We are evaluating Activiti as a process engine to replace our existing home grown work flow engine. We are quite impressed by the capabilities of Activiti especially related to multi tenancy and REST WS.
However, one of the biggest challenge (and probably blocker) to adopt Activiti would be - How we can run or migrate our existing work flow definitions.
As I mentioned earlier, our work flow solution is a home grown one and doesn't adheres to BPMN specifications. There are thousands of templates out there. We can't simply ask our customers to redefine their templates using Activiti. These definitions are stored in proprietary XML format.
Looking at the level of customization in the templates, it would be very difficult to migrate these definitions to BPMN format.
So, does Activiti provides any hooks to run such custom templates. Alternatively, please share your thoughts about migrating the templates from proprietary format to BPMN format.
I suppose such scenario would be common and other people would have faced the same.
I know I am being very vague with this query but at this stage I don't have specific problems that I can discuss.
One option is implement your own proprietary XML parser and parser handlers. Look at org.activiti.engine.impl.bpmn.parser.BpmnParse and org.activiti.engine.impl.bpmn.parser.handler.AbstractBpmnParseHandler and its descendants.
We did it and worked fine.
Related
I am looking for an easy-to-implement solution for form-based ontology editing and I wonder if there are any active projects and which of them is the right path to follow.
I need to create instances of an ontology (lets call it ontology A) using forms (either web or desktop) and store them in a triple store (e.g. Virtuoso). I would like to hide as much details as possible regarding the ontological relationships between the entities defined in ontology A and provide a plain simple user interface for CRUD (Create-Retrieve-Update-Delete) operations based on the entity schema defined in ontology A.
For example, I have found two possible solutions in the protege ecosystem:
PropertyFormPortlet It is not a live feature in the current webprotege version.
facsimile project
As described in the respective paper this is a solution that has been implemented for a specific context. Therefore, adapting it to another domain would not be as straightforward as I would like.
I wonder, is there another solution (even out of the protege ecosystem) that could facilitate such a form-based ontology editing. Could somebody provide some guidance?
Just in case someone lands in this question, I write down my conclusions. Practically, none of which I tried worked, but still I found out some interesting things.
OpenLink Structured Data Editor:
OSDE is a browser plugin which aims at populating RDF graphs in the form of files, based on Linked Vocabularies. In my case it didn't work as my locally hosted ontology cannot play the role of a "Linked Vocabulary". However, OpenLink team said they will work on it.
OData2SPARQL:
In our test, the WebIDE did not manage to create the forms out of the box as suggested by the respective video tutorial. However, we managed to use OpenUI library as a client of the OData services automatically created by OData2SPARQL, providing a web service interface for our ontology.
Ontowiki:
In our test environment, OntoWiki partially worked. We could save data but there were some bugs when trying to add properties etc. OntoWiki developers said that they plan to refactor in order to actively support it in newer hosting settings, but this is not the case right now.
I like the TDD approach to documenting your restful api with spring-rest-docs. However, I love "API Playground" feature enabled by swagger specification. I wish there was a way to get best of both worlds.
Is there a way to build swagger2 specs from spring rest docs? may be via building custom request/response preprocessors.
Do you have any thoughts or recommendations?
There's not out-of-the-box support for this in Spring REST Docs at the moment. The issue that you opened will track the possibility of adding such functionality. In the meantime, your best bet would be to look at writing a custom Snippet implementation that generates (part of) a Swagger specification.
Typically, a Spring REST Docs snippet deals with documentating a single resource, whereas a Swagger specification describes an entire service. This means that the Swagger specification Snippet implementation will need to accumulate state somehow, before producing a complete specification at the end. There are lots of ways to do that (in memory, multiple files that are combined in a post-processing step, etc.). It's not clear to me that one approach is obviously the right one so some experimentation would be useful. If you do some experimentation, please comment on the issue that you opened with your findings.
i have a couple of questions on Alfresco.
Is Alfresco workflow module, point and click configurable or is programming needed?
Can Visio diagrams be used to set workflow?
Can Alfresco process 1-D and 2-D barcodes?
Regards
vish
Currently Alfresco workflows can be developed using the JBPM Graphical Process Designer, which produces XML files you directly upload into Alfresco. Be aware of the fact that in the future the default workflow engine of Alfresco will change to be Activiti.
Barcode generation/scanning is not supported by Alfresco.
At the Alfresco Devcon in New York (2010) Neil McErlean did a presentation on the Alfresco Action Framework where he demoed an example of working with QR codes.
