SVG Working Group Teleconference -- 15 Dec 2011

<trackbot> Date: 15 December 2011

<ed> trackbot, end telcon

<Tav> On one page it does say 20:00 GMT.

<ChrisL> still no booking for the call?

<ChrisL> flights booked for australia

<ed> 20:00-21:30Z

<ed> GA_SVGWG(SVG1)

<cabanier> what is the passcode?

<ed> cabanier: 26631

<cabanier> i says that the conference is full

<ed> Agenda: http://lists.w3.org/Archives/Public/public-svg-wg/2011OctDec/0107.html

<vhardy> Please note the bridge has changed to 26633 (_not_ 26631 anymore)

<scribe> Scribe: Cameron

<scribe> ScribeNick: heycam

F2F with CSS

VH: there's a UEFA championship happening in that week

… so it's difficult to book travel/hotels

… we've updated the page with hotel information and pricing

… it's not too bad, there are hotels 60-170

… one of the worries is about travel

… we sent a questionnaire to the CSS WG

… wasn't sure if CSS/SVG should use the same questionnaire

… otherwise we could just decide that we want to go one particular way in the group

… the CSS WG is leaning towards changing the date of the meeting, at the same location

<ChrisL> the poll shows just as many people wanting to change the location and keep the date

… we (hosts) would also prefer to keep the same location

… but we could also host in Hamburg on the same dates

CL: can't seem to edit the CSS WG's questionnaire to allow SVG people to respond to it

CM: are there specific dates if we keep the same location?

VH: probably before. after I can't.

<ChrisL> change the location and keep the same dates? has 8 yes and 5 yes-and-prefer

CM: same for me

VH: probably the better option is to do it just one week before, in the same place, or we move it

CL: I can't do it one week before, I'm attending a conference

<ChrisL> libre graphics s the week before

VH: some other CSS people can't do it the week before

<ChrisL> ac meeting is the week after

CL: looking at the results so far from CSS, the third option of "change the location keep the date" seems to be the strongest option

<vhardy> https://www.w3.org/2002/09/wbs/32061/css-2012-05/results

CM: I can't do week after

RC: I have no constraints

VH: we need Chris, and some of the other CSS people who can't make it the week before

… I will give an update early January about the F2F

<cabanier> * Adobe office in Hamburg is very nice

<scribe> ACTION: Vincent to email CSS WG to say that the SVG WG prefers to move the location and keep the same date [recorded in http://www.w3.org/2011/12/15-svg-minutes.html#action01]

<trackbot> Created ACTION-3185 - Email CSS WG to say that the SVG WG prefers to move the location and keep the same date [on Vincent Hardy - due 2011-12-22].

telcon schedule with holidays

ED: the four next telcons scheduled

<ed> 22 Dec, 29 Dec, 5 Jan, 12 Jan

<ChrisL> do we have a f2f agenda for preparation over the holidays?

CC: I cannot make 29 Dec

ED: I can make 22 Dec and 5 Jan

DS: I might be able to make 5 Jan, not sure

ED: so let's cancel 29 Dec, and 12 Jan

… let's keep 22 Dec and 5 Jan on

<ChrisL> my regrets for 22 Dec and 5 Jan

<ed> ACTION: ed to send a reminder to the list about the telcon schedule (cancelling 29 dec, 12 jan) [recorded in http://www.w3.org/2011/12/15-svg-minutes.html#action02]

<trackbot> Created ACTION-3186 - Send a reminder to the list about the telcon schedule (cancelling 29 dec, 12 jan) [on Erik Dahlström - due 2011-12-22].

mini-TPAC

<shepazu> http://lists.w3.org/Archives/Public/public-scriptlib/2011Nov/0000.html

DS: there was a suggestion from some of the chairs that several groups get together and colocate a meeting in Silicon Valley

… at least one chair pushed back on that, saying their group needed more lead time to organise that

… Mike Smith said there was still interest in coordinating script library people

… so there's an idea of doing it in either March or May

… if a WG doesn't want to show up, it doesn't have to, but we might want to coordinate informally

… there's value in us getting together with the script library folks

TB: what do you mean by script libraries?

DS: things like jQuery, dojo, d3, raphael. not sure who exactly who would show up

… but definitely the jQuery folks would

… there's a lot of interest from the script librarians about the Component Model, and I think it would be beneficial to be in the same room to talk about these things

… at TPAC we ran out of time

… I think it would be good for the SVG WG to be in there

CM: what's the proposed duration of the mini TPAC?

