Release Notes

This page maintains the FOSSology Project release notes, from the most recent release to the oldest release.

Date: June 2009

1. Email notification of job status: If enabled, FOSSology will email users when their jobs finish. This option will be available on a per user basis. For more information see how_to_use_email_notification.
2. Roll-up of many bug fixes.
3. Several new license templates added.
4. Code cleanup for improved efficiency and maintainability.
5. Tutorial section, with examples, added to fossology.org.
6. Lots of new tests added to the automated test suite.
7. RPM packages for RHEL4, CentOS4, RHEL5 and CentOS5 - see a Simple List of all RPMs Needed for Installing Fossology on RHEL CentOS

• Multi-System updates
  • When upgrading from previous Debian package of 1.0.0, you have to manually update your /etc/Scheduler.conf file to include the new fo_notify agent. For a typical single-system install, the new line in Scheduler.conf will look something like

agent=fo_notify host=localhost | /usr/local/lib/fossology/agents/engine-shell fo_notify '/usr/local/bin/fo_notify %{*}'

• Single System Updates
  • The update should regenerate your Scheduler.conf file to include the new email agent. If it does not, use the line above to manually update the Scheduler.conf file.

• Scheduler -t output changes. For release 1.1.0 when performing a scheduler test the user will now see the scheduler killing the agents during the test. This is normal behavior. It is only the output that has changed since the 1.0.0 release. As long as the scheduler finishes with the statement

2009-05-12 18:21:24 scheduler[28850] : STATUS: All scheduler jobs appear to be functional.
2009-05-12 18:21:24 scheduler[28850] : *** Scheduler completed

then the test passed and the scheduler is ready to run.

Date: 17 December 2008

1. Completely redesigned build and release system that standardizes many file locations, environment variables, and other parameters important to FOSSology. It also conforms to open source software community norms and expectations for software builds and installations; separates the build- time assumptions and dependencies from the install-time, making it possible to create FOSSology packages for Linux distributions (Redhat, SuSE, Debian, etc)
2. Huge optimizations in database schema design which provide a 50x (or better) speed improvement in user interface response times. The time taken to retrieve and display a software license analysis report has been reduced from tens of minutes to just seconds for nearly any size analysis.
3. Addition of license report caching, which lets the system store a snapshot of a previously-run report and supply it to users instantly upon request without regenerating any portions of the report. Combined with the database optimizations, these two features make the FOSSology UI very usable.
4. Huge optimizations in the license analysis algorithm, speeding up license analysis jobs by as much as 10x. A full Fedora DVD release that would have taken a week or longer to process can now be completed in 24-48 hours. (NOTE: analysis times vary significantly based on the system performance, CPU, memory, etc)
5. Major improvements in license detection accuracy: We have spent considerable time refining the license template and license terms that we scan for, so that FOSSology now provides a much more accurate picture of the actual license content of open source software. We will continue this refinement in future releases, and encourage everyone to provide feedback on areas for improvement, but we feel this is a very good balance of accuracy and flexibility in license reporting.
6. Software packaging of FOSSology for the Debian Linux distribution. For the first time, the FOSSology project will be providing a complete DEB package to enable simple installation of FOSSology on most DEB-based Linux systems.
7. A new “self-test” agent that allows the FOSSology system to validate that it was installed and set up properly, and report any configuration issues to the system administrator. Also provides run-time configuration checking to make sure that filesystem permissions, database connections, and other services critical to FOSSology are up and running.
8. An automated test suite based on the open source SimpleTest framework, that includes tests to verify that major parts of the FOSSology web-based User Interface are operating correctly. These tests can be run by anyone installing the FOSSology system, to verify proper operation, and also by anyone developing new features and functionality in FOSSology, to test their work and ensure that it does not create any new bugs in the existing code.

As always you can install FOSSology from source - follow the link to Download to get the tarball from Sourceforge or the live subversion repository. Read the INSTALL file for detailed instructions.

But we're happy to announce the availability of Debian packages! We have a (temporary) apt repository available here:

deb http://fossology.org/debian/ ./

If you add this line to your Deb-based system's sources.list file, then you should be able to simply enter

apt-get install fossology

And the world will become a better place.

The Project Builder (http://trac.project-builder.org/) project is working on RPM packages for FOSSology which should eventually work on most RPM-based systems such as RedHat, SuSE, Mandrivia, etc. These packages are not yet fully tested by us or them! Stay tuned for more information.

