Author tim.peters
Recipients Serge Anuchin, mark.dickinson, pitrou, r.david.murray, rhettinger, serhiy.storchaka, skrah, steven.daprano, tim.peters, vstinner
Date 2015-07-12.21:42:44
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1436737364.93.0.565828370088.issue24567@psf.upfronthosting.co.za>
In-reply-to
Content
[Serhiy Storchaka]
> ... I want to say that double rounding causes not
> only bias from ideal distribution, but a difference
> between platforms

That's so, but not too surprising.  On those platforms users will see differences between "primitive" floating addition and multiplication (etc) operations too.

There's a tradeoff here.  In Python's earliest days, the bias was strongly in favor of letting platform C quirks shine through.  Partly to make Python-C interoperability easiest, but at least as much because hiding cross-platform fp quirks is a major effort and even Guido had only finite time ;-)  As the years go by, the bias has been switching in favor of hiding platform quirks.  I hope Python 3 continues in that vein, but Python 2 probably needs to stay pretty much where it is.
History
Date User Action Args
2015-07-12 21:42:45tim.peterssetrecipients: + tim.peters, rhettinger, mark.dickinson, pitrou, vstinner, steven.daprano, r.david.murray, skrah, serhiy.storchaka, Serge Anuchin
2015-07-12 21:42:44tim.peterssetmessageid: <1436737364.93.0.565828370088.issue24567@psf.upfronthosting.co.za>
2015-07-12 21:42:44tim.peterslinkissue24567 messages
2015-07-12 21:42:44tim.peterscreate