I think the point is that due to network effects the winner actively holds back the losers, so you can't extrapolate from what happened in a world where Skype won to another world where they didn't. It would have been better for everyone if the winner had been an open standard and a very small change could have accelerated that solution to ubiquity rather than Skype, in the same timeframe.
Out of curiosity, another industry where this (the small change accelerating an open standard such that there are more winners [a higher tide overall]) is what happened? I'm not saying there aren't any--honest question.
Can the protocol be reverse engineered? I guess regardless it is dependent on connecting to some central switching service; but maybe there could be an alternative, open central service someday.
Relatedly, if Skype changes the protocol, does it break an installed base of clients, or do those get updated automagically?
SIP, STUN, RTP, XMPP, Jingle, all these are recognizable open protocols. Skype, if they were truly interested in not being a walled-garden, would have either built or migrated to a network based on these.
Skype will be dead in 10 years if Apple makes good on their promise to keep Facetime open.
Funny you mention XMPP in light of how many times people call [it] "Google Talk" (this seems to draw the ire the folks that know about the real 'recognizable open protocol' underneath).
There was 'speakfreely', but I think John Walker gave up on supporting all the different ways to get around fire-walls.
The kazaa code did that pretty good and Friis and Zennstrom figured a way to monetize their bag of tricks they'd created while creating kazaa. Remember how they sold off Kazaa but kept the fire-wall penetrating technology by licensing it back to the buyer (Joltid), then later they did the same thing with ebay and skype.
Given that they'd already pulled that one once before ebay was really pretty stupid to fall for it.
