Today I spent quite a bit of time porting over the project to Windows. This indeed may sound shocking. It involved installing Qt and Boost and building the protocol library using mingw and msvc. I can’t say I was able to get very far with msvc. It’ll take a little bit of work before it compiles in Visual Studio. I also had to install KDE and the KDE libraries to get Parsek to build. It was actually really easy. You can find some cool stuff on the KDE on Windows project here and here. So why am I wasting time doing all this just so I can compile in Windows? Truth be told, I’m not really sure. I’m mostly curious to see if Visual Studio or QtCreator + msvc compiler will help. QtCreator can now support Microsoft’s debugger also. All this may sound like blasphemy but I’m hoping the least it could do is provide some insight into the problem. Besides, KDE software is now meant to be cross-platform.

While I was waiting around for installers to finish, I spent a lot of time looking at libtpproto-cpp code. While thinking about the changes I needed to make to get it compile under Windows, I also thought it might be a good idea to rewrite the Boost-based signals for event handling and callback routines using Qt signals. Basically, the idea would be to, as quickly as possible, overhaul the protocol library so it is more Qt friendly and doesn’t use Boost at all. This would avoid my difficulty in finding a solution to handle Boost signals in the Qt application (the error which was causing my compile failure yesterday). This is not to say there wouldn’t be a need for a light Boost-based protocol library so there could potentially be two versions or a single version that can be configured to use only Boost or Qt, not both. Maybe this sounds like a another lousy idea but I figure it would at least help get a better understanding of the protocol library and help in my ultimate goal of handling the TPSocket manipulation and event handling using Qt classes better integrated with the GUI. I’ll try to think this one over and see if it’s up for discussion. Rewriting a library using a different framework and then trying to compile it under Linux and Windows.. hopefully I’ll only make it a short setback.

This entry was posted on Wednesday, July 1st, 2009 at 11:49 PM and is filed under GSOC 09, Parsek, Progress Report, Thousand Parsec. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.