There are now three key documents distributed with the FOSSology source code:

1. README provides a general description and introduction to FOSSology
2. INSTALL provides detailed instructions for building, installing and configuring FOSSology from source code. (INSTALL.multi provides details for setting up a cluster of multiple FOSSology servers)
3. HACKING provides details about how to modify the way FOSSology is built; this is primarily intended for advanced users, software packagers, etc

As always you can view the complete project documentation online anytime in user documentation. For developers or other technical users, you can peruse the developer documentation.

* On Fedora systems, there is a problem with the authentication method FOSSology uses by default, which is different than the Fedora Postgres package configuration. As a result, an out-of-the-box FOSSology install on Fedora will not work correctly. We are working on a fix now.

* RPMs for FOSSology 1.0 are being built now, and may be available from the Project Builder project. However these are “alpha” quality for now, and need considerable testing.

Date: 2 July 2008

Please refer to the README file contained in the fossology download for first time installation instructions.

If you are updating a previous fossology installation, please follow the update process below.

If the fossology scheduler is running, stop it with:

sudo /etc/init.d/fossology stop

Uncompress the fossology-0.9.0.tar.gz file into a build directory and build.

tar -xzvf fossology-0.9.0.tar.gz
cd fossology-0.9.0
make
make install
sudo ./install.sh -f

After successfully installing 0.9.0, you will need to generate a new scheduler.conf file using

/usr/local/fossology/agents/mkconfig

Invoke mkconfig with the -? option to see which options you should specify for your environment.

Restart the fossology scheduler with:

sudo /etc/init.d/fossology start

Open a browser & browse to the fossology repository:

http:<fossology host>/repo/

Login to fossology as user fossy.

Move your mouse over the “Admin” tab and click on “Initialize”.

After initialization is complete, select Organize→License→Default Terms

Next, select Admin→Database→Check Templates

Know Problem with “Help → About”

Problem: “Help → About” does not return any version/code revision.

Description: see above

Solution: make tar is not putting the version in the pathinclude.h.php.

This issue will be addressed in the next release.

Date: 28 May 2008

Please refer to the README file contained in the fossology download for first time installation instructions.

If you are updating a previous fossology installation, please follow the update process below.

If the fossology scheduler is running, stop it with:

sudo /etc/init.d/fossology stop

Uncompress the fossology-0.8.0.tar.gz file into a build directory and build.

tar -xzvf fossology-0.8.0.tar.gz
cd fossology-0.8.0
make
make install
sudo ./install.sh -f

Restart the fossology scheduler with:

sudo /etc/init.d/fossology start

Know Problem with attempting to Reset or Delete jobs WHILE they are running.

Problem: Attempts to Reset or Delete a job while the job is running will fail.

Description: Job will be marked as failed, it will continue to run and all jobs dependent on it will run.

Solution: Do not attempt to Reset or Delete running jobs.

In the next release, the Reset | Delete micromenus will not be displayed for running jobs.

Known Problems with cp2foss at 0.8.0 release

Problem: cp2foss can't find fossjobs

Description: cp2foss expects to find a program called fossjobs in /usr/local/bin. fossjobs gets installed as fossjobs.php in /usr/local/bin due to a makefile issue.

Solution: as root create a symlink in /usr/local/bin called fossjobs and link it to fossjobs.php.

This problem should be fixed in 0.8.1.

Problem: cp2foss will create duplicate upload folders.

Description: cp2foss does not correctly detect when a file has been uploaded into a upload folder. As a result it will create duplicate upload folders with the same name. This situation only occurs when the same upload folder is used in the same folder path. For example if foo.c is uploaded into upload folder foo.c and that is loaded into folder path /bar/baz/ and then reloaded again in to /bar/baz there will be two upload folders, both called foo.c.

Solution: Don't do multiple uploads into the same upload folder/folder path. If this situation does occur. The Organize→Uploads→Delete Uploaded File menu option should be used to remove the duplicate(s).

This problem is scheduled to be fixed in the 0.8.1 release.

Date: 30 April 2008

Please refer to the README file contained in the fossology download for first time installation instructions.

If you are updating a previous fossology installation, please follow the update process below.

If the fossology scheduler is running, stop it with:

sudo /etc/init.d/fossology stop

Remove the scheduler.conf file (you may want to save a copy if you have customized it):

rm -rf /usr/local/share/fossology/agents/scheduler.conf

Uncompress the fossology-0.7.0.tar.gz file into a build directory and build.

tar -xzvf fossology-0.7.0.tar.gz
cd fossology-0.7.0
make
make install
sudo ./install.sh -f

If you saved a copy of your customized scheduler.conf file, add your customizations into the newly created scheduler.conf file.

