[jboss-user] [JBoss jBPM] - Re: Storing hibernate persistent objects in variables

2006-10-24 Thread stibrian
I'm assuming due to the PEBKAC that you got this figured out, but just in case - are you clear that you don't need to actually do anything to have your custom hibernated types saved by jbpm? View the original post : http://www.jboss.com/index.html?module=bbop=viewtopicp=3980474#3980474 Reply

[jboss-user] [JBoss jBPM] - Re: About BPM miracles and what you can expect in real life

2006-09-09 Thread stibrian
anonymous wrote : Currently BPM is associated with BPEL and integration. We believe that BPM is much better served with a process language that integrates nicely with a general purpose programming language such as Java. That doesn't imply that BPEL is bad. On the contrary, BPEL is a great

[jboss-user] [JBoss jBPM] - Re: rhino script rather than beanshell?

2006-09-09 Thread stibrian
in the for what's it's worth category we were using Groovy for a bunch of scripting work around jBPM - using beanshell only for decision expressions. we actually switched to beanshell for everything just to keep process devs from having to make the mental context switch... Tom's right though -

[jboss-user] [JBoss jBPM] - Thoughts and ramblings on upgrading from 3.0 to 3.1

2006-09-09 Thread stibrian
Good afternoon. We're basing a medium sized application on jBPM, and had been using 3.0. Having modified it to suit our purposes a bit, we noted that many of the modifications we had made were covered in 3.1, so we went at an upgrade. not a terribly difficult process, but there were some

[jboss-user] [JBoss jBPM] - Re: Thoughts and ramblings on upgrading from 3.0 to 3.1

2006-09-09 Thread stibrian
I suppose it's a bit easier when that's link http://blogs.sourceallies.com/roller/page/joedeveloper?entry=jbpm_upgrade_3_0_to1 View the original post : http://www.jboss.com/index.html?module=bbop=viewtopicp=3970548#3970548 Reply to the post :

[jboss-user] [JBoss jBPM] - Re: counter-intuitive behavior persisting domain objects?

2006-09-08 Thread stibrian
I've clarified the fix a bit and posted a more defined explanation from my POV and a fix. http://blogs.sourceallies.com/roller/page/joedeveloper?entry=jbpm_3_1_auto_user View the original post : http://www.jboss.com/index.html?module=bbop=viewtopicp=3970268#3970268 Reply to the post :

[jboss-user] [JBoss jBPM] - Re: how to save data to my existed table but not jbpm_variab

2006-09-08 Thread stibrian
more info, with a tiny bit of example at http://blogs.sourceallies.com/roller/page/joedeveloper?entry=jbpm_3_1_auto_user View the original post : http://www.jboss.com/index.html?module=bbop=viewtopicp=3970269#3970269 Reply to the post :

[jboss-user] [JBoss jBPM] - Re: In-process HSQLDB

2006-09-08 Thread stibrian
to do it with an in mem db you'll also have to set the hibernate config to autocreate the tables for you. see the hibernate docs on how to do this. View the original post : http://www.jboss.com/index.html?module=bbop=viewtopicp=3970270#3970270 Reply to the post :

[jboss-user] [JBoss jBPM] - Re: How many data types do the process variables have?

2006-09-07 Thread stibrian
There is built in support for the primatives, byte[], String, Serializable, and most importantly, automatic and transparent support for using your domain objects in the process if you're using Hibernate as a persistence mechanism. There is one gotcha to the last thought though, and I JUST

[jboss-user] [JBoss jBPM] - Re: Document attachment possible ?

2006-09-07 Thread stibrian
I suppose you could persist it as a byte[], jBPM will persist them, but there will likely be issues with a blob that big (talking on pure assumption here). The alternative is to store some reference to your document in the process, and reference it when you need to with a link or some such

[jboss-user] [JBoss jBPM] - Re: Task is assigned , but is it accepted ?

