Hi Cyril, > Excellent. It looks obvious according to the documentation, I don't > understand why I haven't seen it before submitting the bug.
Well, the man page is certainly not encouraging on the subject. > Was it documented a long time ago? It dates back to Feb 2002 and is in the stable version. > It would be nice if the initscript could check if dictd is supposed > to run from inetd or not with something like (inspirated from > init.d/exim4): Thanks, that sounds like a good idea. (In practice, a stand-alone dictd won't start with port 2628 attached by inetd anyhow, so modifying /etc/init.d/dictd is the correct thing to do, but isn't vital to the process.) I will remove another hurdle by adding "allow inetd" to the access specification of the default /etc/dictd/dictd.conf. I agree that, contrary to the tone of the man page, dictd can be run from inetd, but the upstream author feels that most desires to do so are misguided. dictd memory maps the dictionary database files, which gives it the appearance of being a memory hog, but that mmap'ed memory is not locked. Kirk -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
