Bug 14051 - <svg:image> fails to position correctly when <svg> is inside an inline <div>
Summary: <svg:image> fails to position correctly when <svg> is inside an inline <div>
Alias: None
Product: WebKit
Classification: Unclassified
Component: SVG (show other bugs)
Version: 523.x (Safari 3)
Hardware: Mac OS X 10.4
: P2 Normal
Assignee: Nobody
Depends on: 12207
  Show dependency treegraph
Reported: 2007-06-09 17:31 PDT by Antoine Quint
Modified: 2007-06-16 23:33 PDT (History)
0 users

See Also:

Testcase (1.05 KB, application/xhtml+xml)
2007-06-09 17:31 PDT, Antoine Quint
no flags Details
First attempt (20.09 KB, patch)
2007-06-10 23:06 PDT, Rob Buis
eric: review-
Details | Formatted Diff | Diff
slightly simpler test case (we should land other one however) (435 bytes, application/xhtml+xml)
2007-06-11 06:51 PDT, Eric Seidel (no email)
no flags Details
First attempt (46.37 KB, patch)
2007-06-14 17:45 PDT, Rob Buis
eric: review+
Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Antoine Quint 2007-06-09 17:31:23 PDT
When positioned within a CSS inline block, it seems an SVG-rendered image element will not position itself appropriately, and thus not render on screen. This is not the case for absolutely-positioned elements or CSS-positioned elements with display: block.
Comment 1 Antoine Quint 2007-06-09 17:31:49 PDT
Created attachment 14917 [details]
Comment 2 Dave Hyatt 2007-06-09 17:35:11 PDT
inline-block means something else.  These divs have been turned into straight-up inlines, so they're like spans.  Just clarifying so that whoever investigates this doesn't get confused.
Comment 3 Rob Buis 2007-06-10 23:06:53 PDT
Created attachment 14928 [details]
First attempt

Discussed with Sam, probably worth having this at least for the testcase.

Comment 4 Eric Seidel (no email) 2007-06-11 06:29:07 PDT
Comment on attachment 14928 [details]
First attempt

This is a nice fix, but for the wrong bug.  This is a fix for bug 14064.
Comment 5 Eric Seidel (no email) 2007-06-11 06:51:13 PDT
Created attachment 14934 [details]
slightly simpler test case (we should land other one however)
Comment 6 Eric Seidel (no email) 2007-06-11 06:56:55 PDT
This is yet another bug which will be made easier by fixing 12207.  The only class in SVG which should ever have to care about tx/ty is SVGRenderRoot.  All the rest should just be able to expect to be drawn relative to their SVG viewport, ignoring any html translations.
Comment 7 Rob Buis 2007-06-14 17:45:03 PDT
Created attachment 15037 [details]
First attempt

This should do it.

Comment 8 Eric Seidel (no email) 2007-06-14 18:15:45 PDT
Comment on attachment 15037 [details]
First attempt

looks great.  leave the parentx, parenty param names out so no one else is tempted to use them.  or ASSERT that they're == 0.
Comment 9 Rob Buis 2007-06-16 23:33:27 PDT
Landed in r23544.