Re: [concern-users] evaluating con:cern

[frantisek k. <fra...@gm...>]

Thanks for the link, can't wait to try it out.

Fero

On Tue, Dec 2, 2008 at 10:07 PM, Holger Engels <he...@gm...> wrote:

> Hi,
>
> how to build the osbl is documented here:
>
>   http://osbl.wilken.de/wiki/index.php/Installation_Guide
>
> Yes, con:cern has a notion of tasks. Of yourse you can freely edit the
> object and notify con:cern of the change an con:cern will pick up th
> enew state and continue with the process. But this is merely the
> unusual case. Normally you define a bunch of activities with
> preconditions and postconditions. As soon as the precondition of an
> activity is met, con:cern will either execute it (synchronous system
> acitvity) or enlist it for asynchronous execution. Asynchronous
> activities are mostly user activities. An actor assignes the activity
> to one or more users. If the activitiy is obligatory (it must be
> executed and it will timeout / escalated, if it's not), con:cern will
> create a task for them. If the activity is marked optional, con:cern
> creates an option, that can be executed (by their option).
>
> If you manage to deploy the vacation process demo and study it
> parallely in the process monitor and in the con:cept editor, you will
> get the picture.
>
> Regards,
>
> Holger
>
> On Tue, Dec 2, 2008 at 9:07 PM, frantisek kocun
> <fra...@gm...> wrote:
> > Hi,
> > I'm trying to build the trunk. I it's a little bit complicated but I
> managed
> > to build all the project which demo project depends on (conform, concern.
> > concern-library, osbl, osbl-basics) and finally osbl-demo.  To build
> > osbl-shell I need file osbl-shell\etc\custom-build.properties. How does
> this
> > file look like?
> >
> > Our problem with workflow is what should be modeled as task and what as
> > editation of business object, which has nothing to do with process.
> That's
> > why I'm trying con:cern and osbl. I'm curious how you deal with this. So
> how
> > you deal with this? Do you have something corresponding to task in jbpm?
> Or
> > do you only edit business objects which in turn notifies the process,
> > business object is in (I think this is your case and the action/tasks
> which
> > can be done are determined by process). Or do you have tasks and business
> > objects as well? Is there any article on this topic? I think this is the
> > most important thing for systems dealing with processes and objects. We
> > don't have a classic document-based system. We have a system which has a
> lot
> > of CRUD with a lot of difficult constarints and a lot of business logic
> in
> > it. Modifying one object can trigger creation of tens of another objects.
> I
> > think we have processes in it as well (but we don't program it that way).
> > Sometimes you have an object which can be handled only in one way (or in
> > several ways depending of attributes values) but you can't see this as a
> big
> > picture.
> >
> > Thanks
> >
> > Fero
> >
> >
> >
> >
> > On Sun, Nov 30, 2008 at 9:49 PM, Holger Engels <he...@gm...>
> wrote:
> >>
> >> Hi,
> >>
> >> the con:cern approach has its strength were you model the behaviour of
> >> isolated objects in respect to complex rules and influences with
> >> special cases and the like. The possibility to manually influence
> >> running instances, testability and the simple means to query the state
> >> of an instance and to find out, why things happen or do not happen
> >> make working with con:cern very conventien.
> >>
> >> On the other side, it's sometimes hard to model sequential flows as
> >> you need conditions and state for every step. Also there's not split
> >> and join operation. If activities must be performed on an object has
> >> children (1:n), you need to build a separate process for the children
> >> and to coordinate the the parent's process with the children's
> >> processes.
> >>
> >> However, I'm still convinced of the approach. We've built several
> >> processes with con:cern now and development was always very straight.
> >> Finding errors and reasons for errors is also straight forward.
> >>
> >> Please do not use osbl 1.0. It's quite old. Please checkout trunk from
> >> subversion. Trunk is stable at the moment, as we will release osbl 1.2
> >> by next week. con:cept will load the models from trunk.
> >>
> >> There's no documentation beside the osbl wiki. Though you should be
> >> able to understand it by studying the examples.
> >>
> >> Kind regards,
> >>
> >> Holger Engels
> >>
> >>
> >> On Sat, Nov 29, 2008 at 12:53 PM, frantisek kocun
> >> <fra...@gm...> wrote:
> >> > Hi,
> >> > I'm evaluating OSBL and workflow engines in general. I have no
> >> > experiences
> >> > with workflow systems. I tried two engines and decided for jBPM. But I