The presentation can be found at http://www.slideshare.net/alfresco/custom-action-framework
I am not sure where the source code lives for this customization. So while it can't be done OTB, it is a customization that is possible and has been done as a POC.
More recently, the Activiti workflow engine is the default in Alfresco (though jBPM is still available). There is an Eclipse plugin that can be used to do the initial workflow design graphically (and output BPMN XML files).
However, the BPMN is not useful until it has been manually enhanced with Alfresco-specific features such as the assignee (person or group) for a task, the form to use to display the task, and scripts to create and maintain process and task variables, and/or hook into events triggered during the workflow. Further work may be needed to define the task model (the variables that each type of task needs) and to customise the Share user interface to display the tasks correctly to users. See this article for an example.
So although the initial draft design can be done graphically, detailed programming is also needed. The Alfresco training materials for workflows specifically state that end users cannot create new workflows without assistance from a developer.
The exception is "simple workflows", which don't use BPMN or jBPM, and are simple one-step Accept/Reject decisions that can be created by end users.
You can use whatever BPMN tool as long as it generates an xml BPMN xml output so you can import it into Alfresco activiti
We currently have an application that is based on an in-house developed workflow engine with YAML based DSL. We are looking to move parts of it to Java.
I have discovered a number of java solutions like Intalio, JBPM, Drools Expert, Drools Flow etc.
They appear to be aimed at businesses where the business analyst creates the workflows using a graphical editor and submits them to the workflow engine. They seem geared towards ease of use for non-technical people rather than for developers with a focus on human interaction.
The workflows tend to look like.
Discover-a-file -\
-> join -> process-file -> move-file -> register-file
Discover-some-metadata -/
If any step fails we need to retry it X times. We also need to be able to stop the system and be able to restart it and have it continue from where it was (durable).
Some of our workflows can be defined by a set of goals we need to achieve so Jess's backwards rule chaining sounds interesting but it is not open source.
It might be that what we are after is a Finite State Machine engine or just an Enterprise Service Bus and do everything as JMS queues.
Is there a good open source workflow engine that is both standards-based but also geared towards developers. We don't particular want to use a graphical workflow designer or write reams of XML and it should ideally be in Java or language agnostic (makes REST/Soap calls to external services).
Thanks,
Tom
Both Activiti and Bonita are open source and standard based (BPMN2). See for example this blog post.
Ruote is not standard based but seems close to your DSL approach and runs on a JVM thanks to JRuby.
Intaloi an open source BPM engine it offers a BPMN-support Designer and a BPEL engine. it's written in Java.
Camunda BPM is a developer-friendly Open Source workflow engine that is based on the open standards BPMN 2.0, DMN 1.1 and CMMN 1.1.
While it does come with a comfortable graphical workflow designer it also ships with a fluent API to build workflows programmatically. Camunda is written in Java, but can also be invoked from other languages via its REST API and it can make REST/Soap calls to external services.
jBPM 5 (open source, ASL, BPMN2) is just released and it's the best of Drools Flow and jBPM 4. It's lightweight but it can also integrate deeply with the Drools rule engine to make decisions.
For anyone looking for Python based enterprise grade solution.
Zengine, is GPL3 BPMN workflow based framework with Tornado, Rabbit AMQP, advanced permissions, extensible scaffolding features and more.
Built on top of following major components;
SpiffWorkflow: Powerful workflow engine with BPMN 2.0 support.
Tornado: Tornado is a Python web framework and asynchronous networking library.
Pyoko: Django esque ORM for Riak KV store.
RabbitMQ: Fast, ultrasharp AMQP server written with legendary Erlang lang.
We've used the Validation Block of MS Enterprise Library for some time with great success in conjunction with custom DALs but we've recently started using the Entity Framework and can't get the Validation Block to work with it. The objects are dynamically created in EF and putting attributes on top of them will simply get wiped out when the models are re-genned.
Can these two co-exist? If not, does anyone have any recommendations for what validation library/simple rules engine would be a good candidate to use along with EF?
Thank you.
You need a validator which supports a "buddy class" (like this example for Dynamic Data). This seems to be a work in progress for VAB. I can't find an example of anyone actually using it yet, but it might work.
Validation Application Block supports the concept of configuration based validation. This way you can separate your generated domain entities from validation. You can use the Enterprise Library configuration tool for this. Simply right-click on your configuration file and start adding validation configuration.
I advise you to read the VAB Hands on Lab document (ValidationHOL.pdf) which is included in the Hands On Lab download. After reading that document, read this article. It explains how to integrate VAB with Entity Framework.
Good luck.