[Opensync-users] OpenSync release 0.36

[Daniel G. <dg...@su...>]

Hi everyone,

yet another E.X.P.E.R.I.M.E.N.T.A.L. release of OpenSync. This release is=20
intended for developers, testers and for packages. OpenSync 0.3x releases=20
aren't intended to be stable. We don't recommend this release for productiv=
e=20
use. We don't recommend this release to be packaged as part of a=20
distribution.

=2D-- end of sad/scary disclamer ---

Progress!

25 of 28 Unit Tests are ported and adapted for OpenSync 0.3x API. 25 of the=
 25=20
Unit Tests are passing today.=20

Of the 25 ported unit tests, we're running 230 testcases which covers 67.16=
%=20
of our codebase of the OpenSync framework (no plugins!). Which is only the=
=20
beginning.

We hit an not OpenSync related issue with 2-3 Units which running=20
multithreaded and/or forked tests which break the used Test Framework. We=20
already found the issue within "check" (the Test Framework) and try to fix=
=20
the issue right in the root cause. Please ignore failing tests=20
with: "check_pack.c:107: Bad message type arg" for right now. This issue=20
appears randomly and doesn't influence you if you're not interested in=20
running the OpenSync testsuite (but you should it's awesome!)

Changes since last release:

=2D improved error handling:
  * no slowsync if connect/disconnect fails
  * proper error reporting
=2D timeout handling:
  * no deadlocks if plugin get killed/fails (Error message: "Timeout.")
  * This is brand new and rarely tested, if you get an "Timeout." error,
     Please open a ticket and report also when this "Timeout" happend. While
     connecting, disconnecting, getting changes, committing changes?
=2D improved use of main sink
  * intended for object neutral actions like connecting/disconnecting others
  * no need to lock several connect requests of different sinks
=2D reintroduced detectors while format conversion process
  * osync_format_env_find_path_with_detectors()
  * osync_format_env_find_path_formats_with_detectors()
=2D added facilities to add format conversion configs
  * Allows to parse an individual configuration to the format converter
  * Mostly for enabling format converter extensions, tweaks, workarounds
=2D lots of testsuite improvements
=2D various bugfixes

I want to thank everyone who volunteered for testing OpenSync. Especially I=
=20
want to thank the regular contributors of OpenSync: Alban Browaeys, who fix=
ed=20
lots of format related testcases and reimplemented detectors and introduced=
=20
the format configuration facilities. Felix M=F6ller for keeping our bug tic=
kets=20
clean and assign them to the correct people and component. Bjoern Bricks fo=
r=20
his help with the OpenSync Build Environment to keep it nice and shiny.

Next steps:
=2D Final XMLFormat Schema breakages before 0.40:
  * N levels (childs) in the XML trees=20
     (patchset is ready, see announcement on opensync-devel)
  * Merging/Demerging based on the value of attributes.
     Example: Demerge all number which type CELL: <Telephone Type=3D"CELL">=
=20
=2D Common Plugin configuration XSD for all plugins
  * to make further update processes easier
  * to allow UI developers to write generic config parser and widgets
=2D Implement update facilities to update/convert user and plugin configura=
tion
  * Should we support updates from 0.2x to 0.4x?
=2D Update the OpenSync Whitepaper (already in process)
=2D Write more testcases!
=2D Write API documentation (doxygen)

As you can see the todo list isn't that long anymore. And the core of OpenS=
ync=20
=46ramework is getting more stable, day by day. After breaking the XMLForma=
t=20
schema and doing discovery implementation for plugins, nothing should be le=
ft=20
to do a 0.40 release and get finally rid of 0.2x.

Release tarballs:
http://opensync.org/download/releases/0.36/

best regards,
Daniel