DS: let's follow up on the mailing list

Sydney F2F

<shepazu> http://svgsocial.eventbrite.com/

DS: Chris you may be surprised to see you're on the agenda here

CL: ok, that's fine

DS: I thought it would be a good idea to meet up with web developers

… we may have 50-60 people signed up to attend

… it is a free event

… I figure each of us who is interested can do a demo

… we might have a short panel session

… Dmitri Baranovsky might be interested in presenting too

CM: that's the Thursday night

DS: yes

… location TBA

… we should think about what we should present to them

… I was thinking of doing something brief on integrating SVG and HTML

… maybe Erik you can do some SVG Wow stuff

charter

CL: the charter is expected to go back out to the AC this week

<ChrisL> includes requessted edits on SVG DOM refactoring from RIM and Opera

FOSDEM

<ChrisL> http://fosdem.org/2012/

CL: I've put in a talk to FOSDEM at the beginning of February at the Colour Management session

… I'll be talking about the SVG2 Color chapter, and getting feedback from people there

<ChrisL> talk is about color management in svg2. i have provisional acceptance

SVG2 spec editing status

<ed> http://lists.w3.org/Archives/Public/public-svg-wg/2011OctDec/0082.html

CL: I read through the document you (Tav) wrote

… I don't like cid on portions of something

… it should link to an outermost container for a property definition (or whatever)

… if you're using a <dl> as a block container I'd rather see the id on the <dl> itself

… I like the way you've laid it out, the markup is much better

… I agree we should do properties and attributes in a similar way if possible

… but it would be good to have it styled the same way as in CSS specs, for properties

CC: there are two versions of attribute definitions

… I agree with the change proposed by Tav (the linear gradient one)

<cyril> https://svgwg.org/svg2-draft/pservers.html#LinearGradients

<ChrisL> https://svgwg.org/svg2-draft/pservers.html#LinearGradients

<scribe> ACTION: Tav to ask Vincent for feedback on the spec styling [recorded in http://www.w3.org/2011/12/15-svg-minutes.html#action03]

<trackbot> Created ACTION-3187 - Ask Vincent for feedback on the spec styling [on Tavmjong Bah - due 2011-12-22].

SVG2 requirements, continued

<cyril> http://www.w3.org/Graphics/SVG/WG/wiki/SVG2_Requirements_Input#Advanced_pattern_methods

CC: next is Advanced Pattern Methods

CM: Olaf's dot point here is brief

<ChrisL> There was a call for halftone screening as well, could be done as a pattern

… but I remember he mentioned somewhere else about different mathetmatical tiling methods

… there are 12 of them? I can't remember the name.

TB: 13 I think. symmetries. Inkscape supports those.

… another request was for half toning

RC: half toning is not really a pattern

TB: might make more sense in filters

RC: what is the use case for half toning? print or on screen?

TB: on screen

<Tav> http://tavmjong.free.fr/INKSCAPE/MANUAL/html/Tiles-Trace.html

TB: the star is the pattern

… you can change the size, colour, based on the colour underneath

… this actually uses <use>

RC: programmatically in inkscape itself? not in the language?

TB: yes

RC: for half toning, maybe you could have a filter that takes a spot function

… you would feed in the value of the colour, typically you have four spot functions (CMYK)

… you fill in the value of the colour and out comes the region of the cell you will paint

… half toning is like cells with pixels in them

… you tile that, and feed in the colour values into the spot function

<ChrisL> http://qualityinprint.blogspot.com/2009/06/custom-halftone-dots.html

http://tavmjong.free.fr/INKSCAPE/MANUAL/html/Tiles.html

<ChrisL> the dot screens are at different angles because they are opaque (for ink) but that gives rosette pattersn where the dots form clusters

TB: for the tiling, you take a single object or group and inkscape knows the size of it

<ed> http://tavmjong.free.fr/INKSCAPE/MANUAL/html/Tiles-Symmetries.html

… looking at http://tavmjong.free.fr/INKSCAPE/MANUAL/html/Tiles-Symmetries.html

… it just computes the position of where to put the tiles

<ChrisL> rosettes, see http://3.bp.blogspot.com/_lNzlqVMcIj8/Si6ZsNCWLaI/AAAAAAAAAuw/4TnbKSrvtWM/s1600/Birthday+Bitmapped.jpg

… in the first example, the bounding box extends outside the tiling area, so you get overlap

… in inkscape you can adjust the overlap, the "pattern size"

CM: that sets a clip?

