why is there seemingly no support for "float" (32 bit values) type in thrift?

[Gautam Thaker <gt...@verizon.net>]

I am reading/running/modifying the tutorial as  a way to understand
thrift. I noticed that thrift only seems to support a 64 bit float
value, and not a 32 bit float value. There are times when 32 bit values
would be a better choice.

Is this hard to add?

Gautam

Re: why is there seemingly no support for "float" (32 bit values) type in thrift?

[Bryan Duxbury <br...@rapleaf.com>]

It's true, we don't have a 32 bit float. In practice, I don't think that
most people have suffered too heavily from the omission, since floating
point data isn't as common.

I think that in practice it would be very hard to add a new fundamental
datatype to Thrift because there are so many languages and protocols that
would have to be updated accordingly.

-Bryan

On Mon, Jun 27, 2011 at 2:59 PM, Gautam Thaker <gt...@verizon.net> wrote:

> I am reading/running/modifying the tutorial as  a way to understand
> thrift. I noticed that thrift only seems to support a 64 bit float
> value, and not a 32 bit float value. There are times when 32 bit values
> would be a better choice.
>
> Is this hard to add?
>
> Gautam
>