SVG Working Group Teleconference -- 18 Apr 2013

<trackbot> Date: 18 April 2013

<scribe> scribenick: cabanier

being the telecon 30 min earlier

krit: fine with me

heycam: we should start that next week

… can we resolve that

all: sounds good

resolution: we will start telecon 30min earlier (1:30PT)

<scribe> ACTION: heycam to change the telecon bridge time changed [recorded in http://www.w3.org/2013/04/18-svg-minutes.html#action01]

<trackbot> Created ACTION-3487 - Change the telecon bridge time changed [on Cameron McCormack - due 2013-04-25].

<birtles> http://www.timeanddate.com/worldclock/fixedtime.html?month=04&day=25&year=2013&hour=20&min=30&sec=0&p1=0

meeting at TPAC

heycam: we already decided we would do that

<heycam> http://lists.w3.org/Archives/Public/public-svg-wg/2013AprJun/0063.html

… for our third f2f. there was a note sent to the chairs

… in this email. one note was if we wanted a 3 day meeting instead of a 2 day one

… my feeling is that 2 days is cramped especially if we meet other team

krit: an extra day will likely conflict

heycam: I'm not sure when the extra day is

krit: we have too many conflicts with other working groups

heycam: yes, I noticed that too. Tokyo and TPAC are taken up with joint meeting with CSS

… we should consider that for later meetings

… I can ask the organizers when this extra day would be

… monday-tuesday is wg meetings at tpac

… third day would be thursday or friday. This might overlap with other groups

krit: let's ask first and decide afterward

shepazu: we could resolve that the 3 days if there are no conflicts otherwise we'll have 2

krit: that's fine with me

heycam: css is going to be the big one. We will want joint and separate time

shepazu: also webapps

heycam: is the feeling that we don't want the third day if it overlaps other groups?

… we don't know how we're going to be scheduled.

shepazu: you can ask what their plans are.

… we can ask for 3 days and see if we can use them

<scribe> ACTION: heycam to respond to the TPAC questionnaire [recorded in http://www.w3.org/2013/04/18-svg-minutes.html#action02]

<trackbot> Created ACTION-3488 - Respond to the TPAC questionnaire [on Cameron McCormack - due 2013-04-25].

heycam: we need to say what groups we want to join up with

shepazu: it would be good to meet with CSS and HTML

responsibilities for the test suite

krit: ???

heycam: what we discussed was splitting the spec in sections and assign it to people

krit: I will take masking and filters

heycam: we need to make sure that we have tests for new stuff

… I think that was blocked on me assigning this to people

… also I need to sort out the shepard installation

… we can check in test to the repository without the setup though

… I will move that action higher on my todo list

… does anything else need to be done?

krit: just write the tests

heycam: I think we want the existing tests in the right shape

… I also want to write a script to convert the test to the new template that works with shepard

… it might be good if I do that as well

… and add it to the repository and then assign it to people

krit: we have tests in webkit and mozilla.. Maybe we can reuse them in the new testsuite

heycam: are you planning to move the SVG 1.1 masking and filters tests to the FX repository where the specs live

krit: we can leave them in SVG for now

… at some point they will be in shepard so they will be all in the same repository

heycam: do you already have some tests?

krit: yes

heycam: I will do the migration of the test and then mail out the plan

what to do with features with SVG1.1 that are not interoperable or implement

krit: for SVG2 we want to follow the W3C spec

… do we already have a list of these missing features

heycam: you want to see a list of features?

krit: if there is a feature that is not implemented by 2 user agents?

heycam: you can mark it as at-risk and remove it.

cabanier: This is a 'should' not a 'must'?

shepazu: what are you saying?

heycam: should we drop non-interoperable or missing features?

thomas: from boeing, text-link is something that we were wondering why it was never implemented

… we are wondering what would replace it

<ed> textLength?

… it is of critical importance to us

<ed> that's implemented by more than 2 user agents AFAIK

shepazu: that's an interesting case.

… we can write tests

… so people support it and correctly

… I know Cameron wants to support it for Firefox

heycam: there is no opposition.

thomas: we need to control the exact position of text elements

heycam: I wonder if the advent of web fonts negates the problem

… it might be interesting see what problem you're trying to solve

thomas: I need to be able to control where the test is render

… a wiring diagram with a wire and you put a label. You need the text to be right in the middle.

shepazu: maybe connectors will solve the problem?

heycam: we could decide if certain features won't make the cut

… especially if otherwise we have to write test to drop them

… we can discuss particular features

… is that OK with you dirk?

