weird update problems

[Ed Smits]

especially in a production environment you don't install any patches,
even security patches,  w/o testing them first on a non-critical
server. Because I assume that most of us aren't running production
Ubuntu servers we can be a little more lax, but let's face it, one of
the reasons why we are using Linux over Windows is that the security
concerns are a lot less critical to us - there may be security holes
in Linux but the chances of anyone exploiting them are a lot less than
with Windows machines, and so it behooves us (good word, eh<G>) to
patch our machines intelligently. I'm as bad as the rest - I also
automatically want to patch as soon as Update Manager says there are
patches available, however because I run VMWare inside Ubuntu (and
need it for work) I have learned the hard way not to patch anything to
do with the kernel until all associated parts (headers etc) can be
patched as well.

Lets hope this is cleared up by tomorrow



ED

On 2/8/07, Richard Mancusi <vrman49 at gmail.com> wrote:
> On 2/8/07, Zachary Stern <rollerskatejamms at gmail.com> wrote:
> > For me it IS about always having the newest version, because even though I
> > use linux for work, I like to have the newest apps on my personal computer
> > just to check out the new features. I do a backup before any major patching
> > just in case.
> >
>
> That is a good reason and method for updates.  I also do not understand
> the "... weird obsession ..." comment - but for a different reason.
>
> There are obviously several ways to do your updates.  In my case I use
> Update Manager or simply wait for the notification.  In both cases you
> end up the same place.  A gui that breaks the updates into categories.
> e.g. Proposed, Backport, Important Security, etc.  An argument can be
> made to ignore or wait for many of these categories.  But I fail to see
> the logic behind ignoring "Important Security Updates".  The updates in
> question today were part of these security updates.  As a user we should
> be all over those.  And if you are the administrator of a corporate system
> that fails because you ignored those updates you should be fired.
>
> The real question is how do updates in that category get released
> without adequate testing.  Yes, we are human and make mistakes.  But
> this is not the first time that this has happened.  This forces all of us
> to ignore security updates until proven sound.  But if we all ignore them
> they will never be proven anything.  Note the infinite loop.