[Xy-pic] rotation-invariant arrow-heads?

[Jürgen Koslowski]

Hello,

Since I like to distinguish different types of arrows by means of their
tips, I recently built one-sided tips intended to go on double-shafted
arrows (for lax natural transformations, as opposed to oplax ones):

\newdir{lx}{!/1.5 pt/\dir{}*!(1,.7)\dir{-}*!(-.5,.7):(1,-.3)\dir^{>}}
\newdir{xl}{!/-1.5 pt/\dir{}*!(-1,-.7)\dir{-}*!(-1.3,-.7):(-1,.3)\dir^{>}}

As you can see, the variant going in the other direction is not entirely
symmetric.  But what's worse, these tips fail to work unless the arrows 
are horizontal.  E.g., \xy*{}\ar@{=lx}(20,0)*{}\endxy is fine, but 
replacing (20,0) by (0,-20) leads to desaster.  Obviously, this results
from the offsets being given in absolute terms, rather than relative to
the direction of the arrow.  Unfortunately, I have not been able to come
up with a rotation-invariant way of specifying these tips.  Any ideas?

-- Juergen

-- 
Juergen Koslowski               If I don't see you no more on this world
ITI, TU Braunschweig               I'll meet you on the next one
koslowj at iti.cs.tu-bs.de               and don't be late!
http://www.iti.cs.tu-bs.de/~koslowj      Jimi Hendrix (Voodoo Child, SR)