[cdt-dev] Object IDs in .cproject again

[kesselhaus <kesselhaus@xxxxxxx>]

Hi,

I came across some build tools like, which I currently evaluate.
In our current build environment based on make, cygwin, shell tools and 
per scripts, there is also a script which generates a MSVS vcproject 
file out of the information of the make scripts. Since importing and 
setting up a project into eclipse is rather tedious, I wondered if the 
project and cproject file could be created by the build scripts.

The problem I currently have is, what to use where all the Object IDs 
are used.
Did you guys rather plan to replace the Object IDs with a normal name?
Why are these Object IDs actually really necessary?
Wouldn't a project based name be better?
Also, would it be better, if e.g. the indexer data be stored in the 
.settings directory within the projects directory instead of in .metadata?
I guess the IDs are necessary, because the data is stored in .metadata 
and not within the project.

I think this was discussed previously, but I still think, it makes 
creating project files outside of eclipse very hard.

Regards,
Henning