… if we know that there are features that we know ahead of time will be cut, we can remove

<ed> http://dev.w3.org/SVG/profiles/1.1F2/test/svg/text-text-01-b.svg - test data: passed by 4 independent implementations - http://dev.w3.org/SVG/profiles/1.1F2/test/status/implementation_matrix.html

<ed> that's for textLength

Test the web forward attendees / topics

krit: is someone attending?

cabanier: I am

krit: I will be there too

ed: maybe

<heycam> I will be leaving on Thursday night, so I won't be there

krit: we expect a lot of participants

birtles: I can't go. A couple from mozilla Japan will go too

krit: one of the topics will be filter effects.

cabanier: there will be a blending topic as well

birtles: is there any information?

krit: Google is hosting and we're organizing

ed: it would be good to mail the list as an announcement

… do we need an action to make sure the info is on the site?

krit: no

ed: any other agenda requests?

krit: I changed the style of filter effects and masking for elements

ed: do we want this for the SVG spec?

krit: It's harder to find consensus. Noone agrees so we can leave it for now.

ed: that's fine

… can we publish the filter spec soon?

krit: there are some issues with mask that affect filters as well

… I'm a bit lost with the concepts.

cabanier: those are different security issues, right? It's is CORS

krit: how you fetch resources is an issue

… do you need more info?

ed: if you want to ask the css group again, that's fine

krit: I'm trying to create a wiki with Anne's and Roc's remarks

… it's very confusing

… I will keep the discussion on the mailing list for now

technical documentation

shepazu: many people have said that they want to use SVG instead of WebCGM

… is this of interest to you?

thomas: we've been using CGM in the past and going forward we did a study

… WebCGM was rejected because it's half baked and unstable while SVG is more consistent

… it solves 90% of the problem

… the enduser will see SVG in our world

shepazu: it's interesting that there's a market where SVG is suited

… it's not much of a stretch to go to 100%

… I would love to see what use cases are missing and how we can meet those

thomas: I would love to do that

shepazu: I would to bring up connectors

… there is no format to represent graphs and I would like to make it a first class citizen of the web

… non-scaling strokes are in the spec now

krit: Webkit, firefox and opera have them

shepazu: are there other things?

thomas: I mentioned text-length

… that let's you specify exact text layout

shepazu: Web fonts let you pick an exact font

… but it doesn't let you do distortions

cabanier: he is not asking for distortions

thomas: what we currently do, is export with text-lenght and scale later so it matches the length

<heycam> Firefox bug for textLength - https://bugzilla.mozilla.org/show_bug.cgi?id=569722

… if the spec was implemented, we would not have a problem

ed: I think this is just an implementation bug

heycam: is this because there's a fundamental problem?

thomas: MS and Mozilla told me that it was not important enough

heycam: it's good to hear from people

krit: is this feature what you actually want?

thomas: from reading, textlenght will do what we need it to do

krit: in that case, you should bring it up with the different browser vendors

shepazu: we can make a use case document and have all the tests

… Thomas, if you can do that, it will help us push that feature

<scribe> ACTION: Thomas to write to write a test suite for text length [recorded in http://www.w3.org/2013/04/18-svg-minutes.html#action03]

<trackbot> Created ACTION-3489 - Write to write a test suite for text length [on Thomas Smailus - due 2013-04-25].

heycam: I think we have 1 test

<heycam> This is the only test for textLength / lengthAdjust, I think: http://www.w3.org/Graphics/SVG/Test/20110816/harness/htmlObjectApproved/text-text-01-b.html

shepazu: you need to port the test to the new format. You should also create a use case document

… maybe it should also be a CSS feature?

krit: I think that will be hard to do.

shepazu: I agree that putting it in SVG will be far quicker

… Thomas, is there a community group for technical documentation?

thomas: I don't know

shepazu: maybe we can connect with people from the WebCGM community

… and see if other people want this feature

thomas: another thing is that WebCGM is a plugin. Our customers are very reluctant to rely on plugins.

shepazu: are there other features that are missing?

thomas: let me do a look and see if there are other ones

* cameron, can you send the minutes out? *

- DRAFT -

SVG Working Group Teleconference

18 Apr 2013

Attendees

Contents