Barry Watzman said:
In my opinion, SP2 is a great upgrade.
However, also in my opinion, Microsoft was too confident (incorrectly, as
it turned out) that virtually no users would have any major problems with
it.
A recent survey of 44,000 SP2 ugrades showed 90% had no problems.
That percentage is not equal to "virtually no users would have any major
problems with it." However, I never saw a claim made by Microsoft that
virtually no user would have major problems. Do you have a reference
for your wording from a Microsoft source? I can't believe they said anything
that stupid, but I suppose it is possible... computer just don't work that
way.
In Microsoft's defense, the amount of testing that they did was
unprecedented, and I don't think that there is anything more that they
could have done to have made this release more "problem free". Part of
the "problem" is that the user base is not "techies", and doesn't even
know what a BIOS, Prescott or Firewall are, much less how to use any of
them.
No they are not techies. But I don't think they need to be techies, just
computer literate. Backups are not difficult, and I think computer literacy
is enough to make a backup. So again I disagree with a position that you
use (the other description is "virtually"). The "problem" is that though 99%
of users have heard advice stressing making backups maybe 10% practice
making backups. They have the 'can't be bothered attitude' or the 'it will
never happen to me' superiority complex. I don't think it is Microsoft's
fault or responsibility that people have to learn the hard way.
The lesson here is that you just can't release an upgrade of this
magnitude which is going to be so problem free for everyone that you
should go out and tell the entire world to just turn on automatic updates
and get it automatically. Hopefully, this is a lesson that MS has now
leared as well.
Again, I think there is a problem with this conclusion. In your words:
In my opinion, SP2 is a great upgrade.
I think that means it should be released. Now by what method? What
is the alternative method that eliminates the upgrade from reaching
the hands of people who don't make backups (so that they can fix
any resulting problems) or discriminates their level of computer
expertise. Should they fill out a questionaire in order to qualify?
I think automatic updates is as good as making a large download
or ordering a cd. Yes, a few people have automatically download
and install updates enabled. And if that were not enabled they might
choose not to install SP2; they might decide well I don't have a backup
and I'm a computer dummy so I better hire somebody or just do without.
But any update has the potential of causing a problem on a computer.
So that means all the beneficial updates will not get installed (if you
turn off automatic download and install) by all people who don't know
enough to login into Microsoft and check for updates. There is always
going to be a gap in whatever release method is chosen.
Computers have not evolved to the state where the users do not have to
take responsibility for how the computer is used. If a user knows that
they have not made a backup (and very few users are unaware that they
ought to have a backup) and they are not very competent with using a
computer, then they can hire somebody or turn off automatic updates.
I think it would be a ghastly moral mistake to deprive some users from
this upgrade in order to protect some users like Jim, who exactly deserves
his problem. The "problem" really is that he has no backup of important
email, not that there is an inconvenience of the SP2 installation procedure.
No upgrade can ever be created which will install correctly on all
computers.
So I don't see anything wrong with the MS approch to the best practical
result. It is a common thing in life that there are no rules available to
find
a perfect way of taking census in order to apportion government seats
by population. Neither the census nor the apportioning method.
I mean do you think MS should learn the lesson to turn off all automatic
updates? SP2 is just a collection of updates. Which ones should not
be made available for public distribution? Many of the SP2 updates
were released prior to SP2. There is no possible way of predicting
which update will cause a problem on which user machine. That is a
mathematical principle. Since you might think I am overstating the
case I will provide a reference (also since I asked for one from you).
http://www.cis.udel.edu/~case/colt.html
I notice you did not provide an alternative solution.
Much of this is associated with game theory,
Stephen
