[Issue 323] New: Error: need opCmp for class Bar

[xs0]

Sean Kelly wrote:
> xs0 wrote:
> Ah good point.  So any ideas?  Personally, I'd be happy if the default 
> AA used a linked-list for chaining and opEquals for comparison so this 
> wasn't an issue, but that seems a foregone conclusion.  I'm mostly 
> concerned about providing default behavior that makes sense for the 
> common case.  The problem I've run into is that I have objects which 
> need to be used as keys in an AA but are not inherently sortable.

Well, one possibly good solution would be to templatize AAs and remove 
opCmp from Object. That way, you'd get an error when using comparison on 
objects without opCmp (most other operators work that way) and the 
template could test for its presence and use different implementations 
for each.

Another nice consequence would be the possibility of having optimized 
implementations for specific types; most probably even a single 
implementation could be considerably faster, because the compiler would 
know the exact type at compile-time, producing an optimized version of 
code for each.


xs0