[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: (FC-Devel) Meta-Case tools and mutual interests



Although I am sympathetic to L.s objections, I am left wondering what is a
reasonable enterprise for the FreeCASE community, which is another way of
asking if we have anything to contribute.

I hesitate asking that question, because as I jump in, someone is bound to
reference the existing material on the FreeCASE pages.  This is a fair
admonition, but I must confess I didn't discern it readily.

Anyway, here is my prejudice.  The core(theoretical)problem with CASE tools
is that they don't recognize the importance of matching the language of the
semantic model to that of the application or system domain. Some language
elements are of course, pervasive and generic, and thus the great similarity
among many modeling (especially OO) methodologies.  But CASE tools generally
offer just one notation (albiet flexible and extendable) and force the
developer to impliciti jump the gap between the CASE notation and the
particulars of the application.

(This is the thinking that lay behind DoME, which is perhaps best described
as a meta-modeling toolkit.  And although for rhetorical reasons we
sometimes describe DoME as a CASE tool, it has several obvious limitations,
especially on the process enactment side, and perhaps on the generator side
as well. I'm therefore not suggesting that DoME solves the FreeCASE problem,
and I only entered this discussion to suggest that it may be a useful
component, and maybe even a tool, in which you can prototype FreeCASE or
capture some of its requirements.)

Tom


-----Original Message-----
From: Lindsay.Marshall@newcastle.ac.uk
[mailto:Lindsay.Marshall@newcastle.ac.uk]
Sent: Thursday, February 18, 1999 10:16 AM
To: dav@entel.kiev.ua
Cc: Edman_Tom@htc.honeywell.com; freecase-devel@freecase.seul.org
Subject: Re: (FC-Devel) Meta-Case tools and mutual interests



> So you think that we are unable to create a framework (not all part of
that
> framework) of software development environment ? I'm not going to produce
> all features right now -- just to give a way to expand/integrate different
> modules.

I didn't say that we are unable to createa framework for a software
development environment. However, I will say it now - I don't think we
can create a framework for a software development environment. Not one
that will actually be useful. Not one that a majority of people will
want to use. Lots of other people have tried to do this and failed
miserably why should we be any different? In fact as an amorphous group
we stand far less chance than people who have tried to do it
commerecially or as a research project.

> 
>> see absolutely
>> no hope of us producing anything useful if we set our sites so far off

Sorry about the typo/spello - I meant "sights" of course

>> that we the result is over the horizon. 
> 
> Ok. So we we should develop yet another case tool ? Is there nothing
> similar here ? 

Isn't that what we are trying to do? It's FreeCase, not
FreeSoftwareDevlopmentEnvironment. A case tool would be useful for me,
but I dont want *ANYONE* telling me about other development tools I
have to use. All SDEs I have come across restrict your choices so much
as to be unusable.

> OSF has a lot of thing already done -- drawing tools, object/etc
databases,
> wordprocessors etc. Why not to use those work and create an system of
interfaces
> between different and pre-exisiting modules just to make a consistent
software
> development environment ? 

Um, where are these tools? Are they free? Can I get the source? For
instance I am not aware of a (usable) free wordprocessor that I can get
the source for. If they meet all the requirements lets use them. But I
wont hold my breath. 

Sounds to me like you want to have an IPSE. Remember IPSEs? 15 years ago
they were all the rage and lots of people were going to change the
world with them. They dont work for a whole lot of reasons, not least
of which is that people who are forced to use them tend to hate them!

What's wrong with ed + make + troff anyway ? :-)

L.
-- 
http://catless.ncl.ac.uk/Lindsay