Please! Help! Let's get unstuck!

I'd like to beg you all to please re-frame the debate in terms of
starting with xwt.widget and splitting it step by step, justifying
each step.

    http://lists.xwt.org/pipermail/dev/2003-April/002809.html

I further urge you NOT to consider "organizational reasons" a valid
reason for splitting because:

    1) Unlike the other reasons, it is subjective.  We can argue
       forever about subjective things because nobody can ever be
       proven to be wrong (by example or counterexample, etc).  It's
       kinda like arguing about colors or fonts.  We'll never come to
       an agreement.  Give up now.

    2) [The kicker].  We create xwt.lib.backwards_compatability.  If
       we want to split, say, xwt.widget into xwt.widget.foo and
       xwt.widget.bar later on, we add the line

           xwt.overlay("xwt.widget.foo", "xwt.widget")
           xwt.overlay("xwt.widget.bar", "xwt.widget")

       Into a function makeOverlays() in the static block of
       xwt.lib.backwards_compatability.  That way, apps that were
       written for older versions of the library can use the old
       namespaces, but our cvs checkout doesn't have to have all of
       the foo/bar templates littering xwt.widget!

What I'm driving at here is that there is *no penalty* for delaying
the decision to branch the namespace for organizational reasons.  So
let's put off that decision until we absolutely have to make it.

I think this approach will keep us all truly objective, and avoid
overflowing my mailbox.

  - a

_______________________________________________
http://lists.xwt.org/listinfo/dev