Arrow misses target

[jeffreyrosenbluth]

When using a large head size and curvy shaft the head does not point directly at the oriign.

Apparently this has to do with the shaftScale function

[byorgey]

Hmm, yes, that is a bit off, isn't it? Can you post the code you used to generate this image?

[byorgey]

Also, do you get the correct image if you use the iterative version of shaftScale?

[jeffreyrosenbluth]

Yes I do get the correct image with the iterative version. Code is below.

[jeffreyrosenbluth]

{-# LANGUAGE NoMonomorphismRestriction #-}
-----------------------------------------------------------------------------
-- |
-- Module      :  Diagrams.TwoD.Arrowtest
-- Copyright   :  (c) 2013 diagrams-lib team (see LICENSE)
-- License     :  BSD-style (see LICENSE)
-- Maintainer  :  diagrams-discuss@googlegroups.com
--
-- Test module for Arrow.hs and Arrowheads.hs
--
-----------------------------------------------------------------------------

module Main where

import           Diagrams.Prelude
import           Diagrams.Backend.SVG
import           Diagrams.Backend.SVG.CmdLine
import           Control.Lens ((.~), (&), (%~))


-- | example 0 ------------------------------------------------------------
example = d # connect' (with & arrowHead .~ arrowheadSpike (7/16 :: Turn)
                             & arrowTail .~ noTail
                             & headSize .~ 2 & headStyle %~ fc green
                             & arrowShaft .~ s
                             & shaftStyle %~ lw 0.1 # fc blue)
                            "1" "2"

  where
    c = circle 1.5 # showOrigin # lw 0.02
    t = bezier3 (r2 (1,1)) (r2 (1,1)) (r2 (0,2))
    t' = reflectX t
    l = straight unitX
    l' = straight (unitX # rotateBy (1/6))
    s = trailFromSegments [t, l, t', l, t]
    x |-| y = x ||| strutX 2 ||| y
    row1 = (c # named "1") |-| (c # named "2")
    d = row1

main = defaultMain $ ( example # centerXY) # pad 1.1

[jeffreyrosenbluth]

As far as I can tell from looking at the svg, some of the control points on the arrow shaft differ by almost 10%, e.g. 0.66 vs 0.60 between the algebraic and iterative versions.