Re: [gmt-dev] evolving GMT in a fully open process that allows others to

["Jorn Bettin" <jorn.bettin@xxxxxxxxxxxxxxxx>]

> > > With respect to "getting going in 5 minutes" Fuutje is actually quite
> good,
> > > however it's a different story when it comes to implementing major
> changes
> > > to the tool model, i.e. when you want to work with a completely
> different
> > > meta model. This area needs significant work.
> >
> > ... and this is typically the main usage scenario, I guess.
> >
> I do not think that this is the main usage scenario at all. The current
tool
> model works quite well for me for some 5 years now - with some minor
> modification now and then.

Don't understand, see below.

> What may be needed more often in a layered MDSD world that we intend to
have
> for GMT, is the use of FUUT-je as a tool-factory: You model your (meta)
> model in FUUT-je, generate and develop an application from it, and use the
> result as your domain-generative development tool.

That contradicts what you say above, where you dispute that creating a meta
model for a DSL and then generating a modelling tool for the DSL is the main
usage scenario. All that we are discussing is taking baggage out of the tool
model underlying Fuut-je. This will lead to a more elegant tool factory
(process of building tools). The meta model (tool model) that I'm suggesting
works for MetaEdit, something very similar works for LANSA RUOM, and I'd be
surprised if we come across good tools that use something very different.
And - I think - the Fuutje tool model is not far away from the suggested
meta model anyway.

> I used this to develop an
> HTML/PHP menu system, and earlier to generate complicated
> lifecycle-state-on-many-levels code, EJB's from specific XML structures, a
> database layer from truly odd Rose diagrams, etc.

There you go ;-)

Jorn

Jorn Bettin
jorn.bettin@xxxxxxxxxxxxxxxx
www.softmetaware.com
Tel  +64 9 372 3073 | Mobile +64 27 448 3507 | Fax +64 9 372 3534