W3C home > Mailing lists > Public > www-svg@w3.org > April 2003

Re: Marker Behavior

From: Chris Lilley <chris@w3.org>
Date: Mon, 14 Apr 2003 22:11:41 +0200
Message-ID: <110148762718.20030414221141@w3.org>
To: Brian Cuthie <brian@systemix.com>
CC: www-svg@w3.org

On Monday, April 14, 2003, 8:21:04 PM, Brian wrote:

BC> I was wondering how people feel about marker behavior.

BC> Let me start this topic rolling by stating that I'm not happy with the
BC> way markers work right now.

BC> I have two issues:

BC> Problem #1:

BC> As I understand things (documentation and empirical results) markers
BC> inherit attributes from the <defs> part of the tree, rather than at the
BC> point they're used.


BC> The consequence of this is that if you want an arrow
BC> head to match the stroke attribute of the line its applied to, you need
BC> a separate marker for each color line. This seems, well, dumb.  Am I
BC> missing something?

No, you are not missing anything.

It has become clear that sometimes people want that, and sometimes
they want it the other way.

This will be addressed in a future version of the SVG specification.

BC> Problem #2:

BC> Markers are oriented only one way. This means that you need two arrow
BC> head definitions (potentially in a variety of colors -- see above): one
BC> for the start and the other for the end of lines.

BC> Suggestion:

BC> Add an attribute to markers that would allow them to automatically
BC> orient themselves appropriately for use either at the start or end of
BC> lines. Essentially this means that start markers should be automatically
BC> rotated 180 degrees about their RefX and RefY.

BC> So what do other people think? Does anyone else care?

Hmm thats an interesting suggestion.

Currently you would indeed need two arrow definitions although, for
complex markers, you would have one definition, two use elements as
the actual markers, and put a transform on one of them.

 Chris                            mailto:chris@w3.org
Received on Monday, 14 April 2003 16:11:51 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 8 March 2017 09:46:55 UTC