[Haskell-cafe] Fixity of >>>

[Donald Bruce Stewart]

ross:
> On Fri, Jul 27, 2007 at 12:08:30AM +1000, Donald Bruce Stewart wrote:
> > There's a theory this should work:
> > 
> >     getContents >>= lines >>> map read >>> sum >>> print
> > 
> > But unfortunately we have:
> > 
> >      `(>>=)' [infixl 1] 
> >      `(>>>)' [infixr 1]
> > 
> > Meaning we must write:
> > 
> >     getContents >>= (lines >>> map read >>> sum >>> print)
> > 
> > Indeed, all Arrow ops are infixr. 
> > Are there any technical/compelling reasons for this?
> 
> Not that I can recall.  But it's the precedence rather than the
> associativity that's bothering you here.

Yes, I suppose it is a vote for 0.5 precedence, if they're not to
behave as for >> and >>= ? Which are:

    infixl 1  >>, >>=

So I can write:

    print 'x' >> getChar >>= ord >>> print 

-- Don