Re: [Scheme-reports] Reformulated numeric-tower ballot
Jay Freeman 29 Apr 2014 06:46 UTC
Voting "yes"for 2, 3, and 4:
For (1), there is a problem, and I shall abstain with the following caveat: The presence of arbitrarily large bignums may cause numerical calculations with numbers that are not integers to generate rationals using said large bignums, which will soon be present throughout the calculation and will likely slow it down a great deal.
I suspect that persons who use Scheme for problems which are intensive with number crunching will thereby be displeased.
Thus I suggest that issue #1 not be addressed without also simultaneously addressing means whereby Scheme users who wish to trade accuracy for speed in numerical work, may do so. Regrettably, this concern may require discussion of such matters as I*** fl**t*ng-p**nt. (Partial censorship self-imposed for the sake of the fastidious.)
And despite John Cowan's sneaky attempts at persuasion :-) , I have not joine scheme-reports-wg2; I am afraid someone might ask me to do some work. So this epistle will not appear there unless someone forwards it.
Jay Reynolds Freeman
-------------------------------
Jay_Reynolds_Freeman@mac.com
http://JayReynoldsFreeman.com (personal web site)
_______________________________________________
Scheme-reports mailing list
Scheme-reports@scheme-reports.org
http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports