> I have to respectfully disagree here.
> When something is being "released" and it's not even production, then yes,
> should read them. In reading them, you would discover the wrapper script. I
> know because this is how I discovered it, and I am not a Gimp devel, just a
You are right: you _should_ read them. But reality is that only a few people do
this. And finally we end up with threads like this where a program even does not
start out of the box 'cause it's grabbing a wrong library.
As I said the wrapper script is not only a problem of the development version,
you need this wrapper also with the stable version of The Gimp when installing
it in a non-standard location. Take an average Windows user: usually he runs an
install.exe, can store the application at any place on his harddisks and at the
end of this procedure the program often is even launched ... Done.
Sorry, but I stick with my opinion that the problem is caused by the
procedure of The Gimp. It should install the wrapper by itself. And it could be
done easily: create a wrapper template, during make or make install replace the
path with the --prefix passed from configure, install it and print a message
where the wrapper is so that the user know what to invoke.
Gimp-user mailing list