Sorry, I didn't mean irrational, I did in fact mean non-finite. But you are right about the hopelessness of actually displaying so many digits. As indicated, I don't in reality intend to do so.
Anyhow, as far as I'm concerned this issue is settled. I'm just using 100 now, and will file a bug on the framework inconsistency. Thanks for your time and help. (Sent from my iPhone.) -- Conrad Shultz On Nov 25, 2011, at 0:45, Jean-Daniel Dupas <devli...@shadowlab.org> wrote: > > Le 25 nov. 2011 à 04:26, Conrad Shultz a écrit : > >> On 11/24/11 3:20 PM, Jean-Daniel Dupas wrote: >>> A formatter is used to convert an internal number representation >>> (integer, floating point, fixed point) into a string. Is has nothing >>> to do with the precision of the represented value. >>> >>> If you use double to do your math, you will get as much precision as >>> double provide, whatever the formatter you use. >> >> I realize that, and I'm not using the formatter per se to do the actual >> math - it uses NSDecimalNumber for that, which works quite well. >> >> I was configuring the formatter for maximal *display precision* of the >> result. Since the input (and significant digits therein) is not known >> in advance, and since the calculations performed are of a kind that will >> never produce irrational numbers, in the context of the application it >> makes sense to simply display all the digits that are available. > > Not all rationals numbers have a finite representation (1÷3), and even with a > big screen and lot of times, you will have a hard time displaying > 100001871036415 digits ;-) > >>> If you managed to prove that such case may exist, why not, but I >>> really don't see how having more than hundred of digits can be useful >>> as there is no internal representation able to represent a decimal >>> with such precision. >> >> Yeah, it's not so much that I want absurd precision, more that I >> generally expect to be able to use the defined type maximum or minimum >> constants whenever my intent is "really big" or "really small" >> respectively. In this case, NSUIntegerMax caused things to break >> silently and in such a way that I thought _I_ had a bug. >> I guess I will go ahead and open a Radar. >> >> (Erk... looks like bugreport.apple.com is down again. Not my lucky day, >> I suppose.) >> > > -- Jean-Daniel > > > > _______________________________________________ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
