Meeting Minutes SVG WG from 2012/10/04

Hi SVG folks,

Here are the meeting minutes from the SVG telconf 2012/10/04: http://www.w3.org/2012/10/04-svg-minutes.html

Greetings,
Dirk

   [1]W3C

      [1] http://www.w3.org/

                               - DRAFT -

                    SVG Working Group Teleconference

04 Oct 2012

   See also: [2]IRC log

      [2] http://www.w3.org/2012/10/04-svg-irc

Attendees

   Present
          birtles, [IPcaller], heycam, cabanier, krit, ed,
          +33.9.53.77.aaaa, Tav, Rich

   Regrets
   Chair
          Cameron

   Scribe
          krit

Contents

     * [3]Topics
         1. [4]Telcom time
         2. [5]mask-type keywords on the mask-type property
         3. [6]Dmitry's feedback on SVG
         4. [7]SVG2
     * [8]Summary of Action Items
     __________________________________________________________

   <trackbot> Date: 04 October 2012

   <heycam> ScribeNick: krit

Telcom time

   heycam: time shoifts in australia
   ... europe in 3-4 weeks
   ... japan doesn't
   ... i suggest satying with the current time till all changed
   their time

   cabanier: nothing changes?

   heycam: for the moment

mask-type keywords on the mask-type property

   <heycam>
   [9]http://www.w3.org/mid/ED32692D-189D-47EE-B610-46EA0B8A07E6@a
   dobe.com

      [9] http://www.w3.org/mid/ED32692D-189D-47EE-B610-46EA0B8A07E6@adobe.com

   heycam: we need to allow these keywords for mask references
   inside the property
   ... parsers might not know if the reference is a mask or image
   ... it seems not to make sense to overwrite the settings of the
   property mask-type on a mask
   ... but doing nothing seems worst
   ... so we should define what these keywords do on the mask
   proiperty

   krit: my take is that mask element should know better if it
   allows luminance and alpha or just one

   heycam: I would think so as well, but overwriting, as suggested
   by Brian, seems to make more sense in general

   krit: that might be that a luminance mask, with mask: ref
   alpha, would result in a totally masked element

   heycam: yes, might be
   ... the author should know better.
   ... ignoring seems worst, and it seems not consistent

   krit: maybe we want to add auto to mask-type

   heycam: ah, ok. I still would suggest not to add auto and let
   mask property override the mask-type property for consistency
   on mask type values

   birtles: it seems not entirely unpossible that you get the
   result to the result

   heycam: is it possible to add auto to mask property as well?

   cabanier: there must be a computed value

   heycam: yes, but it matters what it gets calculated to

   cabanier: how is it done in other places?

   heycam: birtles: me thinking that as well

   cabanier: don't think that we have a switch in other places
   dependent on the content

   krit: me neither

   heycam: so we don't have an auto value yet, so no problem now
   ... so don't introduce it
   ... yes, if the mask type value is present, then it will
   override the mask-type property of the mask element

   resolution: if the mask type value on the mask property is
   present, then it will override the mask-type property of the
   mask element

Dmitry's feedback on SVG

   <heycam>
   [10]http://lists.w3.org/Archives/Public/public-svg-wg/2012AprJu
   n/0006.html

     [10] http://lists.w3.org/Archives/Public/public-svg-wg/2012AprJun/0006.html

   <birtles>
   [11]http://raphaeljs.com/reference.html#Element.getBBox

     [11] http://raphaeljs.com/reference.html#Element.getBBox

   getBBox with transforms applied, and x2,y2 as bottom/right

   krit: I would encapsle the element in a g element and get the
   transform from this

   <heycam>
   [12]http://www.w3.org/Graphics/SVG/WG/wiki/SVG2_Requirements_In
   put#Make_it_possible_to_get_the_bounding_box_of_an_element_in_a
   _particular_coordinate_system

