"Harry Johnston [MVP]" wrote:
> There isn't actually anything to uninstall. The add-on is built into .NET, it
> isn't a separate component.
That doesn't even make sense. How can an add-on for a third party
application not part of Windows be considered to be part of the .Net
framework? 70+ per cent of all computers don't even have Firefox installed.
Building this add-on into the .Net framework SP1 appears to be a choice made
to trick users into having Microsoft add-ons in their non-Microsoft browser.
This is one of the best examples I have ever seen of Microsoft using its
power and trust as owner of the operating system to push their other products
onto a public that doesn't want it. If the public wanted it, or if Microsoft
believed for a minute that the customers wanted Click-Once in Firefox,
Microsoft would simply have added the add-on to the Firefox site and
published announcements on their own site. This add-on wasn't even published
on their own site, no banners, no ads, no press release. Only a 3:00 AM
installation similar to Search 3.01 in 2007:
> ... like most other extensions? (The difference, of course, being that most
> other extensions also *install* per-user.)
> What I want to know is how a user is supposed to reinstall the extension having
> previously uninstallled it. I'm going to have to sit down and figure out
> exactly what KB963707 is doing, although knowing that the uninstall is per-user
> is a good start.
And this install should have been per user and should have been installed
only when the user specifically requested to have the installation performed.
And if a user wants to reinstall the add-on after uninstalling, they could
easily go back to the Firefox site and download t he add-on. Or they could
go to Microsoft's download site and request an installation package. Except
that Microsoft did not make an installation package so they could hide this
add-on in a security patch and pretend that it doesn't fit the model of any
other software package or add-on. Really now, not even an MVP can be so
gullible as to believe the installation and uninstall model for this package
was anything other than a way to get this onto the PCs of users who would not
have otherwise installed the package.
> From what I do know, the new system seems like an unnecessary mess. The
> correct solution to the underlying issue would have been for Firefox to disable
> all new global add-ons by default, so each user gets to opt-in. This would
> solve the problem for all add-ons from all vendors, not just this particular
> add-on from MS, and without creating complications.
Oh, I see. It's Mozilla's fault that Microsoft did this because they
allowed it to happen. And the kid who steals a candy bar at the grocery
store isn't at fault; the store is at fault because they didn't put the candy
> > This is a virus. Computer viruses are generally defined as programs that
> > are self propagating and this bug does that - even if only within your one
> > machine.
> Nice rhetoric, but factually incorrect - there's (presumably) only ever one copy
> of the add-on code, it isn't being replicated. (At least, that is certainly how
> the original add-on worked, as I said I haven't analysed KB963707 yet.)
Yep. You got me there.
> Once again, an add-on isn't a patch. The Firefox installation itself is *not*
> being modified.
Yep. Got me there, too. Semantics.