I think you had it right about what you wanted from a .Net framework.

And yes, fighting the framework is a complete waste of time.

Which is actually why I used ProMatrix: because I could extend the framework I intended from before I bought it, without having to fight it.

What we ended up with was a specification of data and ui rules keyed to the data structure, and used at runtime to run the application. If you wrote a custom view (in xCase), it would get generated into the right DBC; the ProMatrix DD would be filled with the appropriate (all) information based on the parent fields (overrideable), lookups for FK fields would be keyed to the specified (and inherited if desired) lookup views from the entity; the definition for the List to select records would follow from the Entity, but was overridable: often the purpose in creating a custom view (all the basic views on entities are created for you, parameterized for the PK and one of the Parent PK's, i.e., a separate view for each Parent) was to get a better viewable List to select records, or to create a Find form (whose definition also followed form the Entity but could be overridden). It goes on and on like this: entity triggers, cascading field triggers, expressions to determine whether the current record can be deleted, or a record added, or changes saved. Similarly for fields, plus whether they are visible, etc.

IOW, you do all your thinking outside your code (although you can write code in things like triggers, you are encouraged to encapsulate that in a testable class, and make the call to the class method).

There is quite a bit more, but I think this gives an idea.

What I've maintained is that we should not confuse typing with application development.

In this method, I put the view into the form's datasession, use a builder to wire it up, and I've got a form that has all the business rules I want it to, including all the goodies we expect forms to have. My thinking has been done before the form even existed.

This is what I want to, well as of today will, do with .Net. With the exception of the builders, I've now identified all the pieces I need to put that together, so I'm pretty happy. And since I have been (for the last 11 years) and am the tool writer, I'm pretty sure I won't find myself at odds with the framework. <s>

Hank

>I guess I do not completely understand the approach of "metadata that follows data to encapsulate business logic" (elaboration most welcome)
>
>My VFP knowledge of "metadata" was the DBCX2 approach we used in VFE that very much relied on metadata to drive the view based apps we encouraged. The first question i always taught my VFE students to ask was "Can this be data-driven?"
>
>But the business object always seemed a convenience rather than a hindrance.
>
>In SF the business object seems pretty flexible, especially since the BO Mapper generates the strong typing of the fields. (though I know you find that very strong typing part of the 'ick' factor <g> ) But of course there remains the implicit disconnect between thinking in "objects" and thinking in normalized tables.
>
>There are those of us from a Fox background in the SF community who are always pushing to make the BO mapper more of a metadata repository but the Microfour guys ( also with a Fox background ) are pretty convinced some approaches that made sense in Fox are performance drains in .NET. Though I still look for ways to data drive stuff (not quite as attractive in a compiled language), they are slowly winning me over to playing to the strengths of .NET <s>
>
>And of course the one element that remains true in VFE or SF is to learn to not fight the framework. ( both, of course, have full source code exposed and can be customized however one likes)
>
>I think we will agree though that as Fox developers move to other development platforms they would be well served to explore all the tools ( like Resharper ) and frameworks ( like SF and MM ) that can make the transition if not painless at least efficient and do-able.
>
>Coming from VFE I still miss some of the incredible RAD tools that allowed me to create VFE/SQL apps amazingly quickly, but I am glad the microfour folks have put their focus on the under-the-hood heavy lifting stuff first. Good architecture is really paying off as we look forward to the WPF implementations to come.
>
>Like you I wanted to see VFP .NET succeed as one can never have too many good tools, but I do think the message needs to go out that there really are lots of ways to approach data and there are lots of pluses and minuses to consider. Tool building is sometimes a very different skill-set than tool-using and if one makes a living writing business apps you can really find some return on investment in using tools created by skilled builders who specialize in creating tools.
>
>
>
>>Hi Charles,
>>
>>I've been a big SF supporter (as well as MM.Net). They are both great approaches.
>>
>>They both do good jobs at making what .Net static languages make a real pain, into something that I would call a lesser pain, and they do so very creatively. If there were only static languages available, I would have difficulty choosing which to use; but I would surely, having looked at everything I know about, choose one of them over all the others. They both get down to the business of writing business apps in a fairly similar conceptual manner, although differently in implementation.
>>
>>That similar conceptual manner, the Business Object, was (as everyone here probably already knows) born out of the need to separate business logic from UI. That is a good thing.
>>
>>But a better thing, and the reason I championed (as it were) ProMatrix, IMHO is the use of metadata that follows data to encapsulate business logic. For me, and from what I have heard from those who have used the system, the idea of going back to business objects brings on (speaking for myself) a great ennui, as in, "how could I subject myself to that?"
>>
>>Any why eTec would blow off Boudewijn (who put a lot more work into VFP.Net than I) and me (who was probably their most vocal supporter in the U.S.) is a real mystery. It's one of the things I've been wondering about. A lot. It was hard to let go. I think they might be continuing development on the product, and I wish them luck. I just can't subject myself, and my clients, to the whims of a demonstrably irresponsible organization. There isn't a day I wish it weren't otherwise. Well, actually, I just found the missing piece of the data layer for the new framework (posted it on the framework group), so I'm a little cheerier. <s> I think that after I get this new piece working in Python (which should be a "gimmee" <s>), I might not miss VFP.Net at all. Well, just a little, probably...
>>
>>Hank
>>
>>
>>
>>>>Well, the web site was still up.
>>>>
>>>>Phone calls had been made to the office.
>>>>
>>>>The emails were not returned (and one individual sent it with return receipt, and it was picked up).
>>>>
>>>>If I had only one guess, and it's only a guess, it was "hunker down because we don't have anything good to tell them" behavior, but that's only a guess.
>>>>
>>>>Hank
>>>
>>>since your background is in psychology I share with you this observation : My biggest reservation about the eTech guys in this circumstance is that they do not seem to understand their own best interest. That makes their behavior unpredictable and that bothers me.
>>>
>>>Given your interest and faith in the project, your long track record as being a positive contributor and booster of projects you find worthwhle ( I wonder what Promatrix woud have been without you, for example ) and your profile in the community that should be their target audience, to just blow you off seems dull-witted at best.
>>>
>>>Oh well.
>>>
>>>Hey, come on over to Strataframe, if only for a visit. Les Pinter's there, some guys you know from Promatrix, and old guys like me and Steve Taylor. <g> They have a pretty balanced approach regarding steak over sizzle and a nice community of developers.
>>>
>>>( forgive me if we've discussed this and I've just forgotten but have you looked into the data handling stuff in SF or at their Enterprise Server? )