Studio can be installed for either tomcat or WebLogic. WebLogic is however required to install OES, so though you don’t need to install it to run Studio you may find it just as well to learn to work with it.
From the studio perspective the paradigm is now that Studio contains multiple Applications, and each Application is tied to a Data Domain. The “Application” is roughly akin to the “Communities” we had before. You can have multiple Applications per Data Domain. The convenience when adding any component is they’ll automatically inherit the data connection for that application. Since this is now enforced there’s no longer a “Data Bindings” section in the control panel.
You can set up your data sources in a parent-child hierarchy that will allow you to combine more than one data domain to a single Application. Doing so enables a property on each component to specify which data source it should use. The child data sources can apply refinement filters or identify separate data domains altogether.
Applications are considered “Certified” if an administrator has checked a box implying they approved its content for sharing. There is also an option to mark an Application as active or inactive.
Creating a new page is much the same as it was before, same as for the templates. The available components don’t appear to have been changed. Range filters are absorbed into the Guided Nav component.
Moving components around seems much more reliable; however as before dragging charts into the tabbed component (for example) does not guarantee they’ll stay there.
Updating the components is a simpler single link to the configuration of the component. Instead of the link to “Return to Full Page” the return link now says “Exit”. The ability to set the permissions on an individual component has been removed.
Administering the site has been changed in a few ways. The Data Sources are unchanged, however there is now the Provisioning Service. From the control panel you only need to update the JSON string and specify the name and port for the OES server. This connection will be used to create any data domains required to support any user loaded datasets (Excel only).
The Provisioning Service runs as an application on the WebLogic server. You’ll need to install the .ear file the same way you installed the Studio application. This is a pretty significant enhancement of the tool, more on it later.
Framework Settings behave largely as before though include some changes to the available settings. As with 2.3 some of the settings won’t appear until first use. For example defaultExporter, mapLocation, maptTileLayer, mapViewer or viewTransitionDefaultContext. They also appear in the underlying file (portal-ext.properties) where they can be edited and added.
Added are settings for provisioningServiceLimit and performanceLogging. The former setting allows a limit to be set on the number of files users can upload, since each time they’ll create a data domain this could rapidly become an issue. The latter since PerformanceLogging is enabled by default. You can specify performance logging on QUERY, PORTLET or CLIENT, by default only CLIENT is excluded. You can directly retrieve and analyze those metrics from the eid-studio-metrics.log file.
Most of the existing sections around managing Users, User Groups, Roles, Password Policies and Authentication seem largely unchanged. Admittedly I didn’t exhaustively compare them.
Instead of the Control Panel there is now a section called “Application Settings”. View Manager has been relocated here, as well as the console for viewing some of the metadata around the attributes and their groups. Though you can add new groups and edit their membership, the properties of the individual attributes can only be edited through Integrator or one of the OES web services.
View Manager has changed its look & feel but otherwise is largely unchanged. Both in how views are created, the fact they’ll disappear on a reset of the data domain, or in how they are interacted with from any chart.
The EQL language has seen some intriguing updates. Existing functions for numbers and dates appear largely unchanged. There are however two string manipulation functions now available (CONCAT & SUBSTR). There are new options around grouping records into custom sets. Sets can contain one or more members and be combined together.
ROLLUP is an extension to GROUP BY. It is designed to calculate subtotals at each grouping level. At each level the granular values are “rolled up” to a final grand total. The other new extension to GROUP BY is CUBE, which calculates subtotals for each combination of attribute being grouped by. That sense of compiled quantitative values at multiple intersection points being the element in common with a traditional OLAP cube.
The chart component is largely unchanged, with the exception of course that scatter plots are available as a chart type. This new chart allows you to display either scatter points or bubbles, so multiple metrics can be visualized simultaneously. The configuration steps for all charts are themselves unchanged, and the settings around how data is displayed, sorted and interacted with are the same.