reintroduce binary exponent float representation - via strconv.ParseFloat

[Dan Kortschak]

What is the likelihood of reintroducing the binary exponent form of
float representation, not via the language proper but via
strconv.ParseFloat?

Currently it is possible to render a float into binary exponent form,
but not to reparse it.

[Rob Pike]

The likelihood improves if you file an issue. Might be above threshold for a proposal, since will require an API change.

-rob

--
You received this message because you are subscribed to the Google Groups "golang-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to golang-dev+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Robert Griesemer]

Please cc me on any such proposal. I'm interested in the details of a canonical representation. The main point of a binary exponent is cheap lossless conversion from decimal to binary and vice versa. Specifically the format should be readable and writable by big.Float.

[Dan Kortschak]

Thanks, will do - simple issue in the first instance (cc'ing gri). I was just checking to see whether that was worthwhile.