> >> > have
> >> > some doubt about it. And especially after reading all the patterns
> from
> >> > workflowpatterns.com, what is really good site. Why such easy things
> >> > needs
> >> > so much effort to realize (tokens, decisions, events... all the
> >> > imperative
> >> > way). Last week when reading Case Handling: A New Paradigm for
> Business
> >> > Process Support by Wil M.P. van der Aalst I realized, that this is
> what
> >> > we
> >> > need. But is just a feeling I don't have any evidence. To gain
> evidence
> >> > I
> >> > must try a case handling system and model various situations, some
> which
> >> > are
> >> > easy to model in jBPM and hopefully stay easy to model in concern, and
> >> > some
> >> > which are hard to model in jBPM and will be easier to model in
> concern.
> >> >
> >> > This is what I found very useful, hopefully I understood it well:
> >> >
> >> > Flow can be data driven but you still have the causal realtionship
> >> > between
> >> > tasks.
> >> >
> >> > I like the idea that subject is the same as process instance in other
> >> > workflow engines, which gave me power to make various queries with
> user
> >> > defined process variables (not only default variables, such as process
> >> > name,
> >> > time created... I made it working in jBPM  with user defined process
> >> > variables as well but it was a lot of additional coding and several
> >> > queries/subqueries are needed to do that job).
> >> >
> >> >
> >> > My questions:
> >> >
> >> > 1.) Is there any working example of concern. I'm interested in
> something
> >> > easy like http://docs.jboss.org/jbpm/v3/userguide/tutorial.html .
> >> > Process
> >> > definition, code and junit tests. Not too much architecture, because
> we
> >> > have
> >> > our own, and I want pluggable solution. Just to get familiar with
> >> > process
> >> > definitions, con:cern api and execution model.
> >> >
> >> > 2.) I downloaded osbl 1.0 and con:cept 1.1 and I can't open urlaub and
> >> > riskmanagement examples in designer. Probably they are using different
> >> > version of emf models.
> >> >
> >> > 3.) What are your experiences with con:cern so far? I have found some
> >> > older
> >> > pages, but that time you were only beginning.
> >> >
> >> > 4.) Is there any comprehensive documentation for all the model
> elements
> >> > in
> >> > con:cern language (preffered English but German is ok as well)?
> >> >
> >> > 5.) Can you point me to some interesting articles concernig case
> >> > handling?
> >> > The best with case handling and workflow examples.
> >> >
> >> > Thanks for help
> >> >
> >> > Fero
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> >> >
> -------------------------------------------------------------------------
> >> > This SF.Net email is sponsored by the Moblin Your Move Developer's
> >> > challenge
> >> > Build the coolest Linux based applications with Moblin SDK & win great
> >> > prizes
> >> > Grand prize is a trip for two to an Open Source event anywhere in the
> >> > world
> >> > http://moblin-contest.org/redirect.php?banner_id=100&url=/
> >> > _______________________________________________
> >> > concern-users mailing list
> >> > con...@li...
> >> > https://lists.sourceforge.net/lists/listinfo/concern-users
> >> >
> >> >
> >>
> >>
> -------------------------------------------------------------------------
> >> This SF.Net email is sponsored by the Moblin Your Move Developer's
> >> challenge
> >> Build the coolest Linux based applications with Moblin SDK & win great
> >> prizes
> >> Grand prize is a trip for two to an Open Source event anywhere in the
> >> world
> >> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> >> _______________________________________________
> >> concern-users mailing list
> >> con...@li...
> >> https://lists.sourceforge.net/lists/listinfo/concern-users
> >
> >
> > -------------------------------------------------------------------------
> > This SF.Net email is sponsored by the Moblin Your Move Developer's
> challenge
> > Build the coolest Linux based applications with Moblin SDK & win great
> > prizes
> > Grand prize is a trip for two to an Open Source event anywhere in the
> world
> > http://moblin-contest.org/redirect.php?banner_id=100&url=/
> > _______________________________________________
> > concern-users mailing list
> > con...@li...
> > https://lists.sourceforge.net/lists/listinfo/concern-users
> >
> >
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's
> challenge
> Build the coolest Linux based applications with Moblin SDK & win great
> prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> concern-users mailing list
> con...@li...
> https://lists.sourceforge.net/lists/listinfo/concern-users
>