Blink incorrectly disregards "width" and "height" attributes on SVG <symbol> element |
||||
Issue descriptionChrome Version: (copy from chrome://version) OS: (e.g. Win7, OSX 10.9.5, etc...) What steps will reproduce the problem? (1) Visit https://jsfiddle.net/mca17zyk/ What is the expected result? Small purple rectangle and square. (Each 20px tall.) What happens instead? The square is huge. Please use labels and text to provide additional information. Per SVG2, the square (really, the <use>-generated <svg> element that gets filled with a 100%-sized purple rect) should take its width/height attributes from the <symbol> element that it cloned. That <symbol> element has width=20 height=20, so its <svg> clone should as well. The rectangle (the first purple thing in the testcase) works correctly because it explicitly overrides the width and height by setting those attributes on the use element itself. But if they're unset on <use>, then we should use the ones from the <symbol>. Blink's behavior here would be correct per SVG 1.1: https://www.w3.org/TR/SVG11/struct.html#SymbolElement ...but it's incorrect per SVG 2: https://svgwg.org/svg2-draft/struct.html#SymbolElement The specific commit in the SVG 2 spec repo is here: https://github.com/w3c/svgwg/commit/ca0f76bf83825505e17d8b69ead26b01d3bc75dc#diff-6e6915b15fe92e713aecbcc7786991f6R449 That commit added x,y,width,height to <symbol>, and added prose saying they should have the same effect on clones-of-<symbol> as they have on an <svg> element. Firefox gives "Expected Results" here.
,
Jan 26 2018
,
Jan 26 2018
I filed WebKit/Edge bugs as well, for the record: https://bugs.webkit.org/show_bug.cgi?id=182172 https://developer.microsoft.com/en-us/microsoft-edge/platform/issues/15641029/ (I'm also attaching the testcase just in case the jsfiddle disappears.)
,
Jan 26 2018
Thanks very much.
,
Jan 30 2018
So there's a potential compat issue hiding here (I gathered from reading the external correspondence...) Fairly limited but still.
,
Jan 30 2018
Yeah, this bug report originated in a webcompat.com issue (where the author was originally depending on the Chrome behavior here & hence Firefox was getting a broken rendering -- but they've since fixed their site to not depend on any particular behavior): https://webcompat.com/issues/15076
,
Feb 1 2018
|
||||
►
Sign in to add a comment |
||||
Comment 1 by dholb...@gmail.com
, Jan 26 2018