Re: [freeelks-devel] HASH_TABLE.valid_index

[Emmanuel S. [ES] <ma...@ei...>]

> Isn't this achieved by:
> 
>     ht: HASH_TABLE [ITEM, frozen KEY]
> 
> Hmmm, probably not. But surely by some of the proposals with 
> type intervals.

It is not achieved since what we want to prevent is key instances whose type is
not KEY but a descendant of KEY. I'm not sure if type intervals would be
practical to use.

I'll change the comment and the postcondition.

Thanks,
Manu