Relationship to Taverna Workbench
Kabak (codename) is a new tool. It allows users to create, edit and run workflows (although they are probably best not called that), created from components (and data) that are well-described and work together.
It is _not_ a re-implementation or extension of Taverna.
It removes all unnecessary complexity of Taverna and allows the power of workflows to be made available to ordinary (not power) users.
| Compexity |
Simplification |
|---|---|
| Activity/Service types |
Components only |
| Output rendering (mime magic) |
Output data is typed from components. Data is always typed in components. |
| Lists (of lists...) |
Only singletons or tabular data |
| myExperiment and BioCatalogue integration |
No myE or BioCatalogue integration except when developing components |
| Dot/cross product |
Fixed in component definition. Palette of choice for behaviour |
| Opening any T1/T2 workflow |
Only open certain workflows |
| XML splitters |
Inside components |
| Nested workflows |
Inside components |
| Error handling |
Explicit part of component |
| Configuration |
Only what's defined in the component |
Users of Kabak will not see what is inside a component. It is possible that they may be able to edit a component which will start a component editor on that component.
Components
In Taverna Workbench
Ability to export a workflow as a component to a library. This includes specification of the types of the input and output data and the palette of possible dot/cross behaviours.
Component editor. Need to analyze how this differs from ordinary workflow editing.
Import a component from a library so it can be edited.
In Component Library (Agora)
Component families
Component viewer
Component editor