TB: the pattern can extend outside

… if you want to pattern something on a rectangular grid, you can choose to do it on a grid smaller than the size of the bounding box

… and then you get overlaps

<cyril> rosettes: http://qualityinprint.blogspot.com/2009/04/rosettes-everything-you-didnt-realize.html

… in some cases, when you're doing hexagonal tiling, if you are not careful you will get content outside the tiling area and you'll get overlaps

… it would be fairly straightforward to specify, not sure how useful it is

… inkscape not only can do those tiling symmetries, see the following pages

… you can add random shifts, shrink or expand the tiles

CC: is there anything in common with the Tiling & Layering work in the Mapping TF?

CL: not really, that's more for multi resolution images

… where you're automatically tiling where they butt together, producing an image by slicing it into multiple pieces

… it's not a repeating pattern

TB: one place where this might be useful is in Mapping

<ChrisL> see http://www.w3.org/Submission/2011/SUBM-SVGTL-20110607/

… if you want a pattern to show a swamp, the density say

… if you could turn a gradient into a density map, add a bit of randomness

<ChrisL> in particular http://www.w3.org/Submission/2011/SUBM-SVGTL-20110607/#Tiling

<cyril> http://www.w3.org/Graphics/SVG/WG/track/issues/2310

CC: the next requirement on the page is also about advanced patterns

<cyril> http://lists.w3.org/Archives/Public/www-svg/2010Feb/0082.html

<cyril> http://www.w3.org/2010/03/01-svg-minutes.html#item06

ED: my email is not something as advanced as these tiling symmetries

… more about tightening up existing patterns

CC: does it make sense to merge these two entries?

CL: I think so

TB: I think they're different

… Erik's is clarification/corrections to the spec

… this current requirement is a new feature

CC: I won't merge then

CM: do you have a sense Tav in how useful these tiling symmetries might be?

TB: not sure it's that useful to include the symmetries

… including screening might be nice, though

… something that can simulate half toning or screening

CM: I'd like to see some brief writeup before accepting it

<scribe> ACTION: Tav to write up a brief summary of what a half toning feature would look like [recorded in http://www.w3.org/2011/12/15-svg-minutes.html#action04]

<trackbot> Created ACTION-3188 - Write up a brief summary of what a half toning feature would look like [on Tavmjong Bah - due 2011-12-22].

<scribe> ACTION: Tav to get clarification from Olaf on what his advanced pattern request is [recorded in http://www.w3.org/2011/12/15-svg-minutes.html#action05]

<trackbot> Created ACTION-3189 - Get clarification from Olaf on what his advanced pattern request is [on Tavmjong Bah - due 2011-12-22].

<ed> http://www.w3.org/Graphics/SVG/WG/wiki/SVG2_Requirements_Mailing_List_Feedback#Video.2Faudio_on_demand

ED: next one is "Video/Audio on Demand"
... we did resolve at TPAC to have video and audio elements simliar to HTML5, or with the same functionality

<ChrisL> same dom or same markup?

CC: we concluded that we wanted to have the HTML5 video/audio elements, unmodified

<ChrisL> because the svg tiny 1.2 ones have better markup but sucky dom

<ChrisL> sigh

<ChrisL> so, no time containers on audio and video then

CM: did you discuss issues of namespaces, or overlap between HTML/SVG elements?

CC: I think it was going to be in the SVG namespace

<cyril> http://www.w3.org/2011/11/04-svg-minutes.html#item12

<cyril> "RESOLUTION: SVG 2 will have a video element in SVG namespace with the same characteristics as the HTML 5 video element"

ED: "same characteristics" doesn't necessarily mean there are no differences

RC: this is for SVG that's not running in a browser?

… because you don't have foreignObject?

CM: in the minutes the question was raised about foreignObject, how it would be different

… and Jen said that IE doesn't do foreignObject

DS: just because we're having at least the same features as HTML's video element doesn't mean we can't go beyond that, right?

CC: I agree with you but I don't think that's what the people from the Media TF wanted

… they wanted a single version of the element

… they wanted us to modify the HTML5 video element if we wanted additional features

… but I shouldn't speak in their stead

DS: ultimately the video element is going to have to change at some point, so we should keep current on any changes… we want it full-featured, but only a single underlying codebase

ED: I think what Olaf is asking for is basically the controls attribute on the audio/video controls, not everything else

CM: olaf's request then doesn't contradict with our plans for audio/video

- DRAFT -

SVG Working Group Teleconference

15 Dec 2011

Attendees

Contents