[PATCH] Postinstall script ordering in setup - take 3

[Igor Pechtchanski]

On 17 Mar 2003, Robert Collins wrote:

> On Mon, 2003-03-17@07:13, Igor Pechtchanski wrote:
>
> > > And for clarity: my suggested tweak is also not sufficient to provide a
> > > weak ordering.
> > > Rob
> >
> > Rob,
> >
> > Your suggested tweak provides a total ordering.  The "unordered(x,y)"
> > [!(x < y) && !(y < x)] relation is false for any x != y [since either
> > (x < y) or (y < x) always holds].
>
> For my example, yes. The problem is when you introduce z.
>
> >   So all equivalence classes have one
> > element: "unordered(x,x)" is always true.  You get transitivity trivially,
> > as "unordered(x,y) && unordered(y,z)" is only true if x == y == z, and
> > then you also have "unordered(x,z)".
>
> But: we break the ordering transitivity.
> > Transitivity
> > x < y and y < z implies x < z [3]
>
> y:= foo: gam
> x:= gam
> z:= bar: foo
>
> both our operators give
> x < y && y < z && !x < z
> (dep)    (dep)    (alpha)
>
> which isn't transitive.
> Rob

Umm, I already replied to that in
<http://cygwin.com/ml/cygwin-apps/2003-03/msg00504.html>, but I'll repeat
it here:

My operator< *is* transitive, because of the "propagateDependences"
step...

So is yours (since you only changed mine).  Neither is transitive without
that step.  FYI, I couldn't think of a better (linear complexity) way of
providing transitivity.  Suggestions welcome.
	Igor
-- 
				http://cs.nyu.edu/~pechtcha/
      |\      _,,,---,,_		pechtcha@cs.nyu.edu
ZZZzz /,`.-'`'    -.  ;-;;,_		igor@watson.ibm.com
     |,4-  ) )-,_. ,\ (  `'-'		Igor Pechtchanski
    '---''(_/--'  `-'\_) fL	a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

Oh, boy, virtual memory! Now I'm gonna make myself a really *big* RAMdisk!
  -- /usr/games/fortune