Subj : Re: Implementing A* algorithm
To   : comp.programming
From : CBFalconer
Date : Wed Jul 20 2005 09:29 am

"Arthur J. O'Dwyer" wrote:
> On Mon, 18 Jul 2005, CBFalconer wrote:
>> Christer Ericson wrote:
>>> cbfalconer@yahoo.com says...
>>>
>>>> You have automatically excluded all replies from those not
>>>> familiar with A* (whatever that may be) and/or Dijkstra's
>>>> (who was a remarkably able programmer and probably published
>>>> thousands of algorithms during his lifetime).
> 
> (That's a good thing.  However, I think you're wrong, in that he
> did /not/ apparently manage to exclude replies from you.)
> 
>>> There's nothing wrong with the OP's question.  There is only
>>> one algorithm of Dijkstra's known as "Dijkstra's algorithm"
>>> and it is taught in many computer science programs. See:
>>>
>>> http://en.wikipedia.org/wiki/Dijkstra's_algorithm
>>
>> Well, on casual examination of my books here, I immediately find
>> Dijkstras algorithm as an extension to N processes of Dekkers
>> algorithm for mutual exclusion.  I can also find his solution to
>> the Dining Philosophers problem.
> 
> So what? Newton developed hundreds of methods, Fermat wrote down
> dozens of little theorems, and Duff probably invented several
> different devices. Does that mean that we should flaunt our
> ignorance every time someone mentions Newton's method, Fermat's
> Little Theorem, or Duff's device? (I say not.)
> 
>>> The A* algorithm is taught in just about every introductory
>>> AI class in the world:
>>>
>>> http://en.wikipedia.org/wiki/A-star_algorithm
>>>
>>> I don't see how you think that anyone unfamiliar with either
>>> or both algorithms would be able to meaningfully comment on the
>>> merits of the one algorithm over the other.
>>
>> I never took an introductory AI class, and I gravely doubt that I
>> ever will.  Since the original problem (IIRC, you snipped any
>> evidence of it) showed similarities to traveling salesmen, I
>> suspect it requires some sort of exhaustive search.
> 
> Come on!  The OP wrote, and I quote: "I have a robot that has to
> find the shortest route through a maze. Options for doing this are
> either the A* or Dijkstra's algorithm."

I had to leave an awful lot unsnipped to make this point.  What you
just requoted was snipped when I read it.  The references did not
exist.

If I say "the counters are 2*421 encoded" does that immediately
bring forth a specific implementation to you?  There is lots of
jargon, and the presence of a clue or two does not significantly
degrade an article.

At any rate, there is no point arguing over this.  I presented my
view, and nobody really has to agree with it.  Of course failure to
agree points out other characteristics of the failer, such as .... 
:-)

-- 
"If you want to post a followup via groups.google.com, don't use
 the broken "Reply" link at the bottom of the article.  Click on 
 "show options" at the top of the article, then click on the 
 "Reply" at the bottom of the article headers." - Keith Thompson

.