     [12] http://www.w3.org/Graphics/SVG/WG/wiki/SVG2_Requirements_Input#Make_it_possible_to_get_the_bounding_box_of_an_element_in_a_particular_coordinate_system

   heycam: I think we discussed it before. You needed a parent
   element and get the transformation of the current element to
   this parent element

   <heycam>
   [13]http://www.w3.org/Graphics/SVG/WG/wiki/Proposals/getBBoxOf

     [13] http://www.w3.org/Graphics/SVG/WG/wiki/Proposals/getBBoxOf

   birtles: this suggestion, to get the bbox in another coordinate
   system, does it transform the points or axis aligned rectangle
   in the target coordinate spece

   heycam: the later one
   ... the alternative, in gecko we have a transformBound function
   ... the matrix you want to transfrom the points by.

   cabanier: it is easy to do it yourself

   krit: yes, but a lot of graphic JS library does it. So why not
   put it in the browser
   ... is it exposed? how does it work?

   heycam: no, we use it internally

   krit: it might be good to do this together with the matrix4x4
   proposal

   heycam: status of matrix4x4?

   krit: no one is wokring on it at the moment

   heycam: if we had a transformBound function, would it still be
   necessary getBBoxOf?
   ... if we had a transformBound function, would it still be
   necessary to have getBBoxOf?
   ... you could do getCTM of the target element, but like
   explicit methods
   ... or current element
   ... or get both transforms and calculate the transformation
   between the two CTMs

   <ed> you'd probably use the getTransformToElement function (in
   svg 1.1)

   <ed>
   [14]http://www.w3.org/TR/SVG11/types.html#__svg__SVGLocatable__
   getTransformToElement

     [14] http://www.w3.org/TR/SVG11/types.html#__svg__SVGLocatable__getTransformToElement

   ed: you would use getTransformToElement function

   heycam: I thought in addition to this function
   ... or as alternative getBBoxIn
   ... might be more existing with current bbox system

   cabanier: getBBoxOf might be more useful

   krit: I am happy with getBBoxX, but wouldn't go further

   birtles: document stuff makes any sense to support as iframe

   <birtles> I wonder if it makes sense to be able to use the
   coordinate space of elements in a seamless iframe

   heycam: you might be right
   ... there is a lot of things that doesn't make sense in
   multiple documents
   ... that is a general issue that we solve once we come to it
   ... or SVG fragments into an HTML document
   ... you can define sth that works
   ... bounding box has the defintion that it does not include the
   transformation, maybe a different name?

   krit: bounding box has the defintion that it does not include
   the transformation, maybe a different name?

   heycam: well, it still makes sense
   ... if you call getBBox on the parent, it still takes the
   transformation of children into account
   ... so it does not always ignore transforms

   krit: you are right

   resolution: Add getBBoxOf to SVGLocatable interface

   heycam: should i come up with a proposal for transformBound?

   krit: I think it should be part of a matrix4x4

   birtles: it doesn't matter to much
   ... it can move later anyway

   <heycam> ISSUE: Consider having a transformBounds method either
   on SVGMatrix or CSSMatrix4x4

   <trackbot> Created ISSUE-2445 - Consider having a
   transformBounds method either on SVGMatrix or CSSMatrix4x4 ;
   please complete additional details at
   [15]http://www.w3.org/Graphics/SVG/WG/track/issues/2445/edit .

     [15] http://www.w3.org/Graphics/SVG/WG/track/issues/2445/edit

   heycam: we can put it on SVGMatrix and make an inheritance
   hierachy and put the stuff into a new matrix proposal
   ... SVGMatrix would inherit from the new definiton

   <scribe> ACTION: krit will talk to Dean Jackson how to continue
   with matrix4x4 proposal [recorded in
   [16]http://www.w3.org/2012/10/04-svg-minutes.html#action01]

   <trackbot> Sorry, couldn't find krit. You can review and
   register nicknames at
   <[17]http://www.w3.org/Graphics/SVG/WG/track/users>.

     [17] http://www.w3.org/Graphics/SVG/WG/track/users%3E.

