Hi Alex,
True, I would lose that advantage, but I think the main advantage of the
grid approach are the local relationships between neighbors, not access by
name.
Cheers,
Wojtek
pon., 1 lip 2019 o 14:42 Alexander Burger <a...@software-lab.de> napisał(a):
> Hi Wojtek,
>
> > I've noticed a limitation if one were to use
> > multiple grids simultaneously.
> > ...
> > Since the fields in the grid contain symbols named along the pattern of a
> > chessboard that get interned when the grid is created,
>
> Correct. 'grid' is intended for stand-alone usage.
>
>
> > be shared between grids. One way to remove this limitation would be to
> use
> > 'box' for every cell, thus guaranteeing the symbols' uniqueness, but at
> the
> > price of having human-readable symbols like 'a1.
>
> In fact, this is the case if the grid is wider than 26 columns. Then all
> cells
> are created using 'box' (i.e. anonymous symbols).
>
>
> > Another thing that came to my mind when porting this approach to Common
> > Lisp would be to have something like 'box' but with a user-supplied
> prefix.
> > For instance, in Common Lisp there's the function 'gensym', which called
> > without parameters returns a fresh, unique symbol,
>
> In PicoLisp you can create such symbols using 'pack' (i.e. transient
> symbols):
>
: (mapcar
> '((N) (pack "prefix" N))
> (range 3 7) )
> -> ("prefix3" "prefix4" "prefix5" "prefix6" "prefix7")
>
> So you might well just remove the 'intern' in the 'grid' code. But then
> you lose
> the advantage of being able to access the cells by name in the REPL.
>
> Another possibility is to create the grids in separate namespaces and
> switch
> namespaces as needed.
>
> ☺/ A!ex
>
> --
> UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
>
