Our VisionX solution store got a new demo application. It's a demo for calculated columns. A calculated column is a virtual column, filled with custom content.

Simply install the application from our solution store:

Demo application - Solution store

The application has 3 workscreens. One demonstrates client-side calculated columns, the second one server-side calculated columns and the third screen demonstrates calculated columns directly in the database. You have different options to use calculated columns, simply check what you prefer.

Our generic REST interface in JVx is well documented and different examples for different technologies are available. The generic approach itself is smart but not super specific because the implementation follows the CoC paradigm. It's awesome for "REST out-of-the-box" without additional coding effort, but the flexibility is sometimes missing.

We tried to break some of this restrictions and get more flexibility. Here is a list of new features to explain some details:

• Fetch on demand

  The old implementation always returned all records from the database. It wasn't possible to implement paging or lazy loading. To solve this problem, we introduced new request parameters:

  _firstRow
  _maxRows

  With _firstRow (starts from 0) you define the start record number. The _maxRows defines the amount of records for the resultset, e.g. start from 0 and return max. 20 records, start from 20 and return 20 records

• Configure visible columns

  The old implementation returned the same column list for the REST call and the GUI/frontend. But sometimes the REST call should return a different list of columns because some internal columns aren't relevant or should be hidden. The new implementation got a new property in the metadata object: visibleColumnNames. If you set this property, the REST interface will return only the configure columns. The GUI won't recognize the new property because GUI controls have own visible column properties.

• Actions with Array parameters

  The old implementation didn't support simple Arrays like BigDecimal[] as parameter types for action calls. Only List was supported. The new implementation supports Arrays but it's not possible to mix Array and List parameters. An example:

  //valid
  public String createPerson(String[] name, String[] title)
  //valid
  public String createPerson(List<String> name, List<String> title)
  //invalid
  public String createPerson(String[] name, List<String> title)
• New admin zone

  We introduced a new admin zone for generic admin options. The URLs are:

  http://.../services/rest/APPLICATION_NAME/_admin/ACTION_NAME
  http://.../services/rest/APPLICATION_NAME/_admin/ACTION_NAME/ACTION_PARAM

  Currently, we support following actions (POST requests): changePassword, testAuthentication

  The request needs a JSON object with username, oldpassword, newpassword or password properties.

• Configurable zones

  We now allow custom zones. The current default zones are:

  http://.../services/rest/APPLICATION_NAME/_admin/...
  http://.../services/rest/APPLICATION_NAME/LIFECYCLE_CLASS/data/OBJECT_NAME/...
  http://.../services/rest/APPLICATION_NAME/LIFECYCLE_CLASS/action/ACTION_NAME/...
  http://.../services/rest/APPLICATION_NAME/LIFECYCLE_CLASS/object/OBJECT_NAME/...

  The zone names are: _admin, data, action, object
  It's possible to set custom names with following parameters in the deployment descriptor: zone.admin, zone.data, zone.action, zone.object

• Hide object names

  If you want to hide the object name(s), e.g.

  https://.../demoerp/services/rest/DemoERP/Customers/data/customer/

  The URL contains the Lifecycle Name (Customers) and the object name (customer). Both are easy to understand. In this example, we return the customer list from our customers object. If you want to use different names, it's possible to change the name in the lifecycle object code, e.g.

  @Replacement(name="business")
  public class Customers extends Session
  {
      @Replacement(name="clist")
      public IStorage getCustomer()
      {
          ...
      }
  }

  With above code, it'll be possible to request:

  https://.../demoerp/services/rest/DemoERP/business/data/clist/

The updated implementation is still a generic solution but it's now more flexible than before and should help you to offer new services. To use the new features, simply use our nightly JVx builds.

The support for ANT and JRE6 with Eclipse Photon.

The plugin was created for:

Don't forget the -clean start (read the original article for more details)!

One more thing you should do:

Photon includes ANT > 1.10. This version requires Java 8. In order to use Java 6, you have to download ANT 1.9.x. Simply download ANT and set the ANT Home to your version, via

If you try to use the bundled ANT with JRE 6, it will fail with an Exception.

Download the plugin from here. It works for us - no warranty!

More Details: Eclipse Oxygen.2 with ANT and JRE6.

We have some great Videos for new VisionX users on our YouTube channel.

Here are the links:

Getting started
How to create your first application
How to use Eclipse & VisionX in real time

Check them out!

These Videos will help new users learning VisionX

I'm happy to announce that our Workflow Engine will work with Oracle Forms.

The current state is beta but we'll release our Workflow engine in the next days. We found some time to test the integration in Oracle Forms and the result is soo cool. It works like a charm.

So it will be possible to embed our Workflow engine in your oracle Forms. It'll be possible for your users to design custom flows and processes. The engine itself is database driven and will work with your Oracle Forms application.

More details will follow in the next days, but here's a first impression:

Workflow Engine - Oracle Forms

It's a standard Oracle Forms window/canvas with our Workflow Designer, embedded as Java Bean.

Feel free to ask your questions

We improved the support for standard MacOS LaF in our Swing UI and VisionX. The LaF has a lot of rendering problems. We tried to use another LaF implementation but had other problems. So we decided to keep the standard Java MacOS LaF.

The problems

• The InternalFrame border problem
• The field border problem
  

  Field border problem

  A simple JTextField has a 5 pixel focus border. This is horrible because the focus is about 2 pixel. The other problem is that a JTextArea wrapped in a JScrollPane doesn't have the 5 pixel border. This makes it hard to create nice looking layouts.

