<krit> RRSAgent: start meeting
<nzimmermann> good :-)
<scribe> Scribenick: AmeliaBR
krit: A reminder that there won't
be an in-person TPAC this year. The TPAC week will be used for
remote plenaray/breakout meetings, but we need to decide if we
also want a meeting.
... We could have meetings with other groups, CSS WG, ARIA WG,
???
AmeliaBR: I don't think there's much benefit of a special SVG meeting. It would just be our regular telcon. But bilateral meetings would be useful, if we need feedback on specific agenda items.
nzimmermann: It would be useful to talk with CSS folk on getting 3D transforms working with SVG better.
krit: We could set up a wiki page
specifically for gathering ideas.
... what was the specific question about 3D transforms?
nzimmermann: I'd have to look up my notes, it's been some months. But there are outstanding bugs in Transforms 2, and fantasai suggested a joint meeting to address them.
krit: Any ARIA issues to bring up?
AmeliaBR: It would probably be useful to discuss progress & how to move forward on SVG AAM, but I'd need to put together a summary.
krit: Any other groups?
AmeliaBR: Depending on how our
testing efforts go, maybe with WPT folks (is that a group?) for
improvements there.
... Maybe web components folks for shadow DOM issues. But
again, need to review & come up with specific topics.
krit: Would you be able to
collect a list of past discussions on shadow DOM, as a
background?
... I'm not sure if there's a specific WG for that, but could
ask they usual people.
AmeliaBR: I could probably get that done by October. Do need to figure out who to talk to. Not sure how Web Platform WG is currently arranged.
krit: I will try to figure out who to talk to.
AmeliaBR: I'm not sure if we have any updates since the past discussion.
krit: Myles was going to look at source code / implementability, but he's not here. nzimmermann, did you have a chance to look at the code?
nzimmermann: No, I didn't.
krit: I know you all had some discussion during I call I couldn't be on. Anything to follow up with that?
<krit> https://github.com/web-platform-tests/wpt
Tav: While I was working on things, one problem was that looking for info on how to create SVG tests, I mostly found out-of-date pages, took quite a while to find the latest guidelines in our wiki.
<Tavmjong> https://github.com/w3c/svgwg/wiki/Testing
krit: The best source for test info is the WPT repo.
Amelia: I think the issue was more about the SVG-specific guidelines.
Tav: The link I posted (↑) is the
most recent info I've found. Although there's one thing I'm not
happy about — the test suite structure says we won't divide the
folders by reftests, scripts, etc., because a testrunner can do
that.
... Now, I'm not sure how to use that tool, but for Inkscape it
is important to get the reftests out so that I can use
them.
krit: So what do you want to change?
Tav: I think it's useful to group
things into folders. Many of the existing tests are already
grouped that way.
... If I go through, people have added lots of tests in the
past 6 months, but I'm not sure what's relevant.
AmeliaBR: Part of the problem is that any organization is difficult to enforce. Tests are getting cross-posted from the browser internal suites, I'm not sure how they get assigned to a particular place?
nzimmermann: So would we need to change the test organization in every vendor?
AmeliaBR: I don't know, not sure how the import process works.
nzimmermann: For imports, are we only talking about reftests? Lots of vendor tests aren't convertible cross-platform.
AmeliaBR: Reftests and JS/DOM tests
krit: There are also manual tests, but I'm not sure that's something we want to promote.
AmeliaBR: Only for interactive things where it's necessary. Otherwise, those are legacy things.
krit: I think we need to emphasize that more clearly.
Tav: For tests that are imported from vendors, is there a review process?
AmeliaBR: They get posted as PR, but are usually automatically approved by bots based on previously reviewed upstream.
Tav: It looks like there is at least guidance about having consistent naming, ref in the name.
krit: That isn't as consistently
enforced, depends on the browser.
... We can maybe ask for WPT folks' advice in the W3C
slack?
AmeliaBR: the #testing IRC channel is also moderately busy
<Tavmjong> https://www.w3.org/Graphics/SVG/Test/svgTest-manual.htm
krit: Tav, can you follow up with trying to find better ways to enforce organization?
Tav: Yes.
... But to go back to the issue of good documentation. This (↑)
is the first link that came up on search. How can we change
this? Who has access?
AmeliaBR: Maybe Chris Lilley?
krit: Maybe. I can get in touch with him. What do you want to change?
Tav: Just a note that it's out of date, with link to new information.
krit: And where should we put the new info. Can we do it on the WPT website?
AmeliaBR: There's a ReadMe in the SVG folder of the repo. Changing means a PR, so don't want to update too often, but we could add a link out to something we can update more readily.
krit: So we could link out to our wiki? And continue that as the formal guidelines.
AmeliaBR: I think so. Once we
have clear guidelines, we could upstream them to make them more
visible/official, but while we're figuring out, best to do it
on our wiki.
... And once we have clear guidelines, we could ask to
contribute an SVG-specific tutorial to the “How to WPT tests”
section on the WPT website.
krit: Yes, generally we want to work with the existing WPT documentation, just add what we need for SVG.
AmeliaBR: Yes, building on the “How to write a RefTest” tutorial they already have.
AmeliaBR: There's some discussion again on the mailing list about what is happening with SVG 2. Need to follow up.
krit: Some is specific about miter and arcs. Tav, you know about that — can you follow up?
Tav: It's really a question of whether implementers are going to accept things & follow up. I can't speak to that.
krit: It looks like someone's done some code, ready for contribution to Skia.
Tav: That's definitely what needs to happen. We need this in the basic rendering engines if we're going to convince browsers to adopt it. I had started a patch for Cairo, but dropped the ball on that.
krit: And what else is there to follow up on?
Amelia: There is the general question of what is the future of SVG 2. Not sure what to say to that.
krit: Yes, it really depends on
whether we get some active editors & tests.
... Any expected agenda items for a meeting in 2 weeks?
AmeliaBR: That really depends on whether I have time to work on the tests.
krit: Likewise for me, haven't had much time for SVG lately.
nzimmermann: Same.
krit: OK, send agenda items or regrets & we'll see if we have things to talk about in 2 weeks, otherwise it will be in 4 weeks.
This is scribe.perl Revision of Date Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Default Present: Tavmjong, krit, AmeliaBR, nzimmermann Present: Tavmjong krit AmeliaBR nzimmermann Found ScribeNick: AmeliaBR Inferring Scribes: AmeliaBR WARNING: No date found! Assuming today. (Hint: Specify the W3C IRC log URL, and the date will be determined from that.) Or specify the date like this: <dbooth> Date: 12 Sep 2002 People with action items: WARNING: IRC log location not specified! (You can ignore this warning if you do not want the generated minutes to contain a link to the original IRC log.)[End of scribe.perl diagnostic output]