   <scribe> ACTION: Dirk will talk to Dean Jackson how to continue
   with matrix4x4 proposal [recorded in
   [18]http://www.w3.org/2012/10/04-svg-minutes.html#action02]

   <trackbot> Created ACTION-3389 - Will talk to Dean Jackson how
   to continue with matrix4x4 proposal [on Dirk Schulze - due
   2012-10-11].

   heycam: the transformation proposal

   krit: I would like to add bug reports and try to address some
   of these in the next level of transforms
   ... CSS Transforms

   heycam: I am no so sure about the shorthands
   ... I am no so sure about the short-hands
   ... I am no so sure about the shortands

   krit: I would suggest to discuss it on the FX TF

   heycam: yeah, just think that shorthand syntax does not fit
   very well to CSS Transforms
   ... We know about the API issues and should come up with a
   proposal

   krit: We still wait for CSSOM

   <heycam>
   [19]https://svgwg.org/svg2-draft/coords.html#InterfaceSVGTransf
   ormList

     [19] https://svgwg.org/svg2-draft/coords.html#InterfaceSVGTransformList

   <heycam> see the getter and setter on there

   heycam: added setter and getter to SVGTransformList so you can
   do transform[$1\47] and get the second entry of the list
   ... transform[2, newItem]

   <heycam> transform[$1\47] = newItem

   krit: ah, functionality great, the defiintion in the IDL seems
   strange
   ... method isPointInside

   <heycam>
   [20]https://svgwg.org/svg2-draft/types.html#InterfaceSVGGeometr
   yElement

     [20] https://svgwg.org/svg2-draft/types.html#InterfaceSVGGeometryElement

   heycam: I wonder if isPointInFill handles it already

   cabanier: does isPointInFill also work on group elements?

   heycam: no, it is stock to SVGGeomentryElement

   krit: getElementsByPoint gives you all elements under a certain
   point, right?

   <ed>
   [21]http://www.w3.org/TR/SVG11/struct.html#__svg__SVGSVGElement
   __getIntersectionList

     [21] http://www.w3.org/TR/SVG11/struct.html#__svg__SVGSVGElement__getIntersectionList

   heycam: I think getIntersectionList does that

   krit: I think there was a problem is that for
   getIntersectionList just gives you the elements under the
   current SVGSVGElement

   heycam: take the root SVG element
   ... but me be good to have a function that just takes a point,
   not only a rect
   ... do we want to allow it on any element and not just
   SVGSVGElement?

   krit: you have a reference element

   heycam: right, but you might expect to put it directly on the g
   ... you don't want to transform your point to the CTM of the
   element that you are interested in
   ... I think would be nice to have it on the element itself

   krit: does it mean that the point also needs to be in the CTM
   of the current element?

   heycam: yes

   krit: and what about the intersectionlist?

   heycam: well I guess you already have your point in this CTM
   ... can we have a isPointInElement like Dmitry suggests beside
   isPointInFill and isPointInStroke

   <heycam> isPointInElement

   <heycam> have that on any SVGLocatable

   krit: does not tell if it is on bbox or visible par

   t

   <ed> we do have pointer-events="boundingbox"

   heycam: it depends on pointer events

   krit: but you might want it independet of this property

   <ed> (well, at least it's in tiny 1.2)

   krit: I still think you might want to have this functoin
   independent of the current setting in pointer events

   heycam: or we have an extra argument that takes a pointer-event
   value

   <heycam> isPointInElement(mySVGPoint, "boundingBox")

   heycam: and can override the settings of the element, just for
   this method

   <heycam> isPointInElement(mySVGPoint, "all")

   heycam: this argument would be optional

   ed: I think I prefer what you suggested earlier
   ... I don't know if it's a good idea to inherit the pointer
   event to the last child
   ... I would like to see more detailed proposal even if I like
   to have this functionality

