Re: [Sml-implementers] A list of ambiguities in SML?

[Andreas R. <ros...@ps...>]

Martin Elsman wrote:
> 
> I think it would be great with a maintained list. I believe Andreas
> Rossberg's work is a good starting point:
> 
>   http://www.ps.uni-sb.de/Papers/abstracts/sml-defects.html
> 
> His list does not mention the problem about "too large" real
> constants, however.

I can keep the issues list up-to-date if there is interest. So far, I'm 
aware of very few new issues, though - the list should probably mention 
the broken structure sharing. If anybody knows about other issues that 
are not listed please let me know and I will incorporate them. Likewise 
corrections to listed issues.

With respect to the "too large" real constants, I'm not sure. I don't 
think there really is an issue with the Definition. The wording in 
section E.1 is a bit fuzzy, but probably as precise as it can be, 
considering that the Definition does not specify anything about 
representations at all - it does not even say that reals have to be 
floating point. Maybe the wording of the Definition should rather be 
refined by the Basis spec?

	- Andreas