Restart the fossology scheduler with:

sudo /etc/init.d/fossology start

Date: 13 February 2008

Please refer to the README file contained in the fossology download for first time installation instructions.

If you are updating a 0.6.0 fossology installation, please follow the update process below.

• Bug fixes.
• New delagent added to improve UI response time.
• Added a check for corrupt bsam file during install.
• Tweaks to vacuum and analyze in an effort to eliminate SQL timeouts.
• Major debugging of bsam-engine.

Also see Known Issues section below.

If the fossology scheduler is running, stop it with:

sudo /etc/init.d/fossology stop

Remove 0.6.0 license information used to create the bsam cache file:

rm -rf /usr/local/share/fossology/agents/licenses/*

Uncompress the fossology-0.6.1.tar.gz file into a build directory and build.

tar -xzvf fossology-0.6.1.tar.gz
cd fossology-0.6.1
make
make install
sudo ./install.sh -f

Restart the fossology scheduler with:

sudo /etc/init.d/fossology start

Date: 18 December 2007 The initial public release of FOSSology.

Please refer to the README file contained in the fossology download for first time installation instructions.

If you are updating a fossology installation created with pre-release bits, please follow the update process below.

A freshly installed system cannot create a new folder in the UI on the first try. The second and subsequent attempts succeed.

The unpack agent extracts files from containers. A container is any kind of file that stores other files. unpack uses utilities, such as, tar, gzip, ar, rpm, zcat, etc. to perform the extraction. In cases where the container is corrupt or otherwise ill-formed, the utilities may fail to successfully extract. The unpack agent will report this as a warning and skip to the next container. The warnings can be safely ignored. An example of a failed extraction and subsequent unpack warning is:

WARNING pfile 1153340 Minor zip error... ignoring error.
LOG pfile 1153340 Minor zip error(2)... ignoring error.
WARNING pfile 1153340 Command bzcat failed
LOG pfile 1153340 WARNING Command bzcat failed:
/srv/fossology/repository//ununpack/0/916132eba75e9fcb069c7b6327f60b3b318ceef0. \
6ab7c5ecb2f03728c63644e4250bde34.656852992.dir/SRPMS/iscsi-initiator-utils- \
6.2.0.742-0.5.el5.src.rpm.unpacked.dir/iscsi-initiator-utils-drop-bad-events.patch

At peak processing times, the show jobs status window will display jobs in a ready or scheduled state for a long time (typically when processing many projects at once or large projects). This is normal behavior and dependent on the number and speed of the agents running.

The following messages are written to the log file each time the agent pkgmettagetta executes on 64 bit processors.

Loading `libextractor_exiv2' plugin failed:
/usr/lib/libextractor/libextractor_exiv2.so: cannot open shared object file:   
No such file or directory
Loading `libextractor_ole2' plugin failed: 
/usr/lib/libextractor/libextractor_ole2.so: cannot open shared object file:
No such file or directory
Loading `libextractor_ogg' plugin failed: 
/usr/lib/libextractor/libextractor_ogg.so: cannot open shared object file:
No such file or directory
Loading `libextractor_mpeg' plugin failed: 
/usr/lib/libextractor/libextractor_mpeg.so: cannot open shared object file:
No such file or directory

Perform the following steps to update and existing fossology database.

invoke psql as user fossy

psql -U fossy -d fossology

cut-n-paste the following lines at the fossology⇒ prompt:

alter table agent_lic_raw  add column lic_url text; 
alter table agent_lic_raw  add column lic_date timestamp with time zone; 
comment on column agent_lic_raw.lic_url is 'URL origin of the license'; 
comment on column agent_lic_raw.lic_date is 'when license text was captured'; 
alter table jobqueue  add column jq_elapsedtime integer default 0; 
alter table jobqueue  add column jq_processedtime integer default 0; 
alter table jobqueue  add column jq_itemsprocessed integer default 0; 
comment on column jobqueue.jq_elapsedtime is 'Elapsed secs working on jq'; 
comment on column jobqueue.jq_processedtime is 'sum of process times (secs) from all agents'; 
comment on column jobqueue.jq_itemsprocessed is 'sum of files processed by all agents';