MathGroup Archive 2004

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Re: Re: Reduce/Solve

  • To: mathgroup at
  • Subject: [mg50008] Re: [mg49995] Re: [mg49990] Re: Reduce/Solve
  • From: Andrzej Kozlowski <akoz at>
  • Date: Wed, 11 Aug 2004 05:53:05 -0400 (EDT)
  • References: <> <> <> <>
  • Sender: owner-wri-mathgroup at

On 10 Aug 2004, at 18:28, DrBob wrote:

> *This message was transferred with a trial version of CommuniGate(tm) 
> Pro*
> Andrzej,
>>> I see no reason to modify anything in it. Do you?
> Since you ask...
> All I did was apply the word "valiant" to your defense of Reduce; it's 
> not a disparaging word. The paragraph you repeated below (from your 
> second message in the thread) admitted something might be wrong, too, 
> so of course you were right. You're almost always right, Andrzej.
> This was preceded by a lengthy defense for Reduce's behavior, however, 
> and look at this statement from your previous post in the thread:
>>> In your particular case Solve produces a warning about using Inverse 
>>> functions and tells you that you may not have a complete solution. 
>>> That tells you exactly the reason why Reduce returns the original 
>>> expression back to you.
> "That tells you exactly the reason..."?
> Maybe not. What do you think?
> Bobby

Of course not. Perhaps I should have paid more attention to this 
particular case and less to general principles. But what I wrote was, I 
think, the rational explanation. We are, however, dealing here with a 
bug and  bugs are like acts of God (and I don't mean Stephen W.) which 
defy rationality, at least as far as mortals like myself are concerned.


  • Prev by Date: how to graphically fill the tails of a normal distribution
  • Next by Date: Re: Re: Re: Reduce/Solve
  • Previous by thread: Re: Re: Reduce/Solve
  • Next by thread: Re: Re: Re: Reduce/Solve