Re: install_initd behavior for non-LSB init scripts.

On Fri, Oct 17, Bart Whiteley wrote:

> This makes absolutely no sense.  What if the script 
> has nothing to do with the scripts immediatly before
> and after it?  

This is the disadvantage of static numbers: Tools don't
know, if the scripts before and after it have something
to do with it or not. Use LSB dependencies, with this
your problem does not exist.

> S80 is far better than S01 (before the network
> interface even comes up), regardless of what is 
> immediately before or after S80.  

Static numbers are always wrong.
And since a very long time, we move this scripts to the end,
not start.

> Also, a collection of third-party services could 
> at least be in the proper order within themselves.  
> 
> S50acme-a, S51acme-b, S52acme-c, etc.  
> 
> Maybe these services need to start in this order, 
> but don't care what other things come before or after
> them.  SuSE's install_initd could potentially set one 
> or all of these to S01 and break the acme suite. 

Than the suite needs to be fixed. You cannot expect that
something, which is not designed for a distribution, works
with this if you use something completly different than the
distribution is using.

> 1. Why S01...  ?  What do you gain from moving
> it here?  

Nothing.
 
> 2. What harm is done by leaving it where it is?  

The same thing can happen as moving it, it will not work.

  Thorsten

-- 
Thorsten Kukuk       http://www.suse.de/~kukuk/        [email protected]
SuSE Linux AG        Deutschherrnstr. 15-19        D-90429 Nuernberg
--------------------------------------------------------------------    
Key fingerprint = A368 676B 5E1B 3E46 CFCE  2D97 F8FD 4E23 56C6 FB4B