rawhide report: 20060716 changes

[Erwin Rol]

On Mon, 2006-07-17 at 20:39 +1000, Rodd Clarkson wrote:
> On Mon, 2006-07-17 at 08:39 +0200, Erwin Rol wrote:
> 
> > Of course it is a bit worrisome that a CTRL-c to stop a yum update can
> > leave behind such a mess, instead of cleaning up more correctly, like it
> > can when a single package stage (the killing of nash) fails. I mean
> > noway in hell would "my mother" been able to fix this system, and those
> > ppl (like i did) certainly will first try to hammer CTRL-C when
> > something hangs. 
> 
> While I appreciate where you're coming from, let's give what's happened
> a little perspective.

And where do I come from?

> Firstly, this has happened in rawhide.  It's unlike your mother would be
> using rawhide and rawhide is rumored to eat babies.

Yes exactly, it is happening in Rawhide, the place to test things and
point out things that could possibly be done better, _before_ my mother
has to use it.

> This hasn't happened in a released version.  It hasn't even happened in
> the testing stage of a released version.

Rawhide is testing, so it did happen. If I don't point it out it might
stay there until it is released. 

> So while it might be a little frustrating, it's not earth shattering.

When someone does a update, and somehow stops this by CTRL-c (or a
system shutdown would probably have the same result) yum will not
recover cleanly, and cleaning up the mess after it needs way more
knowledge than the average user (needs to) have. That was the only point
I wanted to make, and I don't think it is that much out of perspective. 
So if you don't have anything constructive to add to that, why not
simply ignore my mail and go back where you came from.

- Erwin