Dale 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?
Whether you "consider it to be" part of the .NET framework is a matter of
opinion, I suppose. Functionally, the code is located within the .NET code.
Since .NET is monolithic (in the sense that it has no optional components, thank
goodness) this means that there is no reasonable way in which it could be
removed, short of making it an entirely separate product.
> 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.
I'm not sure there's much point in speculating over Microsoft's motives.
Personally I'm not convinced it was an intentionally malicious move, though I
think an ill-advised one. On the other hand, given the existing precedent I
suspect the outcry caught them by surprise.
It does seem surprising that there hasn't, to the best of my knowledge, been any
fuss at all over the fact that Adobe Reader installs a Firefox global add-on
without asking, or providing any way to uninstall it. Perhaps this is because
it's been doing it for at least the last four years! I'm not saying that this
excuses Microsoft's choices, but sometimes it really does seem that they get
picked on. :-)
Then there's the fact that Apple Quicktime does the same thing, *and* also
modifies Sun Java without asking - that last one used to be really nasty,
because it wasn't even done properly. At one point it actually broke Java.
No doubt there are other examples.
> 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. [...]
Not an ideal solution, as it creates additional work for the end user. Not
everyone is comfortable with this sort of task.
> 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.
Technologically and in terms of usability including browser add-ons in the
product they relate to is a superior technique. This avoids people having to
download and install the add-on separately, and avoids people who don't have the
product (.NET) installing the add-on and then wondering why it doesn't work.
(Again, I'm still not saying it was a good idea in the absence of an opt-in
solution.)
>> 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.
No, in my opinion it was a mistake for Microsoft to choose technological
simplicity and usability at the cost of user opt-in. But Mozilla could fairly
easily make it unnecessary to choose between them ... and simultaneously put the
opt-in choice in the hands of each user rather than the administrator, another
issue with the current model.
I should also note that I was presented with a dialog the next time Firefox ran
advising me of the new add-on and giving me the option to disable it. I'd have
preferred it to be opt-in rather than opt-out, but other than that I'm not
entirely sure what all the fuss was about - did this not happen in some cases?
>> Once again, an add-on isn't a patch. The Firefox installation itself is *not*
>> being modified.
>
> Yep. Got me there, too. Semantics.
Not at all. This is an important distinction because it means that disabling
the add-on is just as good as uninstalling it, making the behavioural changes
introduced by KB963707 entirely pointless IMO. If it had made it opt-in rather
than opt-out that would have been a positive move, but simply including a
per-user uninstallation option?
(Note also that keeping the add-on code within the product the best way to
provide add-ons, because it ensures that the order of installation doesn't
affect the result.)
Harry.
-----------------------------------------------------------------------------
Less Spam Better enjoyable experience
Visit :
news://spacesst.com
Linear Mode
