Troubleshooting

Sometimes on Windows, you may find an out of environment space error message when trying to run the scripts that lauch OilEd or the reasoner. This is caused by the fact that the default space for environment variables is quite small. To work round this, you will need to increase this space. This can be done either via editing config.sys or autoexec.bat depending on your OS version or altering the properties of the MS-DOS Prompt. If you do alter configurations, be careful and we take no responsibility for any damage you may do.

The Microsoft support site at http://support.microsoft.com/ has more information on this topic.

When classifying an ontology, OilEd makes use of a temporary directory. This directory is specified in the config.xml file in the installation directory, and is set to the value:

<tmp dir="tmp"/>

by default. If this directory is missing, then an IO Exception will be raised during the classification process. To solve this, either create the necessary directory, or change the value in config.xml to an existing directory and restart the tool.

OilEd 3.4 communicates with FaCT using a CORBA-based interface. This does not work correctly with JDK1.4. If you would like to use the reasoner with OilEd under JDK1.4, you will need to download the JDK 1.4 patch.

OilEd is a pure Java application, and so will run on Solaris, but the FaCT binary distribution is only available for Windows and Linux platforms.

OilEd expects to find a config.xml file that tells it, for example, which parsers and renderers to load up. This file should be on the CLASSPATH. Ordinarily, this should not be a problem, but if for some reason you have edited the startup scripts, be sure to include the directory where config.xml is in order to allow OilEd to find it.

Some aspects of the tool, such as whether inherited restrictions or inferred classes are shown, can be controlled by the Preferences panel. In most cases, the ontology will need to be closed and reloaded in order for such changes to take effect.

OilEd tries to ignore RDF information that is not related to the ontology. As a result, if individuals are introduced but are not explicitly stated to be instances of ontology classes, they may be ignored by OilEd when reading back in ontologies that have been produced. In order to stop this happening, make sure that any individuals you introduce are asserted to be instances of at least one class. This is a "feature" which will be addressed in later versions.

This is related to the way that OilEd deals with namespaces. Check page 17 of the manual.

OilEd can be set up to use an http proxy in order to load URLs -- this helps avoid problems where users have to access the web via firewalls. You can set the value for this proxy either in the config.xml file or through the Preferences panel. It is likely that you will want to change the value to use your local web proxy (ask your system administrators about this if you are unsure).

The basic application is written in Java, so the editor will, in theory, run on any OS with a Java 2 implementation. If you want to use the reasoner though, that has to be run on Windows or Linux. You may have to fiddle with the scripts in order to get the classpaths to work correctly though.

OilEd expectsto find a directory log in the directory in which it is installed. It uses this to record information about the status of the tool. If this is missing, this can cause problems. If you get a message like the above, try creating a log directory.