On Feb 23, 12:28 pm, Chad Perrin <per... / apotheon.com> wrote:
> On Sat, Feb 24, 2007 at 01:11:53AM +0900, benj... / fysh.org wrote:
>
> > What I'd really like to see is a programming environment where the IDE
> > isn't a cunning layer over the underlying reality, but where it's an
> > equally valid interpretation of the data. Then you can work at what ever
> > level you like, and see what ever details you like. This isn't going to
> > happen while we're still storing our programs and data as flat text
> > files though, in my opinion.
>
> I'm not entirely sure what you mean about an editor/IDE providing an
> "equally valid interpretation of the data."  Perhaps you could
> elaborate.

In Microsoft's Enterprise Manager for their SQL Server (and in several
other apps of  theirs), you can create a complex SQL query via a
combination of graphical table joins and filling out information in a
grid. As you do this, you see the SQL statement being fleshed out in
text. The graphical representation drives the SQL.

However, you can ALSO edit the SQL text directly, and (provided you
haven't done something illegal, or that can't be represented) the
graph and grid update when you change the raw code. The SQL drives the
graphical representation.

I'm totally guessing, but I think this is an analog to what the OP is
describing. An IDE that doesn't try to take over the way you work, but
is a valid alternative view of the same underlying code, that doesn't
prevent you from doing raw coding when you want to.