2. Getting started with Taverna

Table of Contents

2.1. Installation
2.1.1. Windows
2.1.2. Mac OS
2.1.3. Linux
2.2. Configuration
2.2.1. Artefact repository location
2.2.2. Proxy configuration
2.2.3. The mygrid.properties file
2.2.4. Default services
2.2.5. Image types and the mysterious 'Error 4'
2.2.6. Setting an alternative location for the 'dot' tool
2.3. Running Taverna
2.3.1. Enacting a predefined workflow
2.3.1.1. Loading a workflow
2.3.1.2. Enacting the current workflow
2.3.1.3. Browsing results
2.3.1.4. Saving a workflow
2.3.1.5. Closing a workflow
2.3.2. Creating a (very) simple workflow
2.3.2.1. Workflow inputs and outputs
2.3.2.2. A single sequence fetch component
2.3.2.3. Connecting everything together
2.3.2.4. Describing the input
2.3.2.5. Enacting the workflow
2.4. Example worklows
2.5. Custom perspectives

This section will walk through the process of installing and configuring Taverna then creating and enacting a trivial sequence fetcher workflow. While this is obviously a task which could (and arguably should) be performed in a simpler fashion it should serve as a reasonable introduction to the basic features of Taverna. This section assumes that you are running on some modern version of the Windows operating system (Windows 2000, Windows XP, Vista), there may be some minor differences with other systems such as Mac OS, Linux and Solaris although the basics should be identical.

2.1. Installation

Installations of Taverna require a Java 1.5 installation to run. Windows machines may not have this by default, so you will need to obtain it from http://java.sun.com. Linux users are more likely to already have this but should check - the procedure for Java installation on Linux is out of the bounds of this document.

2.1.1. Windows

On Windows machines installation should be extremely simple - open the downloaded archive (you probably already did this to read this file!) and unpack it to somewhere on your file system. The precise location doesn't matter but Taverna will not run from within a compressed folder - if you are using Windows XP and have not installed any software such as WinZip then Windows will appear to open the zip file just as if it were a normal folder on the file system, however, this is not the case and the differences will cause Taverna to fail.

2.1.2. Mac OS

Download the Taverna dmg; open it and follow the instructions shown.

2.1.3. Linux

Linux installation is very slightly more complex as you will have to install one extra piece of software first. Taverna makes use of the 'dot' application from the GraphViz package from AT&T research. This package (which is well worth installing for other reasons) is available as an RPM from the AT&T site here: http://www.research.att.com/sw/tools/graphviz/download.html. You may already have it, a simple check is to go to a command line and try typing 'dot', if the terminal just sits there then you probably have GraphViz already, if it shows an error indicating that the application is not found then you'll need to install it. Installation of the RPM will almost certainly require root access so you may need to ask your systems administrator to install it for you. Once you have GraphViz installed then you can simply unzip the Taverna archive to somewhere sensible, you might also want to sym-link to the runme.sh file from e.g. /usr/bin/taverna for ease of access.