Hi,
Actually, the last 2 blocks (about bad practice and mixing) were questions. Let
me try to explain (read carefully): Lets say that you've created a certain
profile (MyProfile) as an extension of the sysml profile (use generalizations
towards sysml element). Now you apply that profile on one of the packages
within your sysml model (MyModel). Can you then still claim that your model is
compliant with sysml specifications? The profiling mechanism is defined within
the specification (chapter 17).
A) If the answer is NO,
1: Does this mean that every profile you create must have 2 parts? Part 1 being
SysML4MyProfile and part 2 being the actual extensions/generalizations you
make. Otherwise you can never use stereotypes defined in sysml but not defined
in your profile if you wish to apply MyProfile.
2: how do you deal with profiled imports?
If you're using a parts
catalog on which for example the SYSMOD-profile is applied, can you
still use these parts (eg. typing a block, generalize to a
SYSMOD-element or even associate with it)? Or do I need to apply the
SYSMOD-profile to the entire package?
B) If the answer is YES, but once you apply a profile on a package you must
stick to it (no pure UML elements inside packages compliant to SysML)
does this mean that you can actually mix up UML and SysML. A UML model may
contain packages on which SysML has been applied, but not visa versa AND it is
possible to have 1 packages containing subpackages compliant to UML and
subpackages compliant to SysML.
C) If the answer is YES,
how do you avoid building a model that "can lead to interpretation mistakes
and mistakes lead to unsafe systems". (quoted from Robert's reply, and I
agree on this problem resulting from mixing UML and SysML).
-------------------------------
Now the important question remains: can or can't I mix up models? What is
considered good/bad practice?
Important to note: I'm not making any statements, these are all questions.
The reason I'm doubting is because I'm currently figuring out how to
include the model of experiments (process models) next to models of
systems (product model) inside the project Model.
Project models must thus (according to me) include packages to which different
profiles apply:
--<Work>
packages (-> SPEM:Activity), which could for example apply the Ptolemy II
profile (and
later on QVTo to Executable Scientific Workflows (Kepler))
--<Product> packages (-> SPEM:WorkProductDefinition) which may be:
------<Concept>:
sysml would do really well on these type of packages (both hardware and
software). Another profile to apply could be SysML4Modelica (allow
the <concept> to be used for estimation of cost). If you're using
part catalogs provided by manufacturing companies within your concept,
these may/will include simulation of their product
------<System>:
Keep on using UML profile as some of the subpackages may be physical,
other may be software. Physical <Systems> may also apply a
specific CAD-vendor profile, which would allow your <concept> to
be transformed (auto generate file hierarchy of .part and .asm files,
download certain parts from manufacturer and set up expressions)
------<Document>:
Apply for example LaTeX profile: nearly EVERYTHING you do will/must
still get documented somehow (even if you disagree on this), set up
presentations/reports linked to the model
--<Reference>
comments (-> SPEM:Guidance): general references and perhaps even discussions
with
colleagues within the model. Big decisions are the result of fruitful
discussions so these could/should be annotating model elements.
Am I allowed to do this?
I'm considering posting the same message on google sysml-forum.
big mail...have fun reading :),
Dieter Van Eessen
Date: Sat, 11 May 2013 21:51:18 +0200
To: [email protected]
From: [email protected]
Subject: Re: [Topcased-users] RE : Newbie question 2, UML vs sysML model
Hello both (and the rest of the list),
OK, I'm not compelled to use the UML diagrams. That's clear.
BUT Dieter you say you do so because there are bugs in the sysML
diagrams (correct me if I'm wrong).
Dieter you also say that it is Bad Practice to take your previous
mentioned short cut.
On the last part I agree with you completely. In designing a
safety critical system you should not take any short cuts (how
inconvenient that may be). In my opinion mixing UML and sysML
should be avoided because that can lead to interpretation mistakes
and mistakes lead to unsafe systems.
Can anyone share an insight on the mention bug?
Best regards, Robert.
On 11/05/13 14:27, Topcased user list where issues are discussed
wrote:
Hi there,
I'm currently not using TOPCASED switched to eclipse kepler to
try out newest CDO features.
But yet, this might be a useful addition to the Raphael's answer
("Notice that you are not compelled to use those diagrams"):
I'm actually glad the UML diagram can still be used in sysml
models and subpackages where the sysml profile has been applied.
On the requirements diagram for example, trying to show the
containment's again (after they have been hidden) has resulted
in an error multiple times and the diagram must be recreated...
A class diagram doesn't have this problem (but it's a little
more work to make all the stereotype properties visible).
This is only one of the multiple reasons I prefer to use UML
diagrams to visualize SysML elements.
