Hi fellow developers,
I have committed modifications of files to trunk (trunk Rev. 2895) that address ...
a the major mis-configuration of the .NET Remoting
other important Server configuration for the
proper 'reclaiming of server memory'
...both of which I had discovered recently!
The previous settings in the Server-postgresql.config file were for the convenience of you - the developers - but were not suitable at all for 'production' (nor for 'testing' by the testers)!
Once you create a new Branch or perform a merge from trunk you will get the corrected settings.
In order for you to still be able to do debugging of the Client you will need to copy the file 'Server-postgresql.config._my' (committed in '\inc\template\etc') to 'Server-postgresql.config.my' and then either run 'nant initConfigfiles' or simply start the server with 'nant startPetraServer' to get the values that you will need for that into the '\etc\Server.config' file. [In case you already have a 'Server-postgresql.config.my' file you will instead need to check whether you have the settings that are in 'Server-postgresql.config._my' in your 'Server-postgresql.config.my' file (likely the case). In case this isn't so then you will need to merge such settings into your existing 'Server-postgresql.config.my' file.]
The values that you as a developer need prevent the 'vanishing' of server-side Objects and indeed the 'vanishing' of a Clients' AppDomain while the OpenPetra client is halted at breakpoints or being 'stepped-through' in the IDE's debugger (in such situations the OpenPetra client is not able to 'keep objects alive' on the server as all its Threads are suspended, and the keeping-alive happens through a Thread).