Re: SH: Integer limits

On 2018-12-12 07:58, Martin Thomson wrote:
> On Wed, Dec 12, 2018 at 5:50 PM Mark Nottingham <mnot@mnot.net> wrote:
>>  From looking at the rest of the spec briefly, it looks like we lose some bitwise and shift operators as a result -- is that correct?
> 
> Yeah, and some of those can be synthesized.  <<1 can become *2.  The
> loss doesn't seem particularly bad. The gain (2^22 times more space)
> seems like a serious upside.
> 
> That this popular language is driving this decision is the only thing
> that bothers me.  Yes, it's a popular language, but it's probably the
> only one that doesn't have native 64-bit numbers (yet --
> https://tc39.github.io/proposal-bigint/).
> 
> ...

In particular as this was one of the reasons to drop the JSON-based 
approach in the first place...

Best regards, Julian

Received on Wednesday, 12 December 2018 07:48:57 UTC