[wx-dev] wxDataViewItem::m_id type confusion

[Robert Roebling]

Vadim Zeitlin wrote:

>  Could you please explain what is wxDataViewItem::m_id under different
> platforms? It's declared as pointer

Yes, e.g. for storing user data, which is typically a pointer.

>  but it is used as an integer in wxDataViewModel::Compare(). 

Only as a fallback since (I think) GTK+ doesn't like to
items being equal when comparing.

> Does it really have to be a pointer (for Mac?)?

Everywhere.

> In this case we need to replace "unsigned int" in the API with wxUIntPtr
> as "unsigned int" is not big enough to hold a pointer under Win64

Not quite sure what you mean. The pointer should be able to hold
an int (and it is), where do we need the opposite?

>  There is also something very strange going on with m_hash and m_useHash
> in wxDataViewIndexListModel. The latter is always false under Mac AFAICS
> but why then do we even compile all the code using m_hash there? I'm not
> worried about inefficiency here as much as about code clarity: having 2
> versions of each function isn't really ideal IMO. Why not provide 2
> versions of the (base) class, one using the array and the other one not
> using it?

Because the other platforms flip-flop between using a hash not
not using it at runtime.

>  And as a final question, shouldn't we use wxVector instead of
> WX_DEFINE_ARRAY in this (new) code?

wxVector probably wasn't finished then. I'm also not sure 
what support for sorting we have in wxVector without STL.

 Robert