"Perry E. Metzger" <perry@piermont.com> writes:
> "Michael K. Sanders" <msanders@confusion.net> writes:
> > Is this really the case?  I thought it was generally accepted that for
> > some people with production systems running 1.3.3, an upgrade to 1.3.4
> > would be a lot less painful than an upgrade to 1.4.
> 
> I seriously doubt that -- 1.4 is a big change internally, but not such 
> a big change externally.

"Uh, No."

(1) how many new/different config files are there going from 1.3.3 to
1.4 vs. from 1.3.3 to another release in the 1.3.x line?

(2) how many potential or real new bugs are there, for either
transition?

(3) how is either upgrade going to change the expected behaviour of
the system as seen by users of the system?


>From both a reliability, 'known problems experienced by users of the
production system,' and 'conversion work' standpoint, updating a
production server to 1.3.N from 1.3.(M<N) is a minor undertaking and
upgrading from 1.3.x to 1.4 is a major one.

Trying to claim that 1.3.3->1.4 upgrade isn't a significantly greater
undertaking than a 1.3.2->1.3.3 or 1.3.3->(hypothetical 1.3.4) ugrade
is just rubbish.  If you honestly think otherwise, then i seriously
doubt your ability to maintain a production system...


> People are better off upgrading to it, when
> it becomes available.

This simply does not follow.  it's an assertion, made by you, with no
necessary backing in a real production environment.

see the list of questions above.  also take into account other
upgrade-related factors (disk space usage expansion, etc.).

While it's probably true that new installations are better off using
it (if it's stable; not even that has been shown, since it's not been
released), that says nothing about existing installations.

for existing installations, often "a few bug fixes for serious
problems, and a few minor features" is better than "complete change."



cgd
-- 
Chris Demetriou - cgd@netbsd.org - http://www.netbsd.org/People/Pages/cgd.html
Disclaimer: Not speaking for NetBSD, just expressing my own opinion.