Re: convergence and undoing changes

[Mark Burgess]

On Mon, 2005-11-21 at 12:27 -0500, Alva Couch wrote:
> A few of you have expressed concern about my bluntness in
> discussing cfengine's limitations. I apologise to Mark
> and the list for this; they deserve better from me.
> 
> I am in awe of Mark and cfengine and do not mean to detract
> from either cfengine or its author. I know that it represents
> the easiest tool to use that is currently available for
> managing networks.  This is why I stopped developing my own
> tools and study it instead; to figure out how to make the 'best
> thing so far' even better.
> 
> I have replaced the talk slides mentioned previously with what I
> hope is a more equitable description of the limitations that I
> am trying to address. Please try not to take these limitations
> as destructive criticism; they are in fact issues outside the
> declared scope of cfengine and may need to be addressed from
> outside of the framework that it provides, through practice or
> external tools.
> 

I should also point out that Alva and I are the best of friends and that
we more often agree than not. As everyone who has seen Alva deliver a
talk knows, he is a marvellous speaker and showman. I think the true
interpretation of the offending slides was likely lost on the casual
viewer due to its lack of accompanying show - with explanation. But it
is good to clear up these issues so that no misunderstandings arise.

I value Alva's work and opinions highly and look forward to implementing
some of them in cfengine 3. Of course, everyone is invited to the
presentation BoF at LISA.

Mark