On Thu, May 20, 2010 at 8:58 AM, Brian Schott <[email protected]> wrote:
>        Your comment about spreadsheets being scalar seems
> to off the mark a little to me, because there are plenty of
> spreadsheet formulae that refer to cell ranges, such as
> Excel's SUM(A1:E4).

In that case, you are producing a scalar result.  Granted,
you are using a range expression but when you go examine
that range, it is composed of 20 independent scalars which
you have selected from a larger batch of scalars.

If you want results which are larger than a scalar you either
hand craft each element of the structure or you use copy and
paste.  [That said, the spreadsheet does have support explicitly
designed to aid copy&past programming -- for example A1 vs.
$A$1]

> To me the bigger problem is making a cell aware of its own
> geographic position relative to other cells: the whole cell
> referencing thing.

I think that that would be a bad idea, most of the time.

>        But thinking more about your reply, it occurs to me
> that I should be more specific about the problem I have been
> contemplating writing a small app to solve. The scenario is
> 4 guys playing bridge for 5 hours or so, and changing
> partners after every rubber. The spreadsheet at the
> hyperlink below shows the scheme where columns A, B, C, and
> D are used for each rubber and columns H, I, J, K, and L
> contain running sums of the rubbers, and would typically be
> a separate sheet, not columns on the same sheet with A, B,
> C, and D.
>
> http://www.pixentral.com/show.php?picture=1PClHxSiRzbGytd3ExEn1RZOSvA16
>
>        Henry Rich's bridge scoring J program is very good
> at computing the totals produced in rows 25 and higher from
> inputs entered in rows 2 through 18, once the inputs are
> translated into integers from strings. But how does one
> produce a J program or a J/spreadsheet program combination
> to accomplish this? (If the compactness of the outputs in
> rows 25 and up are a special problem, I would be happy for
> each of those rows to simply have one entry instead of some
> having two entries, and other accommodations for
> simplification would be fine.)

At the moment, my idea of tying J to google spreadsheets is
pure vapor.  I have not been using tables/excel nor tables/tara
but if I had an immediate project I imagine that I would be using
one of them.

-- 
Raul
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to