Re: [boost] Re: Re: Numerical Integration - continued

From: Neal D. Becker (nbecker_at_[hidden])
Date: 2002-10-23 10:34:49

>>>>> "Martin" == Martin Weiser <weiser_at_[hidden]> writes:
    Martin> Neal, 

    Martin> sorry for the ironic comment above. I figured you mean this situation. 

    Martin> However, the assumption of equidistant integration nodes (constant dx), 
    Martin> and the assumption of known dx (didn't seem to be given as parameter, but 
    Martin> your can easily change that, of course), is quite a specific setup, and 
    Martin> may be totally different from what you'd want to have for integrating 
    Martin> functions you can evaluate wherever you like. 

    Martin> Thus I think these two tasks are different enough to justify different 
    Martin> interfaces. 

I agree (actually, I was trying to say the same thing).