Author Topic: Instances of Packages  (Read 599 times)

Paolo F Cantoni

  • EA Guru
  • *****
  • Posts: 5909
  • Karma: +71/-80
  • Inconsistently correct systems DON'T EXIST!
    • View Profile
Instances of Packages
« on: March 19, 2010, 05:22:47 pm »
Here's another of my "left field" questions...

For reasons that I can't disclose at present (but one could consider related to "meta-modelling"), I started playing around with "instances of Packages"...

With appropriate "surgery" I can create such beasts and EA seems to handle them OK.

Can anybody suggest any (conceptual) reason why I shouldn't be able to use them - acknowledging that Sparx can change the universe from under me?

TIA,
Paolo
Inconsistently correct systems DON'T EXIST!
... Therefore, aim for consistency; in the expectation of achieving correctness....
-Semantica-
Helsinki Principle Rules!

g.makulik

  • EA User
  • **
  • Posts: 355
  • Karma: +0/-0
    • View Profile
Re: Instances of Packages
« Reply #1 on: March 20, 2010, 04:24:11 am »
No! No disagree.
But would be interested what kinda "beasts" these are as soon you're able to disclose it. Most likely I could imagine a kind of library or plugin artifacts specialized (instantiated) for particular deployment environments, languages or machine architectures.

VBR
GŁnther
Using EA9.3, UML2.3, C++, linux, my brain, http://makulik.github.com/sttcl/

Paolo F Cantoni

  • EA Guru
  • *****
  • Posts: 5909
  • Karma: +71/-80
  • Inconsistently correct systems DON'T EXIST!
    • View Profile
Re: Instances of Packages
« Reply #2 on: March 20, 2010, 10:05:13 am »
Hi Gunther,

Can you please give the reasons why you disagree?  I'm sure people would disagree, but I'm REALLY interested in why.

(Your imagination isn't too far from where I'm heading....  ;))

TIA<
PAolo
Inconsistently correct systems DON'T EXIST!
... Therefore, aim for consistency; in the expectation of achieving correctness....
-Semantica-
Helsinki Principle Rules!

g.makulik

  • EA User
  • **
  • Posts: 355
  • Karma: +0/-0
    • View Profile
Re: Instances of Packages
« Reply #3 on: March 22, 2010, 05:33:45 am »
That was a misunderstanding Paolo. Of course I should have written more precisely 'No! I don't disagree.'  :)

Have a nice day,
g.

P.S.: Plz remember sometimes: I'm not a native speaker ...
Using EA9.3, UML2.3, C++, linux, my brain, http://makulik.github.com/sttcl/

Simon M

  • EA Administrator
  • EA Guru
  • *****
  • Posts: 6216
  • Karma: +49/-5
    • View Profile
Re: Instances of Packages
« Reply #4 on: March 22, 2010, 08:28:02 am »
Conceptually, they don't inherit from Classifier and are therefore not valid to instantiate.

Of course, if you're happy to work with your own metamodel (instead of UML) you can do what UML does.  Create a new package with a merge relationship back to UML, add your package and an inheritance relationship back to Classifier...
Simon

support@sparxsystems.com

Paolo F Cantoni

  • EA Guru
  • *****
  • Posts: 5909
  • Karma: +71/-80
  • Inconsistently correct systems DON'T EXIST!
    • View Profile
Re: Instances of Packages
« Reply #5 on: March 22, 2010, 10:28:17 am »
Quote
That was a misunderstanding Paolo. Of course I should have written more precisely 'No! I don't disagree.':)

Have a nice day,
g.

P.S.: Plz remember sometimes: I'm not a native speaker ...
Understood...  I will admit I was a little surprised at your negative response...

Paolo
Inconsistently correct systems DON'T EXIST!
... Therefore, aim for consistency; in the expectation of achieving correctness....
-Semantica-
Helsinki Principle Rules!

Paolo F Cantoni

  • EA Guru
  • *****
  • Posts: 5909
  • Karma: +71/-80
  • Inconsistently correct systems DON'T EXIST!
    • View Profile
Re: Instances of Packages
« Reply #6 on: March 22, 2010, 10:34:37 am »
Quote
Conceptually, they don't inherit from Classifier and are therefore not valid to instantiate.
Yes, thanks Simon, I understood that...  That's why I asked the question (although I meant from a generic modelling point of view not UML specific).
Quote
Of course, if you're happy to work with your own metamodel (instead of UML) you can do what UML does.  Create a new package with a merge relationship back to UML, add your package and an inheritance relationship back to Classifier...
Well, I hadn't really thought of formalising it in this way, but that sounds like a good way to go...

(In effect) Is that what Sparx has done with its extensions to UML?

Paolo
Inconsistently correct systems DON'T EXIST!
... Therefore, aim for consistency; in the expectation of achieving correctness....
-Semantica-
Helsinki Principle Rules!

Simon M

  • EA Administrator
  • EA Guru
  • *****
  • Posts: 6216
  • Karma: +49/-5
    • View Profile
Re: Instances of Packages
« Reply #7 on: March 22, 2010, 01:49:38 pm »
I would say that all of our extensions can be expressed in those terms.
Simon

support@sparxsystems.com