IRC log of svg on 2016-06-22

Timestamps are in UTC.

07:25:39 [RRSAgent]
RRSAgent has joined #svg
07:25:39 [RRSAgent]
logging to http://www.w3.org/2016/06/22-svg-irc
07:25:41 [trackbot]
RRSAgent, make logs public
07:25:43 [trackbot]
Zakim, this will be GA_SVGWG
07:25:43 [Zakim]
ok, trackbot
07:25:44 [trackbot]
Meeting: SVG Working Group Teleconference
07:25:44 [trackbot]
Date: 22 June 2016
07:26:30 [nikos]
chair: Nikos
07:26:32 [nikos]
scribe: nikos
07:26:34 [nikos]
scribenick: nikos
07:28:27 [nikos]
present+ nikos
07:28:32 [nikos]
present+ AmeliaBR
07:28:35 [nikos]
present+ Tav
07:29:18 [nikos]
Meeting: Amsterdam editors meeting 2016 Day 3
07:29:51 [nikos]
TabAtkins: You didn't camelCase meshGradient ... was that on purpose? ;)
07:30:42 [Tav]
Tav has joined #svg
07:43:12 [stakagi]
present+ stakagi
07:43:19 [nikos]
Topic: Mesh gradient rendering
07:45:11 [nikos]
AmeliaBR: I prefer treating this just like a path, so fill-rule has an effect and the rendering is clipped to the outer path (which means the back side will not show)
07:45:43 [nikos]
Tav: I quite like the idea of overflow as a control
07:45:48 [nikos]
... whether the back faces are shown or not
07:47:10 [nikos]
... If you want to be able to stroke the path, I would like to implement using the same code we use to draw and fill paths, which means it would be clipped
07:48:00 [nikos]
nikos: The important thing imo is to be able to render in the usual mesh gradient way - with back faces showing
07:49:26 [nikos]
... I don't mind having a control. I would suggest overflow should be the default because that matches other implementations of mesh gradients
07:49:56 [nikos]
... whether you have overflow enabled or not, you'll still be able to stroke and add markers to the 'outer' path (which might not be the actual outermost path)
07:51:56 [nikos]
nikos: We might want to call it something other than overflow, because we're kind of overloading the term here. It's not a viewport that's being clipped and there'll never be scrollbars
07:55:53 [nikos]
AmeliaBR: We're going to want fill rule to control what path is clipped to
08:00:37 [nikos]
nikos: I don't think fill-rule is very important. It's not going to affect what is stroked
08:00:52 [nikos]
AmeliaBR: But it will control whether backfaces are filled or not
08:02:41 [nikos]
... we can define as always nonzero for now and add evenodd later if authors request
08:04:13 [nikos]
RESOLUTION: Mesh gradients will have some sort of overflow control that defines whether any path exists outside of the author defined 'outer' path (which will be the equivalent path). Painting will be performed with fill-rule equal to nonzero.
08:51:52 [AmeliaBR]
AmeliaBR has joined #svg
08:52:04 [AmeliaBR]
Regarding https://github.com/w3c/svgwg/issues/26
08:52:22 [AmeliaBR]
Test case for SVG links inside links: http://jsbin.com/nidixomute/1/edit?html,output
08:53:10 [AmeliaBR]
Currently, Chrome and Safari do not render the content inside the nested link (the orange ellipse is visible, no blue link to the GitHub issue)
08:56:27 [AmeliaBR]
Firefox and Edge do render & link the blue ellipse correctly
08:56:40 [AmeliaBR]
("correctly" being how we'd like to spec SVG 2)
08:57:34 [AmeliaBR]
(Technically, Chrome & WebKit are correct by SVG 1.1: don't render SVG elements that are in places they don't belong.)
11:43:50 [nikos]
Topic: Gradient transforms
11:44:07 [nikos]
https://github.com/w3c/svgwg/issues/135
11:44:23 [nikos]
Tav: The gradient transform is supposed to be applied last
11:44:48 [nikos]
... that is, it is supposed to be post multiplied but it says 'inserted to the right of' any previously defined transformations
11:44:53 [nikos]
... which in matrix notation is actually the opposite
11:44:59 [nikos]
... the transform matrix on the left gets applied last
11:45:31 [nikos]
AmeliaBR: Can we get rid of left and right and just say post or pre multiplied?
13:03:07 [AmeliaBR]
AmeliaBR has joined #svg
14:55:17 [nikos]
RRSAgent, make minutes
14:55:17 [RRSAgent]
I have made the request to generate http://www.w3.org/2016/06/22-svg-minutes.html nikos
14:56:46 [stakagi]
Thanks for your hard work!!
14:57:28 [nikos]
Thanks Takagi-san. It's such a shame you couldn't join us .
14:58:18 [stakagi]
I almost feel better!
14:58:45 [stakagi]
Thank you for your concern.
14:59:10 [nikos]
I'm very glad to hear that
15:26:06 [TabAtkins]
nikos: ...yes? We agreed not to camelcase any new element names ever again.
15:26:30 [nikos]
My recollection was things could be camelcased for consistency with camelcased names
15:26:38 [nikos]
meshGradient fits with linearGradient, etc
15:28:08 [TabAtkins]
That violates the entire reason for avoiding new camelcase in the first place...
15:28:56 [nikos]
yeah but it gives people a chance to guess what they should type
15:29:07 [TabAtkins]
"But just one more, we promise!" isn't very convincing. ^_^
15:29:24 [nikos]
maybe we should call it gradientmesh ;)
15:29:34 [TabAtkins]
I'd rather swap the names around somehow, yeah.
15:30:01 [TabAtkins]
Or like, <mesh> and <meshcolors>
15:30:08 [nikos]
so mesh is the geometry element
15:30:17 [nikos]
we separated the geometry element and the paint server element
15:30:27 [nikos]
cos having to switch interfaces, etc was messy
15:30:32 [TabAtkins]
Yeah, I got that.
15:31:22 [TabAtkins]
(Can <mesh> specify its geometry on its own, or is it required to get that from a <meshgradient>?
15:31:23 [nikos]
I'd be happy enough calling it gradientmesh - A Gradient Mesh (as opposed to a mesh based gradient) is a specific implementation, but ours satisfies that now we have smoothing
15:31:25 [TabAtkins]
)
15:32:42 [nikos]
mesh has a meshGradient child, in the same way that rect can have a paintserver child now
15:33:43 [Tav]
Tav has joined #svg
15:33:43 [TabAtkins]
Right. I'm asking if that's *required*, or if it can specify its shape on its own and, like, fill itself with just "red" or something.
15:35:01 [nikos]
no the meshgradient is required now. I would like to expand to flat fill later though
15:35:16 [TabAtkins]
Okay, cool, I was just wondering.
15:37:23 [AmeliaBR]
You can still have a path, circle, polygon, etc that uses meshGradient as a paint server. But a mesh without a meshGradient is specced to act like a path without a d attribute.
15:45:21 [Tav]
Tav has joined #svg
15:47:23 [TabAtkins]
Yup, that all makes sense. Cool.
17:23:59 [Tav]
Tav has joined #svg
18:08:54 [Tav]
Tav has joined #svg
20:19:41 [shepazu]
shepazu has joined #svg
20:58:28 [Tav]
Tav has joined #svg
22:21:10 [shepazu]
shepazu has joined #svg