FIXED: Re: [Bug 57571] New: crash on calling code completion

[jalal]

I've had this bug for a while now, and it seems to be fixed by 
rebuilding the databases formed by "r++".

A couple of points tho.
It seems like the program shouldn't just fall over on something as basic 
as code completion just because the database is made with the wrong 
version of r++.
r++ won't run if there is an existing database (it should maybe 
overwrite the old one?), nor if the directory doesn't exist (it could 
create the directory, neh?).

Otherwise, now it works brilliantly...

cheers
jalal