Stream: t-lang/wg-unsafe-code-guidelines

Topic: integer float bool


gnzlbg (Nov 27 2018 at 17:16, on Zulip):

@avadacatavra I don't think the overview is worth having, it is not as precise as the sections below, which makes it a bit confusing

gnzlbg (Nov 27 2018 at 17:17, on Zulip):

I think it is worth having a section for the "fixed-width integer types", you probably want to spell out that i8 maps to int8_t, etc.

gnzlbg (Nov 27 2018 at 17:18, on Zulip):

I also think that it is worth adding a section for floating point types, in particular, it is worth mentioning that when it comes to Layout, the different revisions of the IEEE standard do not touch it, so it does not make sense to say that they are IEEE754:2008 or IEEE754:2018 compatible, because the layout of the 32 and 64 bit wide floating point types did not change

gnzlbg (Nov 27 2018 at 17:19, on Zulip):

the section should however mention which of the floating-point formats specified in IEEE754 the f32 and f64 types have

gnzlbg (Nov 27 2018 at 17:20, on Zulip):

for f32 that's binary32 and for f64 that's binary64

gnzlbg (Nov 27 2018 at 17:21, on Zulip):

I don't know if we should hyperlink wikipedia here, but this link gives an overview: https://en.wikipedia.org/wiki/IEEE_754#Representation_and_encoding_in_memory (for anything else we can hyperlink the IEEE754:2008 standard or the latest draft of the IEEE754:2018 standard)

Last update: Nov 19 2019 at 17:35UTC