So let us try to follow the suggestion of Jordi Cabot, who volunteers
to set up a model repository, and see where this leads us.
Everyone that has models available that he would like to share can
then send store these models in the public model repository,
preferably in various formats. The model repository should allow to
specify very clearly with which tool and in which format the models
have been generated. (Which version of UML, which version of XMI,
which modeling tool, and so on.) In addition, it could always be nice
to have, next to this, a human-readable format of the models, for
example as a collection of diagrams in pdf format.
I suggest to link the model repository to the well-known PlanetMDE
website. Once the model repository website is finished, I can ask
Jean-Marie Favre to provide such a link and to make some
advertisement for the model repository.
P.S. Note that I attended the IDM 2006 workshop yesterday, the French
version of MDE, and we discussed the need of such a model repository,
and many persons were very much in favour of, and willing to
contribute to, this repository.
On 27 Jun 2006, at 15:45, Rodolfo Campero wrote:
> 2006/6/27, Greg O'Keefe <email@example.com>:
>> On Sun, 2006-06-25 at 00:40, Islam Essam wrote:
>> > Does any body could help me in finding such a sw UML
>> > design (includes Class Diagram, Sequence Diagram,
>> > State Diagram,...) to test/verify my tool on it???
>> I do not have a model for you Islam, sorry.
>> But I thought I'd grab this opportunity to jump on the soap-box again
>> and say: We need a library of models!
>> This would be point of comparison for different tools and approaches.
>> Some of you might know about TPTP (thousands of problems for theorem
>> provers) which is enormously useful to the automated theorem proving
>> people. Linguists have huge databases ("corpus") of text. The famous
>> model transformation example (classes -> database schema) was really
>> useful in this way.
>> I have considered entering a sample of the toy models found in the
>> formalisation literature into EMF/UML2, which might be a useful
>> beginning. It would be much better to obtain some real-life
>> models from
>> industry though.
>> Anyone interested?
> Me too I consider it would be very useful to have a set of model
> samples, ranging from toy to industrial / real-life complexity.
> I maintain a library for manipulating UML 2.0 models with
> serialization to XMI 2.1, and the lack of samples is a drawback,
> because several problems (mostly related to XMI) are detected by users
> when they try to consume models exported with other tools. Having a
> set of models available would help avoid this kind of problems.
> I also agree with D H. Akehurst:
> regarding that the models should be available in several formats (all
> versions of XMI and other formats) whenever possible.
> To remove yourself from this list please mail puml-list-
> with a message containing the word "unsubscribe".
Received on Wed 28 Jun 2006 - 07:37:19 BST