UML

[Christopher Browne]

On Fri, 27 Aug 1999 17:30:56 +0100, Alan Gauld
<alan.gauld at gssec.bt.co.uk> wrote: 
>> > Nutshell books are reference books. I tried that one as well and was
>> > disappointed too.
>> 
>> I didn't like _UML in a Nutshell_ either, it's the worst O'Reilly
>> book I've ever read.
>
>Its not even a good reference! It misses several bits out
>(cf coverage of Activity diagrams)

I have not yet heard any good report on that book.

>I prefer Instant UML from Wrox press - but it does have 
>a slightly limited view of UML, and tries to explain UML 
>in terms of UML which may not suit everyone!
>
>Fowler's book gets the best reviews so far.

The Instant UML book has a (seemingly unique) section on mapping UML
to IDL, which strikes me as being one way in which to make sure that
it is actually useful.

Automating a particular diagramming notation isn't that useful if
you're still left with having to write up manually how it corresponds
to the actual code.

As near as I can tell, the real value in having UML would be in
diagramming the relationships between the objects whose formal
interfaces would be outlined in IDL.  

There don't seem to be many tools yet that provide that
correspondence, perhaps unfortunately...

-- 
A student, in hopes of understanding the Lambda-nature, came to Greenblatt.
As they spoke a Multics system hacker walked by.  "Is it true", asked the
student, "that PL-1 has many of the same data types as Lisp?"  Almost before
the student had finished his question, Greenblatt shouted, "FOO!", and hit
the student with a stick.
cbbrowne at ntlug.org- <http://www.hex.net/~cbbrowne/lsf.html>