome to mimic that
                      in SVG. It's the LOD nodes that I think have been overlooked.
                      Wouldn't they be handy in GIS for zoomable maps? I don't know
                      anything about GIS, but it would seem that it'd be a pretty big deal.

                      - josh
                    • neumannandreas
                      ... prototypes ... that ... deal. ... yes LOD would be cool - not only for GIS. If you urgently need it now you can already do it by javascript and dynamic
                      Message 10 of 20 , Sep 3, 2002
                      • 0 Attachment
                        >
                        > j> I agree. Having done some VRML stuff in the past, I was very
                        > j> disappointed in SVG's "reuse" capabilities. IIRC, VRML
                        prototypes
                        > j> were also parameterizable.
                        > j> --
                        > j> John Breen
                        >
                        >
                        > You do remember correctly, but it's not too cumbersome to mimic
                        that
                        > in SVG. It's the LOD nodes that I think have been overlooked.
                        > Wouldn't they be handy in GIS for zoomable maps? I don't know
                        > anything about GIS, but it would seem that it'd be a pretty big
                        deal.
                        >
                        > - josh

                        yes LOD would be cool - not only for GIS. If you urgently need it
                        now you can already do it by javascript and dynamic loading
                        (.getURL(), .parseXML()).

                        Andreas
                      • masterqmann
                        hi andreas and josh since the bounding boxes of lod nodes in vrml are spheres (one center point of which the distance to the avantar is calculated) they are
                        Message 11 of 20 , Sep 3, 2002
                        • 0 Attachment
                          hi andreas and josh

                          since the "bounding boxes" of lod nodes in vrml are spheres (one
                          center point of which the distance to the avantar is calculated) they
                          are useless for maps in 2d (like svg) or vrml without that geovrml
                          extension (not projected maps). so you had to build your on bounding
                          boxes, use
                          proximitysensors and do some scripting. (are i am wrong?)

                          a lod node approach in svg should be something like a cube (thats
                          only my opinion)!

                          something like protos would be really cool. but i think, all this can
                          be done by defining your own entities or dtds. it is just different
                          to vrml.

                          cheers georg
                        • neumannandreas
                          ... (thats ... it should relate to the currentScale Factor or the viewBox (represented as a rect) - I don t see why it needs to be a cube, since we are at 2D
                          Message 12 of 20 , Sep 3, 2002
                          • 0 Attachment
                            > a lod node approach in svg should be something like a cube
                            (thats
                            > only my opinion)!

                            it should relate to the currentScale Factor or the viewBox
                            (represented as a rect) - I don't see why it needs to be a cube,
                            since we are at 2D graphics. But I think the SVG WG is already
                            aware of the LOD problem (at least they discussed it also at
                            SVG.Open).

                            > something like protos would be really cool. but i think, all
                            this can
                            > be done by defining your own entities or dtds. it is just
                            different
                            > to vrml.

                            of course it can be simulated by js/events whatever - but it
                            would be easier to do it with something like protos.

                            I just recently wrote a scrollable selectionLists for a SVG GUI
                            and it required 300+ lines of javascript to be abstract enough
                            for re-use and behave like a similar selectionList on Linux or
                            Windows.

                            So - of course I can implement anything on my own using
                            ECMA-script - but things have to get easier --> to be able to
                            compete with Flash, that already offers GUI-elements within
                            FlashMX.

                            Andreas

                            PS: of course it is not an easy job for the SVG WG to define what
                            is essential enough to serve most people's demands or what would
                            be too specific - and what can be implemented in a reasonable
                            amount of time for the svg viewer programming teams (Adobe/Batik,
                            etc.). A complicated spec without being implemented is useless ...
                          • Jim Ley
                            Tobias Reif wrote in message news:3D74966F.6080506@pinkjuice.com... ... on ... onZoom. If you tag an element with detail for
                            Message 13 of 20 , Sep 3, 2002
                            • 0 Attachment
                              "Tobias Reif" <tobiasreif@...> wrote in message
                              news:3D74966F.6080506@......
                              > neumannandreas wrote:
                              > Does this
                              >
                              > http://www.w3.org/TR/SVG2Reqs/#req-interactivity
                              > "5.4. SVG should provide a mechanism to trigger dynamic content based
                              on
                              > the level of zoom or location of the viewport. SVG may also provide a
                              > mechanism to support a document with elements tagged with level of
                              > detail information (e.g. maps). [SVG 2.0]"
                              >
                              > describe what you need? I'm not sure if it include loading of detail
                              onZoom.

                              If you tag an element with detail for high-zoom that references an
                              element in another document you'd get that wouldn't you? Anyway it's
                              surely to early to speculate exactly the features of such a mechanism
                              beyond what's written.

                              Jim.
                            • Tobias Reif
                              ... Does this http://www.w3.org/TR/SVG2Reqs/#req-interactivity 5.4. SVG should provide a mechanism to trigger dynamic content based on the level of zoom or
                              Message 14 of 20 , Sep 3, 2002
                              • 0 Attachment
                                neumannandreas wrote:


                                > But I think the SVG WG is already
                                > aware of the LOD problem (at least they discussed it also at
                                > SVG.Open).


                                Does this

                                http://www.w3.org/TR/SVG2Reqs/#req-interactivity
                                "5.4. SVG should provide a mechanism to trigger dynamic content based on
                                the level of zoom or location of the viewport. SVG may also provide a
                                mechanism to support a document with elements tagged with level of
                                detail information (e.g. maps). [SVG 2.0]"

                                describe what you need? I'm not sure if it include loading of detail onZoom.

                                Tobi

                                --
                                http://www.pinkjuice.com/
                              • Tobias Reif
                                ... I think it s important to discuss the requirements documents, and discuss the details of the listed features. Tobi -- http://www.pinkjuice.com/
                                Message 15 of 20 , Sep 3, 2002
                                • 0 Attachment
                                  Jim Ley wrote:


                                  > Anyway it's
                                  > surely to early to speculate exactly the features of such a mechanism
                                  > beyond what's written.


                                  I think it's important to discuss the requirements documents, and
                                  discuss the details of the listed features.

                                  Tobi

                                  --
                                  http://www.pinkjuice.com/
                                • masterqmann
                                  ... based on ... a ... yoh that s exactly what could be provided by an event. if the user is zooms in and enters a certain zoomlevel, an event is triggered,
                                  Message 16 of 20 , Sep 3, 2002
                                  • 0 Attachment
                                    > http://www.w3.org/TR/SVG2Reqs/#req-interactivity
                                    > "5.4. SVG should provide a mechanism to trigger dynamic content
                                    based on
                                    > the level of zoom or location of the viewport. SVG may also provide
                                    a
                                    > mechanism to support a document with elements tagged with level of
                                    > detail information (e.g. maps). [SVG 2.0]"

                                    yoh

                                    that s exactly what could be provided by an event. if the user is
                                    zooms in and enters a certain zoomlevel, an event is triggered, which
                                    loads new data for example. the same for zooming out.

                                    with that, something like dynamic reloading could be realized easily.
                                    and if the x y coordinates are recongized, you could realize
                                    something like a map viewer; and that without coding very
                                    much.

                                    is that useful?

                                    cheers georg
                                  Your message has been successfully submitted and would be delivered to recipients shortly.