Comment 1536 for bug 1

Actually it's much easier than we thought:
the root cause of all this mess is a simple fact that traditional
model of personal computing is proprietary to hell and back - you've
got proprietary hardware running proprietary code processing
proprietary data.
Proprietary solutions are designed to impose vendor lock-in on their
users and that's why we're completely screwed until we somehow manage
to establish free and open alternative as a widely enough recognised
standard or until we completely reverse engineer at least the most
commonly used ones in order to provide solid interoperability.
Notice that open standards tend to dominate areas which M$ and the
likes never touched or where they were too late to the party, because
they actually empower their users instead of tying them down in
exchange for questionable short-term "benefits"; it's just hard to
break the vicious circles that are already there and I have to insist
that being unable to get any machine you like without mandatory
initial dose of proprietary infection certainly doesn't help.