On Mon, 16 Dec 2002 15:53:27 -0700, Bernard Johnson
<[EMAIL PROTECTED]> wrote:

> Perhaps I have missed this in an earlier discussion... Is there a
> particular reason that tmda-ofmipd does not fork to the background
> like a normal daemon would?  I have a patch that I'm using on my
> system now that makes it do this, as the init scripts parts of redhat
> seem to expect that behaviour. 
> 
> Is there a particular reason it doesn't do this already?

According to DJB (not that I want to put the man up on a pedestal or
anything, but there's no denying that he speaks with knowledge and
experience), it is bad software design for a daemon to put itself into
the background.  A well-written daemon should keep its own thread to
better allow the spawning program to monitor it (see
http://cr.yp.to/daemontools.html for a good example of a supervisory
program).  In fact, DJB even mentions how to work around poorly
written daemons that do disconnect themselves by using a pipe:
http://cr.yp.to/daemontools/faq/create.html#fghack

Gre7g.
_________________________________________________
tmda-workers mailing list ([EMAIL PROTECTED])
http://tmda.net/lists/listinfo/tmda-workers

Reply via email to