   <scribe> ACTION: Cameron to come up with a proposal for
   isPointInElement and pointer events [recorded in
   [22]http://www.w3.org/2012/10/04-svg-minutes.html#action03]

   <trackbot> Created ACTION-3390 - Come up with a proposal for
   isPointInElement and pointer events [on Cameron McCormack - due
   2012-10-11].

   heycam: getElementsByPoint seems to be interessting
   ... since you often don't want to have a rectangle

   <scribe> ACTION: Cameron Will also investigate in
   getElementsByPoint [recorded in
   [23]http://www.w3.org/2012/10/04-svg-minutes.html#action04]

   <trackbot> Created ACTION-3391 - Will also investigate in
   getElementsByPoint [on Cameron McCormack - due 2012-10-11].

   krit: maybe we should discuss the other methods on the mailing
   list

   heycam: maybe

   <scribe> ACTION: Dirk will start discussion to other proposed
   methods [recorded in
   [24]http://www.w3.org/2012/10/04-svg-minutes.html#action05]

   <trackbot> Created ACTION-3392 - Will start discussion to other
   proposed methods [on Dirk Schulze - due 2012-10-11].

   <heycam>
   [25]http://lists.w3.org/Archives/Public/public-svg-wg/2012JulSe
   p/0214.html

     [25] http://lists.w3.org/Archives/Public/public-svg-wg/2012JulSep/0214.html

SVG2

   SVG2!!!!!!

   heycam: keep it SVG 2 but allow SVG2 as well outside the spec

Summary of Action Items

   [NEW] ACTION: Cameron to come up with a proposal for
   isPointInElement and pointer events [recorded in
   [26]http://www.w3.org/2012/10/04-svg-minutes.html#action03]
   [NEW] ACTION: Cameron Will also investigate in
   getElementsByPoint [recorded in
   [27]http://www.w3.org/2012/10/04-svg-minutes.html#action04]
   [NEW] ACTION: Dirk will start discussion to other proposed
   methods [recorded in
   [28]http://www.w3.org/2012/10/04-svg-minutes.html#action05]
   [NEW] ACTION: Dirk will talk to Dean Jackson how to continue
   with matrix4x4 proposal [recorded in
   [29]http://www.w3.org/2012/10/04-svg-minutes.html#action02]
   [NEW] ACTION: krit will talk to Dean Jackson how to continue
   with matrix4x4 proposal [recorded in
   [30]http://www.w3.org/2012/10/04-svg-minutes.html#action01]

   [End of minutes]
     __________________________________________________________


    Minutes formatted by David Booth's [31]scribe.perl version
    1.137 ([32]CVS log)
    $Date: 2012/10/04 22:32:49 $
     __________________________________________________________

     [31] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm
     [32] http://dev.w3.org/cvsweb/2002/scribe/

Scribe.perl diagnostic output

   [Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.137  of Date: 2012/09/20 20:19:01
Check for newer version at [33]http://dev.w3.org/cvsweb/~checkout~/2002/
scribe/

     [33] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/there/their/
Succeeded: s/axis alive/axis aligned/
Succeeded: s/rect angle/rectangle/
Succeeded: s/it be/it still be necessary/
Succeeded: s/havegetBBoxOf/have getBBoxOf/
Succeeded: s/I don't want/I don't know if it's a good idea/
Found ScribeNick: krit
Inferring Scribes: krit
Default Present: birtles, [IPcaller], heycam, cabanier, krit, ed, +33.9.
53.77.aaaa, Tav, Rich
Present: birtles [IPcaller] heycam cabanier krit ed +33.9.53.77.aaaa Tav
 Rich
Found Date: 04 Oct 2012
Guessing minutes URL: [34]http://www.w3.org/2012/10/04-svg-minutes.html
People with action items: cameron dirk discussion krit start talk will

     [34] http://www.w3.org/2012/10/04-svg-minutes.html

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.



   End of [35]scribe.perl diagnostic output]

     [35] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm

Received on Thursday, 4 October 2012 22:35:55 UTC