Re: Failover in copy...

[Chip Seraphine]

On Thursday 16 September 2004 11:59, Christian Pearce wrote:
> I know I talked about this before.  But isn't this just the inverse of
> define?  If the copy failed the define class wouldn't be set and the
> failover would.  If the copy was successful just the opposite.  Isn't
> this redundant?  All you have to do is ! the class and you have the
> other condition.  Is there something about failover that I am missing?

Order of operations.  Does "!copygood" mean that the copy in question has not 
run yet, or that it failed?
 
> I think it might be of interest to have it try a second IP address.  So
> the server would take comma separated values.

That would be extremely, massively cool.  I would recommend following the 
iterator conventions rather than introducing a new one, however.

-- 

Chip Seraphine
Unix Administrator
TradeLink, LLC
312-264-2048
chip@trdlnk.com