[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [PROPOSAL] Cron jobs

On Fri, Mar 12, 1999 at 12:04:46PM -0500, Cristian Gafton wrote:
> No, it does not. It is one thing to ask an user to take a look at a script
> somewhere in /etc to modify it for his system and another thing to ask
> him to hack soime script in /usr/bin. There is the argument that those
> cron scripts don't have necessarily to go in /usr/bin or /usr/sbin, that
> one can create an additional directory for them, at which point I fail to
> see what we saved over not going with /etc/cron.daily in the first place.

Okay, I'm persuaded.  Limit the system-specific configuration to /etc
and its subdirectories, and I'll quit arguing this point.

> anacron is perfectly compatible with vixie-cron, so there are no worries
> about switching and maintaining compatibility.

Yes, it is, but the "better mousetrap" might not be.  Anacron hasn't changed
in a while, which is good.  If it's maintainer moves on, though, we've
a problem, but I can raise the same issue of vixie-cron, too. 

> > Is specifying the behavior we want worse than specifying a particular
> > implentation?
> No, but the problem is that we can not always define things in some
> standard way and it is much easier to tell people exactly waht we are
> after....

Noted, and agreed.  On the other hand, that's what "standards" are supposed
to do.  Nope, it isn't simple, and the result can be an impenetrably difficult
to read standard, but you've not locked yourself into a single product.
Specifying behavior and letting vendors implement it as they see fit is 
a wiser, more general approach.

All software expands until it can read email.

Reply to: