search

certik / yaml-cpp

Automatically exported from code.google.com/p/yaml-cpp
MIT License
0 stars 0 forks source link

unable to round trip floating point numbers #272

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
The code used to control the precision of floating point numbers is too 
restrictive, the limit should be defined more like this

(2 + std::numeric_limits<float>::digits * 3010/10000)

see

http://www2.open-std.org/JTC1/SC22/WG21/docs/papers/2005/n1822.pdf

not sure if the default should be changed, for consistency I would

Original issue reported on code.google.com by Kevin.J....@gmail.com on 30 Jan 2015 at 1:57

Attachments: