After finishing my SoC and relaxing a little bit afterwards i have been
back hacking on beagle for the last two weeks. I was curious about how
to use NUnit tests and monodevelop. So i decided to try and work with
both on the things i wanted to improve of my SoC project (will post
later about that).
Because there was some discussion about using MonoDevelop some time ago
i thought i might share my experiences. I used MD not only to browse and
edit the code but also to build beagle (or at least the parts i am
interested in). MD can handle dependencies between the different
libraries and build them in the right order without using the makefiles
etc. This is done by splitting the program (solution in MD) up in
different parts (projects in MD) that have their dependencies etc and
create one library each.
MD can use this structure to create the autotools files out of it but it
can't read them to load the structure. So it's quite a bit of work to
create all the projects and dependencies etc. but then it worked fine
for me.
The Project and Solution files of MD are in xml format and easy to edit
by hand. So it's easy to include changes from the autotools files in the
MD files as well. MD also supports different so called configurations
which allows you to define different symbols for debuging and release
etc. But one thing that's really missing - especially for beagle is the
equivalent to --enable-blablub flags for configure. IIRC i read some
discussion about implementing that somewhere on monodevelop sites but it
looks like that might take some time.

There is a nice NUnit plugin to MD that still is a little buggy but
worked quite well for me after some playing around. It can also display
how long the tests took - guess that's not really exact but it gives you
a clue how your changes might affect performance.


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]