20100719 Release Notes
The goal for this release was primarily to provide new
metadata for the corpus. A consequence
has been an improvement in general quality of the entries of the corpus,
as the computation of the metadata has provided an additional check on
This release added no new systems, but there are 95 new versions of 14
systems, giving a total of 495 versions overall of 100
systems, with 23 systems having more than one version.
New metadata is being added. The documentation is changing to
describe this new metadata. All new metadata documentation can be
found here. This affects
the overall organisation
and the the system version structure.
- System name change: "Vuze" -> "Azureus"
The developers of Azureus have decided to change its name to "Vuze".
However, many of the Java files (more than 2000 out of probably no more than
3000 total) are identical between versions, so it seems there is no
significant structural change due to the renaming. We will continue
to use 'azureus' in the corpus. The version numbering continues as
well, so using the old name will not lead to problems in this respect.
The fullname property
has the name "vuse".
- Added systems
No new systems have been added.
- Added versions
- Checked source
With the change in the metadata, this is no longer relevant and will
be removed in the next release.
- Changed properties
The domain field in each .properties file has been updated to only use
a set of predefined values. The old domain values have been appended
to the notes field.
- Changed install
- The script for installing the corpus has been updated to allow installation
of specific systems and/or versions, as well as listing available systems and versions.
See the documentation for install.pl for more details.
- Previously, most second level source archives (i.e. source archives within the
compressed archive) were not unpacked. In this distribution, all versions
of jung, junit, myfaces_core, weka and xmojo will unpack these archives into the src directory.
In addition, the latest version of eclipse_SDK will also unpack its source archives. The packed
source archives still exist inside the compressed archive and instructions for retrieving them
are included in the QC_README text file within each version directory.
- The installation procedure for some systems (ivatagroupware, james, jspwiki and webmail)
has been modified to always install the "reduced" version.
- Changed src
The source code for the following systems are now unpacked from second level jar archives. Previously, these
archives were left packed.
- eclipse_SDK (latest version only)
- Changed docs
A new criterion (most recent versions are
final) has been added (and the criteria discussion has been tidied a
little). New documentation for the new metadata (and relevant software),
and changes to other parts of documentation impacted by this change. Various