>>> if (!SHGetSpecialFolderPath(NULL, tmppath, CSIDL_APPDATA, FALSE))
>>> return false;
>>> snprintf(ret_path, MAXPGPATH, "%s/postgresql", tmppath);
>>>
>>> This needs to be tested of course.
>
>> Uh, by default this directory is not going to exist, right?
>
>The /postgresql subdirectory probably ought to be created by
>the Windows
>installer. I don't think libpq ought to do it.
Um. Then we wouldn't support installs made any other way, so I think
not. Also, this directory is *per user*, not per machine.
But actually, looking at the files, only the history file is one that
would be created by a program - the rest are all supposed to be created
by the user to make a libpq program act in a special way? If I'm right
in that, I would suggest that *psql* creates the directory when it needs
to write the history file. But libpq should never need to, since it only
reads the files.
//Magnus