Re: tables, positioning of `#+Plot:' lines

[Nicolas Goaziou]

Hello,

Mario Frasca <mario@anche.no> writes:

> I was wondering about the position of the `#+plot:' lines.  we have
> a table, and if we want to have formulas, we put these in a `#+TBLFM:'
> line following the table.  the documentation of org-plot states that
> `#+PLOT:' lines are looked for: following, or preceding the table

Then the documentation needs to be fixed.

> but then only the "preceding" strategy is implemented.  I tried just
> out of curiosity, what happens if I put my `#+PLOT:' lines following
> the table but before the `#+TBLFM:' line.  in this case also the
> `#+TBLFM:' line is not found.
>
> I see a couple of problems with this approach and in the current
> implementation:
>
> - only one single `#+TBLFM:' line is recognized,

IIRC, this was a feature request, even though we could allow multiple
TBLFM lines. Why some users requested to have only one active TBLFM line
at a time, I don't know.

> it must follow the table, there cannot be any other meta information
> in between the table and the `#+TBLFM:' line.

What kind of meta information?

You don't need to create monster-tables. You may simply write a table,
and use it as input in a dedicated gnuplot source block.

> my main doubt is that if we acknowledge meta lines at both ends, we
> end up with an unmanageable mess, where similar information can be
> very distant in the document.  I think it was a mistake to let PLOT
> directives be placed elsewhere than formulas, but I don't know if it's
> still worth changing this.

Since TBLFM lines are the odd ones, the mistake was to allow them after
the table, while every other affiliated keyword in the Org universe goes
before the object it applies to.

IOW, I don't think also allowing PLOT affiliated keyword at the end of
the table is a good idea, and I'm sure that allowing it on both ends is
a bad one.

Regards,

-- 
Nicolas Goaziou