From: Randy Kramer (rhkramer_at_fast.net)
Date: Fri Nov 07 2003 - 08:52:44 EST
On Thursday 06 November 2003 10:13 pm, Martin Sevior wrote:
> On Fri, 2003-11-07 at 12:54, Leonard Rosenthol wrote:
> > Worrying about the text layout engine comes MUCH later...
>
> Well this is where my own bias as a bottom-up developer comes in to
> play. I need to see how the details will work before I can get the big
> picture :-)
Sorry, I want to ask a followup question which takes this subject even further
OT -- feel free to reply directly to me rather than the list to minimize the
OT "noise".
Background: I started programming before top-down / bottom-up was an issue,
lived through the top-down "craze", and have always felt (still do, and
finally recently found someone brave enough to say it in print), that you
need a combination of both. Maybe a partial way to express my feelings is
that top down is good, but you have to know enough from the bottom up
viewpoint to know that your top down approach is implementable. When you
have doubts, you have to work bottom up to determine what is implementable,
perhaps given the restrictions of a particular language. (Ignoring the idea
that many things are implementable (except NP complete problems or similar).
When I hear someone advocate one or the other (like Leonard *seems* to favor
top down by saying "Worrying about the text layout engine comes MUCH later.")
while Martin (more explicity) recognizes the need for bottom up in describing
a perhaps more balanced approach, I wonder what, in their background,
education, experience, or whatever, makes them advocate one approach over the
other.
Any insights appreciated!
Randy Kramer
This archive was generated by hypermail 2.1.4 : Fri Nov 07 2003 - 08:40:47 EST