[Uml-devel] Heyas

[Nick Bronson]

At 01:29 PM 27/04/2003 +0100, Jonathan Riddell wrote:
>On Sun, 27 Apr 2003, Nick Bronson wrote:
>
> > Hey guys, i'm new to the list. I downloaded 1.1.1 for use in my project and
> > I got to say its pretty good. (Save for the bug which wouldn't let me have
> > overloaded operators).
> >
> > Just thought i'd point out a couple of bugs and make a comment or two ;)
> > Hopefully i'll eventually have time to help out on the development once my
> > honours thesis is finished.
>
>Best way to get round this is to do a dissertation which /is/ on
>developing Umbrello :)

Unfortunately, my topic is now set in stone.  In case anyones interested, 
i'll be designing a web-based system (using java probably, ew.) to allow 
first year university students to write, and compile programs without a 
compiler (in a web browser), and get immediate automated feedback on the 
results.  This will also allow them to participate in a variety of 
excercises that don't necessarily require a whole program to be written 
(say, prewrite a half-program and they need to fill in the rest...)

> > Firstly, I know the bug I mentioned above has been fixed in ver 1.2, but i
> > couldn't track down the actual bug and 1.2 is too unstable to use for me
> > (wouldn't even let me save.)
>
>It should certainly let you save.  Can you explain more about what it
>won't let you do?

Basically it saved for me, however when I went to load the saved file, it 
throws an error.. Something cryptic like "an error occured while loading 
this document."

Co-incidently, I think the same error occured when I saved a file under 
1.1.1 which had duplicate names in it (and hadn't crashed), perhaps its 
related to that? (All the documents I tested with 1.2 had duplicate names, 
that was the impetues behind moving to 1.2 ;)


> > Plus I found that if you go through and create a bunch of classes and then
> > go to document them after the fact, it all works fine except for parameters
> > of operations. For some reason, after its initial creation, it wont let you
> > change the default value or the documentation of a class-function 
> parameter.
>
>Yes, I think that's know about, although I'm not sure if it's in
>bugs.kde.org (you might want to add it if it isn't).

I'll check later.

> > Finally, do you think an option may be added to 1.2 to allow the code
> > generation to ignore aggregations and relationships? I tested out the code
> > generator last night and it adds a pointer to a variable wherever I have a
> > "contains xxx" type relationship, which would be alright except a) I
> > already added that variable in, and b) I wanted it to be a reference, not a
> > pointer.
> > It's an interesting idea, but if different options for its representation
> > aren't possible, could we have a way to turn it off? :)
>
>This is just the sort of thing that hopefully Brian Tomas will be adding.

Coolness.

> > Don't mean to sound like i'm winging, it's an excellent program - i've
> > found it easier to learn and start producing with it than with Rational
> > Rose which I had to learn the year before last.
>
>Cool bananas.
>
>Jonathan Riddell
>
>


Ciao,
Nick Bronson


>-------------------------------------------------------
>This sf.net email is sponsored by:ThinkGeek
>Welcome to geek heaven.
>http://thinkgeek.com/sf
>_______________________________________________
>Uml-devel mailing list
>umbrello-devel at kde.org
>https://mail.kde.org/mailman/listinfo/umbrello-devel