2006-09-05 Thread stibrian
using a custom extension to TaskInstance and some custom logging you can get the desired behavior. Patience and use of the source will guide you - we needed and got the same thing working, so I know it's possible. Good luck. View the original post :

[jboss-user] [JBoss jBPM] - Re: passing data to AssignmentHandler best practice?

2006-09-05 Thread stibrian
using the names of the swimlanes and a custom assignment class is another trick that works well for me. Or using a named user that is stored in the process as a variable and then again using swimlane name as a key into which one you want - nearly no code, but nice flexibility. View the

[jboss-user] [JBoss jBPM] - Re: advanced examples

2006-09-02 Thread stibrian
I agree with the first response - many of the advanced things in jBPM are really customizations/extensions that are revealed as you read the user guide and come to understand the possibilities. for example we've built: emailing on task assignment that is configurable at runtime (not based on a

[jboss-user] [JBoss jBPM] - Re: how to save data to my existed table but not jbpm_variab

2006-08-28 Thread stibrian
yeah, so everything was working fine with the unit tests (transactions are not being simulated perfectly apparently), as in the actual web app there is a problem with the jBPM perfectly recognizing the hibernateable types and doing the right thing. Long story short, my domain class is

[jboss-user] [JBoss jBPM] - Re: how to save data to my existed table but not jbpm_variab

2006-08-26 Thread stibrian
I agree that you don't want to be using the jBPM tables in your domain, I'd encourage you to look at either writing converters OR switching to jBPM 3.1. You can write converters in either version, allowing you to reference full domain objects in the process (including in scripts via the

[jboss-user] [JBoss jBPM] - counter-intuitive behavior persisting domain objects?

2006-08-26 Thread stibrian
Good evening, I've been working with 3.1.2, and was in love with the concept of transparently persisting hibernatable domain objects within a process. There is a slight, what I would call counterintuitive problem with the default variable mapping config file related to this though... objects

[jboss-user] [JBoss jBPM] - Re: allow swimlanes to perform assignment at each reference?

2006-08-25 Thread stibrian
I'll post a patch when I implement the real solution. Currently we want this to happen always, so we made a quick modification to TaskMgtInstance only. To impl the patch fully there will be a bit (just a little) more effort. I'm waiting to hear from any of the jBPM devs on the general

[jboss-user] [JBoss jBPM] - Re: Passing information from UI into taskInstance

2006-08-25 Thread stibrian
there are a number of issues here, starting with the UI - what kind of UI? You'll programatically attack things slightly different is you're coming from the web tier rather than a fat client. Then variable type comes into play - from the web tier specifically you'll have issues as you'll have

[jboss-user] [JBoss jBPM] - Re: Assign a task to 2 users, how please?

2006-08-25 Thread stibrian
the previous post is dead on in my experience, and we use the null actorid versus a non-null extensively to determine where a task is to be displayed. You're using the identity module... I'm not using the Identity module - we looked at it and we have better luck with a handful of custom

[jboss-user] [JBoss jBPM] - Re: JBPM API integration with AD

2006-08-25 Thread stibrian
jBPM has a very open API, so yes, you can connect it. I'm assuming you want to do task based assignments based on AD roles/groups? Do a bit of LDAP programming and you'll see that there are a number of possiblities. Related to logging into a jBPM application with AD credentials... you'll

[jboss-user] [JBoss jBPM] - Re: In-process HSQLDB

2006-08-25 Thread stibrian
don't recall the exact setting, but you'll need to add mem to the connection URL - this indicates to HSQL that you want full in-memory mode. Not sure what to do in JBoss specifically, but this is how we run our unit/integration tests. View the original post :

[jboss-user] [JBoss jBPM] - allow swimlanes to perform assignment at each reference?

2006-08-24 Thread stibrian
Swimlanes are a very effective way to perform assignment, and we have implemented several that are nicely reusable. A swimlane though, isn't always (to my mind) the same actor or actors throughout the life of a process instance. In our case (and we've several use cases where this is