It's pretty silly that some people seem to think there should be only one thread for a topic ever. There are forums like that, and it just leads to 100+ page threads where it's impossible to sort out all the useful stuff without wading through each page. Not to mention that thread necromancy can be a PITA sometimes too. Obviously there's no fixed rule, but I don't think it's unreasonable to start a new thread when new info comes out and there's not a recently updated one.
/digression
Back on topic, Anandtech had a
good article on the newest Light Peak developments. To my mind the most interesting one is that it looks like there won't be a native Light Peak protocol, but it will just be a transport for other protocols (USB, HDMI, SATA, etc.). The second part we sort of new already, but it raises a few questions. Specifically what would be required to hook up an existing device over. If it's more than a tiny dongle (which seems likely), it's not going to be very effective in the short term to hook up existing devices.
It makes it more likely we'll see things like Light Peak docking stations, where one cable runs from a laptop to a converter box hooked up to a bunch of peripherals. That certainly has its advantages, but most of the gain is simply convenience. It also means that rather than replacing all the ports on a laptop, it's going to be in addition to them, if you want to use existing devices on the go - at least until (if?) we get devices with native Light Peak connectors.
It certainly opens up a lot of possibilities, and it's clear why Apple is interested. With a Light Peak connector, you could replace the USB and DisplayPort connectors on a MacBook Air, giving you display and much better I/O options while reducing the connector count. An Apple display could have a Light Peak and Magsafe connector only, but offer connectivity for the display, iSight, and as many USB, FW or eSATA devices as they offered ports for on the back (or even a second daisy chained display).
One drawback is that it could make connectivity a problem when you don't have a hub available (i.e. most mobile situations). Laptops are going to need standard I/O ports in addition to Light Peak to be useful, since I don't anticipate cheap dongle adapters to go from say LPK->USB, nor do I think we'll get a ton of native Light Peak devices overnight (particularly not cheaply). If you don't eliminate some/all of the ports, what's the point?
The other problem I see is that if there's no native protocol, you become dependant on the protocols it's carrying. I.e. there could be a nice shiny new USB4 Light Peak device that comes out, but your Light Peak compatible laptop can't make use of it because it wasn't made with a USB4 controller. Perhaps there's a solution to this (or in fact there is a native protocol) that hasn't been detailed yet, but from the sound of the Anandtech article, it seems like it could be an issue.
All in all, it opens up a lot of interesting possibilities, particularly when it comes to eliminating the I/O limitations of a lot of Macs. However there are a lot of implementation details that could make it a hit or miss.
