General questions

Is Aduna software open source software?

Yes. And it can be used for free. Not only the semantic technology components, like Sesame, are open source, but also AutoFocus is open source. Aduna distributes the software under the Open Software License version 3, except for Sesame that is distributed under a BSD-style license.

But we don't like open source. Can we pay for the software?

For organizations that want additional warranties and imdemnifications, Aduna offers a proprietary license, called the Aduna Commercial License. This paid license includes a support contract and the license does not have the obligation to return source code to the community.

What is the difference between AutoFocus and AutoFocus Server?

AutoFocus is a desktop client that is capable of indexing file systems, websites and e-mail sources and present search results in Cluster Maps. AutoFocus Server is a full-text and metadata indexing server that can be queried by AutoFocus but it also has its own web-based UI. It is the web-based counterpart of AutoFocus. Using AutoFocus in combination with the AutoFocus Server means that, rather than letting AutoFocus deal with the task of crawling and indexing sources, this task is performed by the AutoFocus Server. Where in AutoFocus you previously would use a certain source, you can then define a AutoFocus Server source. One of the benefits of this approach is that the AutoFocus Server can be queried by an unlimited number of AutoFocus clients, i.e., you can share an index with multiple users. Also this makes disclosing the information in a web-based format possible.

Support

Can I use my own sources like file systems and databases in AutoFocus and AutoFocus Server?

At this time AutoFocus (Server) support the following sources: file systems, websites and IMAP email sources. If you want to include a source of another type, like a database, AutoFocus will need a specific connector. AutoFocus uses the Aperture Java framework for extracting and querying full-text content and metadata from various information systems. More information on: http://aperture.sourceforge.net

What about my own metadata and ontologies?

AutoFocus and AutoFocus Server extract and store a fixed set of metadata properties such as document titles, authors, etc. It could very well be that your documents contain other, company-specific metadata properties as well. In order to extract, index and query these, one would need to adapt AutoFocus and/or AutoFocus Server code. We can offer consultancy and development support for achieving this.

Where can I find the sources of products and components?

The sources and other technical information can be found at the Aduna developer website: http://www.aduna-software.org.

Where can I get developer support?

If you are a developer and you want more in-depth technical information on Aduna products and components, please go to the Aduna developer website: http://www.aduna-software.org. Sesame has its own website: http://www.openrdf.org. On both websites you will find documentation, the sources, issue trackers and more. Note: you can buy an Aduna Development Support Contract. This will give you direct access to the technical team. Read more...

Advanced topics

Where are AutoFocus' data files located?

There is an AutoFocus data directory in your home directory. The actual path to this directory depends on your platform:

• Windows Vista
  C:\Users\<USERNAME>\AppData\Roaming\Aduna\AutoFocus

• Windows 2000, 2003, XP
  C:\Documents and Settings\<USERNAME>\Application Data\Aduna\AutoFocus

• Mac OS
  /Users/<USERNAME>/.aduna/autofocus

• Linux
  /home/<USERNAME>/.aduna/autofocus

Where to find AutoFocus logs?

Look at logs subdirectory of AutoFocus data directory (see the question above).

Is it possible to change AutoFocus data directory (e.g. for using it from a USB key)?

When using AutoFocus.exe on Windows, the data dir can be specified by invoking "AutoFocus.exe -dataDir <somedir>" instead.

When using the Java runtime directly, the default application data dir can be overridden by specifying the aduna.platform.applicationdata.dir system variable to the Java runtime. When set, AutoFocus will use this as the "Aduna" dir and create an "AutoFocus" dir below it.

On Windows, you can do this by adapting the bin\run.bat script provided with AutoFocus. In there you are free to specify and alter any JVM settings you like. Add -Daduna.platform.applicationdata.dir=<YOUR_PATH> into two lines where JRE is started (after java.exe):

java.exe -Daduna.platform.applicationdata.dir=<YOUR_PATH> -Xmx300M -jar ...
...
"%JAVA_HOME%\bin\java.exe" -Daduna.platform.applicationdata.dir=<YOUR_PATH> -Xmx300M -jar ...

On the platforms where bin/run.sh script is used (Linux and Mac OS X), modify the script by similar way:

${JAVA_HOME}/bin/java -Daduna.platform.applicationdata.dir=<YOUR_PATH> -Xmx300M -jar ...

Does AutoFocus support stopwords lists to exclude common words from indexing?

Although Lucene does provide functionality for applying stop word lists, we do not use this in AutoFocus. In fact, including all words in the index makes the use of phrase searches possible.

AutoFocus does contain stop word lists for about 20 languages, but they are used for a different purpose. During indexing, AutoFocus uses a simple heuristic algorithm to determine a list of significant keywords for each document. These keywords serve as a sort of document summary and can also be used as search suggestions. The suggestions in the Keyword Search facet are in fact derived from these significant terms. The stop word lists are used in this process to suppress these common words from ever being suggested as a significant term or search suggestion.

Troubleshooting

How to avoid "Out of memory" error?

OutOfMemory exception usually arises when very large documents are indexed. To avoid it, you need to increase the amount of memory, available for AutoFocus ("heap size").

Please note: On Windows, the AutoFocus.exe launcher does not pass any arguments to the Java runtime. Therefore you will have to use the bin\run.bat script provided with AutoFocus to launch it.

By default, the startup scripts set the heap size limit to 300 MB. If OutOfMemory exception still arises with this value, you can increase it in Java runtime starting lines of the run.bat script:

java.exe -Xmx512M -jar %AFP_HOME%\lib\aduna-autofocus-desktop-4.0.jar %AFP_CMD_LINE_ARGS%
...
"%JAVA_HOME%\bin\java.exe" -Xmx512M -jar %AFP_HOME%\lib\aduna-autofocus-desktop-4.0.jar %AFP_CMD_LINE_ARGS%

The heap size in the run.sh script is set by similar manner.

Cannot search using chinese or japanese characters

We do not specifically support these languages yet. I will put it on the todo list, as we seem to have a growing number of Asian users.

How a file type is determined? Why the files with non-standard extensions are not indexed?

Generally we look for certain binary markers (so-called magic numbers) that identify certain file types. When this detection process fails to produce a detected file type, we fall back to a list of known file extensions per file type. In the case of a non-standard (unknown) file extension, the file type cannot be determined.

Why some characters are ignored in search queries?

This is caused by what is called the analyzer: before a document can be indexed, the analyzer breaks down the text in order to determine the individual words used in it. This analyzer throws away whitespace characters, punctuation characters, etc. The same analyzer is also used to break down your query into the individual terms.

As non-letters and non-digits are ignored, for example, the queries "searchterm", "searchterm/" and "searchterm " (with an extra space at the end) for example all end up being equivalent.