Rational-Rose and Python ...

Warren Postma embed at geocities.com
Wed Mar 29 12:41:37 EST 2000


> It sounds like the modelling wasn't your primary problem on that project.

Yes, probably. However, I find most new technologies, in attempting to
overcome a developer's resistance to trying something new, end up oversold.
A nice thing about Python, IMHO, is that because it's a language to which I
have the source code, it's not oversold. If it can't do something, I can
make it do something [like Christian Tismer's Stackless Python].

I think the same would be true for UML.  The Tigris stuff looks exciting for
the same reason. If there's something my Case tool won't do, I can go in and
add it, send a patch, and it might even get put into the program.  In other
words, the latest ROSE doesn't do much for me, even with it's much enhanced
two-way tools, but an open source UML tool might end up being just as handy
as Python itself. Even better if it worked from inside
Python/PythonWin/IDLE/etc, and was truly "compact" in much the same way as
Python is.


Warren






More information about the Python-list mailing list