Hope these will be solved in the future. if the TOPCASED release
for kepler arrives, I'll definitely use it again! For now, the
UML diagrams still offer a solution for these bugs.
Is it considered BAD PRACTICE of you mix up models using
different meta-models, eg. UML models (using registered UML),
SysML models (using registered SysML profile) and others models
using even other profiles?
Currently got about 5 or 6 profiles which I use, does applying a
profile to a package mean I should stick to using THAT profile
within that package?
kind regards,
Dieter
> To: [email protected]
> Date: Sat, 11 May 2013 09:35:22 +0000
> From: [email protected]
> Subject: [Topcased-users] RE : Newbie question 2, UML vs
sysML model
>
> Hi Robert,
>
> you are right, communication and class diagrams should
not be part of SysML tool as they are excluded from SysML
language.
>
> Papyrus is still recent and efforts have been made on
extending UML to support SysML specific diagrams. No effort
has been put to restrict UML so that we support SysML in a
strict way.
>
> notice that you are not compelled to use those diagrams
;-)
> regards
> raphaël
>
>
> ________________________________________
> De : [email protected]
[[email protected]] de la part de
Topcased user list where issues are discussed
[[email protected]]
> Date d'envoi : vendredi 10 mai 2013 23:33
> À : Topcased user list where issues are discussed
> Objet : [Topcased-users] Newbie question 2, UML vs sysML
model
>
> Hello All,
>
> I'm back, sorry but the tool is harder to grasp then I
hoped.
>
> What I did was create a new project "File > New >
Papyrus Project".
> Named it PPtest and selected my Diagram Language "UML".
Then I selected
> my first diagram "UML UseCase Diagram". After that I hit
Finish.
>
> Everything look good so far. In the Project Explorer I
have a project
> "PPtest" and in the Model Explorer I have a model
"model".
> If I right click on "model" in the Model Explorer and
select "New
> Diagram" I will get a selection of UML diagrams I can
choose from. This
> is what I expect, in this selection I don't want to see
any sysML
> specific diagrams.
>
> BUT, if I create a sysML model under the PPtest project
and do the same,
> right click on the SysMLModel and select "New Diagram" I
will get ALL
> available diagrams including the "UML Communication
Diagram" and "UML
> Class Diagram".
> As far as I know (I'm not a sysML expert, more a sysML
novice) these
> diagrams are no part of sysML, so why are they included
in this selection?
>
> I didn't find any answer in the
>
TPC_5.3.0_UML&SysML_withPapyrus_features_tutorial-1.pdf
tutorial.
>
> Thanks again for your help. And by the way if the answer
is "it is as it
> is and you should know which diagrams to choose if your
in a sysML
> model" that's fine to I just would like to know ;-) (and
maybe help
> improve the tool).
> Best regards,
> Robert.
>
> _______________________________________________
> Topcased-users mailing list
> [email protected]
>
http://lists.gforge.enseeiht.fr/cgi-bin/mailman/listinfo/topcased-users
> ________________________________
>
>
> Ce message et les pièces jointes sont confidentiels et
réservés à l'usage exclusif de ses destinataires. Il peut
également être protégé par le secret professionnel. Si vous
recevez ce message par erreur, merci d'en avertir
immédiatement l'expéditeur et de le détruire. L'intégrité du
message ne pouvant être assurée sur Internet, la
responsabilité du groupe Atos ne pourra être engagée quant au
contenu de ce message. Bien que les meilleurs efforts soient
faits pour maintenir cette transmission exempte de tout virus,
l'expéditeur ne donne aucune garantie à cet égard et sa
responsabilité ne saurait être engagée pour tout dommage
résultant d'un virus transmis.
>
> This e-mail and the documents attached are confidential
and intended solely for the addressee; it may also be
privileged. If you receive this e-mail in error, please notify
the sender immediately and destroy it. As its integrity cannot
be secured on the Internet, the Atos group liability cannot be
triggered for the message content. Although the sender
endeavors to maintain a computer virus-free network, the
sender does not warrant that this transmission is virus-free
and will not be liable for any damages resulting from any
virus transmitted.
>
> _______________________________________________
> Topcased-users mailing list
> [email protected]
>
http://lists.gforge.enseeiht.fr/cgi-bin/mailman/listinfo/topcased-users
_______________________________________________
Topcased-users mailing list
[email protected]
http://lists.gforge.enseeiht.fr/cgi-bin/mailman/listinfo/topcased-users
_______________________________________________
Topcased-users mailing list
[email protected]
http://lists.gforge.enseeiht.fr/cgi-bin/mailman/listinfo/topcased-users
_______________________________________________
Topcased-users mailing list
[email protected]
http://lists.gforge.enseeiht.fr/cgi-bin/mailman/listinfo/topcased-users