  We fixed this problem:

  Field border without gap

  Looks better. One problem is that the JScrollPane doesn't paint the focus if the JTextArea has the focus. This is a LaF problem and we didn't find a workaround for this problem.

• Application Menubar

  Standard Java appliacations show the menubar in the same frame

  Application menu

  If you set the system property:

  System.setProperty("apple.laf.useScreenMenuBar", "true");

the menu will be placed in the menu bar of MacOS:

macOS Menu

We fixed the problem that setting the menu to null will keep the menu as it was.

• Comboboxes

  The height of Standard Comboboxes is wrong:

  Combobox wrong

  We fixed the problem as good as possible:

  Combobox fixed

• VisionX Toolbar height

  The toolbar height was wrong because of layout calculation problems:

  Wrong Toolbar height

  We fixed the problem and the toolbar height is correct:

  Toolbar fixed

• VisionX Wizard gaps

  The image gaps in VisionX wizards was not recognized:

  Wrong gaps

  It looks better without gaps:

  No gaps

• Wrong button rendering

  If you had a button without an image and a button with an image, the height of the button without image was higher than the button with an image. This is a LaF rendering bug:

  Wrong size and font

  The fonts are also different. We didn't know why there has to be such a difference, but we fixed the problem in Swing UI:

  Button with correct size and font

All changes are available in our nightly JVx builds

This article is a short introduction of our brand-new product. It's a super simple workflow engine. Our workflow engine is database driven and can be configured and extended for your needs. The engine itself contains a simple board which allows you to create workflows. The configuration will be read from and saved into a database. The workflow steps are configured in the database and the engine simply executes the configured steps.

We tried to find a workflow solution for our needs or a simple product which supports drawing workflows. We found a lot of complex workflow engines and tools but not a simple one. We didn't like complex solutions so we started with our own workflow enigne.

And here are some impressions

Workflow Designer (loop)

Workflow Designer

The left area contains the list of available workflows and all used steps for each workflow. The search option is a full-text search for workflows and contained steps.

The right area contains all available/configured steps. The steps are configured in a database. We offer a simple Java interface which makes it possible to implement your own workflow steps. The area also contains a mini version of the board. This makes it smooth to search through the workflow.

The middle area, the board, contains the workflow designer. It supports Drag and Drop.

The created workflows can be used from any JVx based application. With next VisionX release, it'll be possible to use the new workflow engine.

The workflow engine will be available as AddOn for VisionX and as standalone version for product bundeling.

We're happy to anounce that VisionX 5.1.8 is available!

It's a simple update release of VisionX 5 with smaller bugfixes and one awesome new feature.

The bugs fixed

• Support for STARTS_WITH and CONTAINS Filter conditions

  We didn't save this conditions correctly.

• Keyboard Navigation in HTML5 UI

  It wasn't possible to use arrow keys for navigation in HTML5 tables.

• Floating License

  We had some problems with floating license check with 1 user and the HTML5 live preview.

The new feature

• Swagger UI support

  We now create Swagger documentation for application' REST services. It's possible to test your REST services with your browser - super easy.

The download area already contains links to VisionX 5.1 binaries.

What is Xamarin?

Read all details here.

It's a company and a really cool platform for mobile application development. It enables you to use one codebase for iOS and Android.

Since 2016, it's more or less free for everyone.

We tried Xamarin some years before 2016 and it wasn't free. So it wasn't an option for us. Not because it's using C# as programming language. We already have mobile clients for JVx applications but they don't share the same codebase. There is one client for Android and another one for iOS. The source code is licensed under Apache 2.0.

Our native clients work great but we think they could be better.

In the last days we made some tests with different tools, to create a new universal client with one technology. A HTML5 client wasn't an option for us because of different criterias. A real native client was important during our evaluation.

Some links to different opinions:

• https://techbeacon.com/10-cross-platform-mobile-development-tools-enterprises
• https://medium.com/@MasterOfCodeGlobal/best-10-android-frameworks-for-building-android-apps-d2d0ee48e464
• https://www.lifewire.com/top-tools-for-multi-platform-mobile-app-development-2373505
• https://www.iqvis.com/blog/9-cross-platform-mobile-development-tools-to-use-in-2017/
• https://www.outsystems.com/blog/free-cross-platform-mobile-app-development-tools-compared-2017.html

We made good first progress with Xamarin and because of its pricing, it was the winner of our evaluation. Sure, Xamarin is C# but it's better to write C# than JavaScript. Another plus was that we already have a simple JVx connection port to .NET based on C#. So we could reuse this code.

Why not JavaFX?

The problem with JavaFX is, that it's unclear what will happen with JavaFX in the future. There are different articles about the future of JavaFX, but the one from Jonathan Giles makes it clear

Sure, there's Gluon and some other individual developers, but the overall performance isn't relly comparable to native apps. It's awesome what happened in the last years, but the progress is missing.

We still have our JavaFX based UI, but we're not sure if JavaFX will survive or it will be commercialized. So it's better to have an option.

Why not CodenameOne?

It's a really cool technology and works great with some smaller performance glitches (simply try the current KitchenSink from the App stores)..
WORA just works!

The cloud build isn't what we prefer and the pricing is not good for our use-case. Our client should be free, open and easy to use for every developer.

Here is a simple test application for REST access to our Heroes application, used in our Angular test application.

iOS client

Android client

The source code for this test: