People on MR were only noticing now that the MacBook Pro Retina models were non-upgradeable and had dispensed with Firewire? The MacFixIt article on which that piece was based is about six months old, and the issues don't end there.
Apple offers a thunderbolt-to-firewire converter and pretends that's enough. Unfortunately, the conversion only goes from thunderbolt to firewire, not the other way around. Which means that, if you're a small studio with bags of firewire I/O hardware, you'll need a
Sonnet Echo Express Card to use your own costly and hard-won kit.
Additionally, if you're a former Mac user who wants to migrate data, you can't use Apple's Migration Assistant with the Firewire adapter or even the Echo Express. This is because MA only supports firewire-to-firewire and thunderbolt-to-thunderbolt (USB isn't supported at all). The whole point of MA was to make moving from one machine to another transparent, so that all of your software and passwords are still recognized and usable. Effectively, Apple has sabotaged their own claim of seamlessness, meaning your choice is to create a disc image and transfer it from one machine to the other via USB 3.0.
"It just works," indeed.
For people whose
industries are Mac-based, the obvious solution is to forgo the Retina MBP and buy the MBP with conventional construction, which is still being sold by Apple and still offers user-upgradeable RAM, batteries and cards. If you must have the best laptop screen possible and worship the Retina display, then wait for a comparable external 233-ppi-or-better display to appear in 2013-4. Meanwhile, you'll have a laptop that can be upgraded and repaired by you, and will doubtless age more usefully.
If I weren't bound to Apple by the music industry, I'd have built my own tower and had done with it ages ago. As things stand, the iMac is as powerful as the old Apple quad-cores but heats up and isn't reliable as a replacement for people running loads of, uh, loads.
If Apple doesn't release a tower eventually, we'll all build hackintosh towers and use those until we've migrated our systems to whichever platform feeds the music production niche more reliably. Linux appears to be the obvious endstop, but it isn't because music production encompasses so many variables and levels of user that it has to be idiot-proof. I won't even tell you what clients have done in the middle of a mix when I was elsewhere adjusting physical settings or downstairs accepting a food delivery. The idea of what they'd do in a Linux environment is terrifying.