From: Douglas Gregor (gregod_at_[hidden])
Date: 2003-12-07 12:18:09
On Sunday 07 December 2003 11:37 am, AlisdairM wrote:
> A while back Joel de Guzman pointed out that tuple, variant and optional
> were all variations on a theme. There is a similar relationship between
> boost::array and tuple, where tuple is a container for N heterogenous
> elements, array is the homogeneous couterpart.
> It would be relatively trivial to supply the same discovery interface for
> array as is required for tuple i,e, get<>, tuple_size, and anything else
> I'm missing.
> The fact we have not had to implement this already probably indicates there
> is no real need for it, but there is a pleasing symmetry <g>
> Is this worth pursuing, or am I off in a world of my own again?
Ye might want to look into Joel's "Fusion" (part of Spirit, currently, but I'm
sure it'll come up for review in hte future). Tuples and arrays both fit
nicely into this framework.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk