I agree. In particular note that rmname and rmelem are completely
different operations:
- An rmname will never create an orphan, at least not in any easily
detectable way. Assume you're going to remove a directory called
d from a parent directory p. If d existed in earlier versions of p then
rmname doesn't really "remove" anything - it just breaks a link from
one version of p. The links from the older versions still exist. If d did
not exist in earlier versions of p (i.e. you just did a "ct mkdir") then
d will be moved to lost+found, along with any contents that it has.
The files in d will still be accessible via lost+found - note that the
files in d will not themselves be in lost+found, just d.
- I disagree with Christain on one point - I feel strongly that engineers
should never use rmelem. They can use rmname, then once every
few months I go in and use rmelem to clean up lost+found. If you use
this approach then you don't need to worry about rmelem ever creating
orphans.
I have a feeling that we're missing something in your request - if you
can explain more clearly what an orphan is and why you want to avoid it,
then maybe we can give a more appropriate answer.
Lezz Giles
Christian Goetze wrote:
> On Mon, 23 Jul 2001, Schiltz, Pete (MED) wrote:
>
> > Does anyone have a pre-rmname, pre-rmelem trigger to prevent
> > orphans, that does more than just check to see if the directory
> > we're about to remove has any children, and fail if it does,
> > and that you'd be willing to share?
>
> Why do you want this? If I accidentally create an element that I don't
> want, why should I be forced to keep it?
>
> If my removal "orphans" the element, then it is very likely that this
> element has no purpose anyway.
> --
> cg
>
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
>
-- Leslie Giles Software Release Manager Avici Systems Inc lgiles@avici.com Phone: 978.964.2030 Fax: 978.964.2100"Do not fear the merge"
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
This archive was generated by hypermail 2b29 : Tue Jul 31 2001 - 22:04:00 EDT