Re: List Strawman JC01

["Marin David Condic" <dont.bother.mcondic.auntie.spam@[acm.org>]

I don't have an objection to arbitrarily tacking a name on both ends -
"Front" and "Back" are useful, but still kind of arbitrary. Why is one door
on your house a "Front" door and another the "Back" door? I use the "Back"
door of my house far more frequently than the "Front" door - so in a sense
it is more "Primary". Similarly, "First" and "Last" are kind of arbitrary,
yet still acceptable designations for ends of a doubly linked list. For that
matter, we could use "Left" and "Right" and rather arbitrarily decide that
those concepts will relate in some sense to the scanning of the list as if
it were text. The end you "Push" onto is the "Left" end and the end you
"Enqueue" onto is the "Right" end. Or arbitrarily, we could imagine the list
as vertical and use "Top" and "Bottom".

I think that the end-user of the list is going to be the one to impose some
meaning on the direction or ends of the list. If the application has some
sort of visual presentation then there may be some sense in which the ends
are going to be "Top" or "Bottom" or "Left" or "Right". Pick something and
don't worry too much that it isn't the perfect representation of the
abstract concept. I'd vote for being consistent with the nomenclature in
Ada.Strings.* just so that the end user doesn't have to keep switching
mental contexts when changing the data structures he's working with. There
might even be advantages to using the same enumeration types to avoid
needless proliferation of otherwise identical enumerals.

Or, for Cat In The Hat fans, we could name the ends "Thing 1" and "Thing 2".
:-)

MDC
--
Marin David Condic
Senior Software Engineer
Pace Micro Technology Americas    www.pacemicro.com
Enabling the digital revolution
e-Mail:    marin.condic@pacemicro.com
Web:      http://www.mcondic.com/


"Ted Dennison" <dennison@telepath.com> wrote in message
news:uV4Q7.52774$xS6.86977@www.newsranger.com...
>
> The answer of course is that you *can't*. What we have here is a truly
> direction-neutral structure. We could always go "eeny-meeny-miny-mo" and
pick
> one of course. But just because we did that doesn't suddenly make a
directional
> bias inherent in the data structure. Its something the user imposes, and
its
> there only in as much as the user believes in it and adheres to it.
>
> I'd like to hear from one of the instructors in the audience about this.
After
> all, you are the ones who have to teach students data structures using
this
> thing.
>