Notes from the OpenOffice.org sprint (Mar 10-12)

[Matthias Klose]

[ sent to both the Debian and Ubuntu lists, please note that
  messages to the Ubuntu list are moderated for non-subscribers ]

Last weekend, Martin Kretzschmar, Rene Engelhardt and I met in Berlin to
work on OpenOffice.org packages. The focus of the sprint was on
addressing issues having an impact on good usability and which still had
a chance to make into the current packages for dapper. Most of the
results are now available in the archive, packaged as version
2.0.2-1ubuntu3 (not on the flight-5 CD). For bug reports, please use the
appropriate bug tracking system, do not discuss bugs on the lists or forums.

- The packages are based on the upstream version 2.0.2, released last
  week.  Compared to the OOo version found in breezy, many issues were
  addressed by upstream and ooo-build developers. Thanks!

- Scanning the Ubuntu bug tracking system, about 2/3 of the 240 open
  reports could be closed on the weekend and the following week.
  Reproducible reports were forwarded upstream, and about a dozen
  fixed on the weekend. Some bug reports found in both the Debian and
  the Ubuntu bug tracker were addressed as well, although we were
  running out of time ...

- The size of the openoffice.org-l10n packages could be reduced to
  20%-70% of the previous size by eliminating duplicate files (most
  of it untranslated template documents). So hopefully we can ship
  more localization data on the CD's.

- The upstream test tools are now available (and installable) in the
  archive. These tools allow doing a quick check, starting all OOo
  applications (install the openoffice.org-qa-tools package, and run
  oosmoketest). The README.qa describes what can be done with the
  other 50MB of test scripts and data. Thanks Martin for packaging
  these tools.

This week we got lot of feedback from the Ubuntu L10n sprint, and some
testing done for asian locales. Outstanding is still additional
packaging of dictionaries and hyphenation patterns. The problems with
these data is tracking down the correct licenses of this data and then
see if we can include it in our archives.

Overall a nice step forward towards more usable packages, so I hope we
are able to repeat such a short sprint in summer/autumn.

  Matthias