Skip to content

Latest commit

 

History

History
282 lines (208 loc) · 11.3 KB

README.md

File metadata and controls

282 lines (208 loc) · 11.3 KB
Raw

Taverna workbench

Taverna Workbench product build and installers

Released by myGrid

(c) Copyright 2005-2014 University of Manchester, UK

Introduction

This repository is a Maven source code trees that do not themselves contain any code, but which are used to assemble the distribution, add launchers and configurations files, and package this as installers.

You can think of a product as the last build step before having a downloadable/executable software distribution. Each product includes their required code from the common Taverna source code, by retrieved the compiled JARs from the myGrid Maven repositories unless they have already been built locally with Maven.

This is the build product for the desktop user interface Taverna Workbench.

In most cases you do not need to compile Taverna, instead you can download the official distribution for your operating system from the Taverna download pages.

For usage of this product, see the Taverna user manual.

Editions

This repository have several "edition" branches, one for each domain-specific builds. Each domain-specific build extends the core branch by modifying the plugin set (and ensuring the repository folder is populated for those plugins), but may also modify configurations such as the default service sets.

Branches:

  • core - for any domain - general activities (WSDL, REST, Beanshell, Component, Interaction, etc) - also basis for the other branches (can be considered the 'master' or 'trunk' branch)
  • bioinformatics - formerly maintenance, adds BioMart, BioMoby, Soaplab, WebDAV, corresponding default services and BioCatalogue
  • digitalpreservation - core + WebDAV
  • biodiversity - core + WebDAV + BioDiversityCatalogue
  • astronomy - core + AstroTaverna
  • enterprise - core with all possible plugins

To check which edition you are building, verify these bits of pom.xml:

    <artifactId>taverna-workbench-${edition}</artifactId>
    <name>Taverna Workbench [${edition}]</name>

    <properties>
            <edition>core</edition>

For more information, see http://dev.mygrid.org.uk/wiki/display/developer/Taverna+products

Creating a new edition

To create a new edition, make a new branch based on core and modify the above line(s). You would then typically also modify src/main/resources/plugins/plugins.txt and add the <dependency> sections from the new plugin(s) to the pom.xml so that the new plugin is added to the generated repository/ folder.

Note: You MUST use <scope>test</scope> in pom.xml for the added <dependency> blocks to avoid adding them to the lib/ folder, which would bypass the plugin-system and potentially cause conflicts with other parts of Taverna.

To verify the build, run executeworkflow.sh. A new user home directory for your edition should be created, e.g. ~/.taverna-astronomy-2.5-snapshot-20140228t1156/ for the astronomy edition.

Often when adding plugins, Maven will not download all the same versions required at runtime by Taverna's plugin system Raven. The first execution of executeworkflow.sh will download these to the repository folder of the edition-specific user home directory.

To avoid this download phase for other users and permanently add these "missing" pieces, try the equivalent of:

stain@biggie:~$ cd .taverna-astronomy-2.5-snapshot-20140228t1156/repository
stain@biggie:~/.taverna-astronomy-2.5-snapshot-20140228t1156/repository$ find . -type f > ~/files
stain@biggie:~/.taverna-astronomy-2.5-snapshot-20140228t1156/repository$ cat ~/files
./axis/axis-jaxrpc/1.4/axis-jaxrpc-1.4.pom
./axis/axis-jaxrpc/1.4/axis-jaxrpc-1.4.jar
./axis/axis-saaj/1.4/axis-saaj-1.4.jar
./axis/axis-saaj/1.4/axis-saaj-1.4.pom

There is a Python script that will read files from the current directory and show the XML that needs to be added to pom.xml:

stain@biggie:~$ python src/taverna-workbench-product/src/main/python/generateArtifactItems.py 
                            <artifactItem>
                                <groupId>axis</groupId>
                                <artifactId>axis-jaxrpc</artifactId>
                                <version>1.4</version>
                                <type>pom</type>
                                <outputDirectory>${project.build.directory}/repository/axis/axis-jaxrpc/1.4</outputDirectory>
                            </artifactItem>
                            <artifactItem>
                                <groupId>axis</groupId>
                                <artifactId>axis-jaxrpc</artifactId>
                                <version>1.4</version>
                                <type>jar</type>
                                <outputDirectory>${project.build.directory}/repository/axis/axis-jaxrpc/1.4</outputDirectory>
                            </artifactItem>
                            <artifactItem>
                                <groupId>axis</groupId>
                                <artifactId>axis-saaj</artifactId>
                                <version>1.4</version>
                                <type>jar</type>
                                <outputDirectory>${project.build.directory}/repository/axis/axis-saaj/1.4</outputDirectory>
                            </artifactItem>
                            <artifactItem>
                                <groupId>axis</groupId>
                                <artifactId>axis-saaj</artifactId>
                                <version>1.4</version>
                                <type>pom</type>
                                <outputDirectory>${project.build.directory}/repository/axis/axis-saaj/1.4</outputDirectory>
                            </artifactItem>

Now copy and paste to add those artifactItem blocks within the beginng (or end) of the <artifactIems> block in pom.xml. On next build, the repository/ folder of the distribution should now also include these files, and Taverna would no longer attempt to download them on startup.

