It's positively encouraging to see usability mattering again,
and the "mozilla.dev.usability" newsgroup now even come alive again
in my version of TB -- for ages I never saw a new post appear in it,
for no reason that I could fathom, except for possible death of the group.
I normally don't try to mix topics, but not long ago I had an impulse
to post a short, multiple-choice quiz, asking whether TB is
that type of client which can support only one user per computer,
or is it designed for multiple users, and even as to each user,
can (s)he have only one instance of a TB profile, or many?
Once the non-controversial correct answers were agreed upon,
my next question was to ask how many copies of the program
normally exist per machine, and then, if by any chance
there were to exist an option by which the computer owner
were to be able to tell TB not to update that single copy,
where would be the best place to store that option --
would it be in just one of the many profiles
which just one of the many users might be making active,
or might it better be located with the program files
or "Registry" (for Windows), so that it could have any hope
of really having any effect to protect the single copy of program files?
Being the maker of outrageously graphic (but memorable) analogies that I am,
I compare what occurs in my computer with these unstoppable updates
to "date rape," where if she tapes a sign on one breast to keep hands off,
the guy just grabs the other one, and even if finally signs are pasted
over most every body part, then the guy simply tries groping the
date's sister (any other user) instead. I often have to test or start with
TB's original state via an empty new profile folder, which is pretty much like
running unclothed through a bar filled with gropers -- well,
I'm sure everyone gets the idea: why isn't said option
saved in a place that makes more sense for a product
already designed for a multi-user, multi-profile-per-user system,
but having taken no heed of that fundamental fact
in its useless and ineffective location for any
"don't try inseminating any new version into my computer
without my explicit permission" option, entirely defeated
if even a single profile (on a remote share, say) ever remains unprotected?
Making even "about this program" start causing an update
is bad enough, but the above even dwarfs that issue
of "whose computer is this, anyway?" and by the way,
I never even was asked to "agree" to any of this policy,
which is another reason to consider it as outright "computer rape."
--