Synopsis - Issue Tracker

Stefan Seefeld via synopsis issue tracker wrote:

> To break the circular dependency the PTree module shouldn't depend on
> Scopes, but Scopes (which I'v tentatively put into the 'SymbolTable'
> module) have to depend on PTree, as that's what Scopes and Symbols operate
> on.

I think that's OK, too

> > There's another question relevant to Scope::lookup. How do you pass
> > N::n<1>::m to it? As you can guess I would not like Scope::lookup to
> > accept Encoding (and dependend on all PTree::). But another issue is that
> > qualified name must be parsed element-wise.
>
> ok, I understand your point, and agree. So we need two 'lookup' methods,
> i.e. lookup_qualified() and lookup_unqualified() (say). 

Hmm.. the papers you've linked say there are some other cases, but I haven't 
looked closure.

> I'm not sure about 
> the Encoding yet. I agree we have to clarify a lot more what Encodings are
> so we can see whether we can enhance the semantics through a better design,
> but the question seems mostly unrelated to the SymbolTable / lookup issues,
> so we can keep the Encoding discussion on task27.

Agreed.