To verify, now the repository folder should be empty:

mvn clean install
rm -rf ~/.taverna-commandline-*
cd target
unzip *.zip
cd *SNAPSHOT
sh executeworkflow.sh
find ~/.taverna-commandline-*/repository

Licence

Taverna is licenced under the GNU Lesser General Public Licence. (LGPL) 2.1. See the file LICENSE.txt or http://www.gnu.org/licenses/lgpl-2.1.html for details.

If the source code was not included in this download, you can download it from http://www.taverna.org.uk/download/workbench/2-5/#download-source or http://www.taverna.org.uk/download/source-code/

If you installed a OS-specific distribution of Taverna it may come bundled with a distribution of OpenJDK. OpenJDK is distributed under the GNU Public License (GPL) 2.0 w/Classpath exception. See jre/LICENSE.txt or http://hg.openjdk.java.net/jdk7u/jdk7u/raw-file/da55264ff2fb/LICENSE for details.

Taverna uses various third-party libraries that are included under compatible open source licences such as the Apache Licence.

Building

First, make sure you have checked out the edition you want to build.

Build requirements:

The below will assume you want to build the core edition.

git fetch --all          # fetch all branches/editions
git checkout core        # replace 'core' with your edition
mvn clean install

The packaging will take several minutes. The first time, this might take up to an hour as several libraries are downloaded from Maven repositories.

[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 5:39.145s
[INFO] Finished at: Thu Feb 27 01:06:08 GMT 2014
[INFO] Final Memory: 47M/1200M
[INFO] ------------------------------------------------------------------------

After a successful build, the file target/taverna-commandline-core-2.5-SNAPSHOT.zip (or equivalent) contains the platform-independent distribution of the Taverna Command Line Tool.

Installers

If you would like to build the platform-specific installers, you will need an installation of Install4j.

First configure Maven to find your Install4j installation.

stain@biggie:~$ cat .m2/settings.xml 
<settings>
  	<profiles>
  	    <profile>
  		<id>dist</id>
    		<properties>
    		    <install4j.home>/opt/install4j5</install4j.home>
		    </properties>
        </profile>
    </profiles>
  </settings>

Start install4j to install your license key or run a 30-day trial, then quit the user interface.

To generate the installers, run:

mvn clean install -Pdist

This process can take anything from 2 minutes to 30 minutes depending on your hardware. You can speed up the process by adding install4j-maven-plugin parameters, for instance generating only the Windows 64-bit installer with -Dinstall4j.buildIds=winx64. The buildIds are:

  • winx64 (Windows .exe 64-bit)
  • wini386 (Windows .exe 32-bit)
  • osx (Mac OS X)
  • linux_amd64 (Debian .deb x64)
  • linux_x86_64 (Redhat .rpm x64)
  • standalone (Windows .zip)
  • unix.tar (Unix .tar.gz)

Note that on first execution, this will also download OpenJDK binaries for embedding in OS-specifc installers. OpenJDK is licensed as GPL 2.1 with Classpath exception.

The media/media folder will contain installers for different operating systems.

md5sums
output.txt
taverna-workbench-core-2.5-SNAPSHOT-linux_amd64.deb
taverna-workbench-core-2.5-SNAPSHOT-linux_x86_64.rpm
taverna-workbench-core-2.5-SNAPSHOT-macos.tgz
taverna-workbench-core-2.5-SNAPSHOT-unix.tar.gz
taverna-workbench-core-2.5-SNAPSHOT-windows-x64.exe
taverna-workbench-core-2.5-SNAPSHOT-windows-x86.exe
taverna-workbench-core-2.5-SNAPSHOT.zip
updates.xml

Icons

If you would like to customize the icons and splashscreen used by the generated installers and launchers, see src/main/resources/lib - note that there are different icon files for different operating systems.

For more information on Taverna installers, see http://dev.mygrid.org.uk/wiki/display/developer/Taverna+installers

Support

See http://www.taverna.org.uk/about/contact-us/ for contact details.

You may email [email protected] for any questions on using Taverna workbench. myGrid's support team should respond to your query within a week.