The MySQL GUI team is currently working full steam ahead on the much
anticipated MySQL Workbench project. Development is on schedule and the
first Beta is to be expected in Q3.
The application's core - the GRT Environment allowing for full
scriptability and extendability via plugins - has been advanced to
version 3. It now features a C++ module loader together with C++ header
generation. This makes it possible for us to convert all application
logic from the old C and Lua code to type safe C++ modules. This will
greatly improve stability and maintainability. Further, the thread
handling and event system has been shifted from the UI to the library
code using glib and sigc++.
In contrast to the current tools we now draw a much cleaner line between
the back-end and front-end implementation. To make this work with
managed .Net/C# UI code on Windows we have to wrap the C++ back-end
libraries with C++.net wrapper classes. This turned out to be much more
straight forward than expected (a step that is not required on Linux).
The rewrite of the OpenGL canvas using the Cairo graphics library has
been almost completed and gives promising results on OS X and Linux. We
are currently working to get the Glitz compositing library working well
on Windows - thanks go to the Mozilla team for their major work in this
respect.
We are also making good progress with reusing the SQL grammar from the
server for the reverse engineering process. This will guarantee a much
improved handling of the SQL syntax throughout the tool.
The new Subversion source repositories will be published together with
the very first Beta in Q3 and we are hoping to receive more patches from
the community than before. These are now easier to write because of our
switch to standard tools on Windows and our move to cleaner C++ code in
general.
The same is true for plugins which should be written preferably in Lua
and C++. We will also try to update the Python support and get it back
into the Workbench later this year, as requested by members of the
community, including Google and others.
Let me repeat that those who are anxiously awaiting the final release of
the MySQL Workbench will not be disappointed and that the tool will be a
major step in Open Source database application development.
