I like the idea of being able to use Wayland with my current Nvidia binary drivers. That way more people would be able to move to Wayland before Nvidia optimizes their driver for it (giving them a reason to do so in the first place).
Asserting that no effort should be made to accomodate blobs is a mistake imho. After all, nvidia has a driver that works on both Windows' protocol and X (on Linux, BSD and Solaris, no less) with minimum adaptations already. If your shiny new protocol can't accomodate that, you can hardly fault nvidia for not being flexible enough. Remember, they will have to dedicate resources to make adjustments anyway. And to test. So, taking a "my way or the highway" approach is not exactly productive.
Of course, it's not unthinkable the Wayland releases with OSS drivers support only and works out blob issues a little further down the road. After all, they don't (he doesn't?) have unlimited resources either.
Bull. Pure and utter crap. I've got a 3-core Llano running the r600g driver on my MythTV machine in the living room at home. It records and plays HDTV at a 1920x1080 output resolution without any issues. It plays Bluray quality movies without issue. HDMI audio out works perfectly. Flash videos play fine.
I realize that you probably think that VDPAU/VAAPI is absolutely necessary for an HTPC, but you're wrong. If the CPU is up to the task, there's no need for hardware decode acceleration. Those CPUs are up to the task and then some.
Haha, yeah great idea - let the CPU do all the work and let the GFX chip burn even more power because your precious OSS driver does not even support proper power management.
If you have a 3 core Llano, you probably have a A6-3500. That CPU alone has a TDP of 65 W! Yes, I really can see how this is an alternative for a low power, low noise, HTPC. My Intel Atom system with Nvidia Ion does not even have a fan and it's power supply has a maximum output of 25W.
Have fun with your OSS driver and your crappy HTPC.
I propose we not slough off something mature like X. It doesn't make sense to foist an entire upheaval of the graphics stack for very little (if any) added value, all because of one guy's pet project.
This has the potential to be more disruptive than the transition to PulseAudio, except without the nominal features that PA brought... and at a time when such a disruption would be very unfortunate. (I.e., while trying to launch a serious gaming effort on Linux.)