Double precision in B3D?

Blitz3D Forums/Blitz3D Programming/Double precision in B3D?

Jimmy

(Posted 2011) [#1]

Is there ways to get double precision in Blitz3D? I noticed
http://www.blitzbasic.co.nz/codearcs/codearcs.php?code=1905#comments
were pointing to some dll doing this but link seem long gone by now

Floyd

(Posted 2011) [#2]

You can find the discussion which started this here.

I've uploaded what I think was the final DLL to my limited free web storage Double DLL.

Note it doesn't allow .zip files, so I have named it blitzdouble.txt, which you will have to rename to blitzdouble.zip to render it usable.

In general I find these attempts at using doubles in Blitz3D to be extremely awkward since can't simply write something like a = b + c. I would go so far as to say that if you really need double precision then Blitz3D is the wrong language.

Jimmy

(Posted 2011) [#3]

Thanks.

In general I find these attempts at using doubles in Blitz3D to be extremely awkward since can't simply write something like a = b + c. I would go so far as to say that if you really need double precision then Blitz3D is the wrong language.

Don't go that far. I find Blitz3D very nice for prototyping. This will fill the void testing algorithms that need plenty bits and precision. This may suit perfectly. I'm not after effeciency now, only the possibility to do accurate testing.

Thanks again, will give this a try.

EDIT I see your point regarding notation, its not ideal. Better than nothing though.

Last edited 2011