A hypothesis concerning the RF mount

Messing up 3rd party lens compatibility would be a bummer, but will guarantee more Canon lens sales.

I don't understand the point of mounting an EF lens directly onto the mirrorless camera body. It will be much too close to the sensor. How could it possibly work?

Brilliant. Locks out 3rd party lens makers for 17 years for the RF mount. However, the remaining EF mount 3rd party lense should be accessible for the mean time. I certainly hope there is a EF-M full frame option as well. Auto crops to APS-C with EF-M, and with EF adaptor takes EF-S and EF lenses. I want compact options.

if they could pull this off, it would be quite a brilliant solution

How's that supposed to work, EF-M without an adaptor? Any FF mirrorless mount would be as wide as the EF mount or even wider.

the source suggested that they have seen EF lenses mount directly onto the camera, if that's the case EF-M lenses would not be able to be used. Simply EF/RF lenses on the mount. It may not lock out lens makers but it MAY lock out Sony FE adapters to support RF lenses

Maybe I'm not understanding the hypothesis (I'm slow, give me a break)... How would the flange difference be rectified? If there is no flange difference, than what benefit could there be in switching mounts? Or is the hypothesis that the mount would be switched and the camera can support two protocols?

I would consider it a mistake if Canon doesn't shorten the flange distance in its new mount.
As a sexy fix for EF lenses, someone wrote about a moveable sensor which would be really interesting but I don't understand why an adapter wouldn't work.
Why wouldn't it work? A Canon adapter, made by Canon for Canon, would work flawlessly and be extremely fast.
Canon would then have a new future-proof mount and the adapter would make all or their EF lenses work.

Is there really a reason to have a native EF mount (or an EF-R with an EF flange distance) on a Canon mirrorless? If so, what is it?

I'd been wondering a while back if they'd use the same mount but create mirrorless lenses which can extend further back internally, behind the mount. With no mirror in the way perhaps they can support EF lenses and potentially more compact overall mirrorless oriented lenses that way.

You'd need some really bizarre rear lens caps for RF lenses if that's how they're designed, but so be it.

Sensor movement, mount extension, lens design (so technically no change in flange distance but a lens that extends into the body). There's tons of possibilities so I guess we'll just have to wait and see.

it sounds like they are using the 44mm registration distance if they saw EF lenses mount natively.

Let's face it. Canon has over 100 million reasons to use the EF mount and continue using it. Shortening the registration distance only improves a couple of handful of lenses. Canon RF lenses sound like they have a different (faster) lens protocol and ALSO could sit back into the mirrorbox similar to EF-S lenses.

Again, this was a hypothesis based upon the limited information we have and patent applications that were released.

For the Sensor to more back and forth waht 12mm? would there be an extra 12mm of space lost to accommodate old lenses?
How could you get the mechanical sensor moving reliably for say 10,000 movements?

Someone came up with an idea, and someone else a graphic to illustrate it, of EF fitting normally and RF lenses having the rear element/s protruding into the camera body.

This way you retain 100% adaptor-less compatibility with EF lenses and RF lenses can still benefit from getting closer to the sensor & have a small(er) camera+lens size.

I don't know enough to have any idea whether this will restrict potential designs for RF lenses.

For this reason (reliability), I believe the sensor will be fixed at the standard EF flange distance and new EF-R lenses will simply extend into the body and sit much closer to the sensor. Very simple, reliable and the best of both worlds in terms of short flange distance and full native EF support.

The only drawback I can see is the camera body won't be able to be as slim as sony/nikon. But that's only a drawback for those who care about a thin camera body.

IBIS systems imply that it is already possible, it's more a question of logistics. Realistically, the same mechanisms that drive the focus group in the lens could drive the sensor plane as well, as long as there are flex cables attached to everything. And those systems actuate far more frequently. Part of the new lens protocol could be a key value to position the sensor. That would open up the lens design for any flange distance that is optimal for the lens, not forcing them to use one constant one for everything.

Guess 1: Body approach -- An EF lens mounts directly. When an RF lens mounts on the camera, a sliding shaft is engaged which pulls the mounting assembly with the lens further in towards the camera. A mount within a mount, in effect. Hopefully they found a way to do this that does not impact unit cost too much.

Guess 2: Lens Approach -- All RF lenses will have a rear element that extends back towards the sensor once mounted. The basic EF mount stays the same. If a future new lens doesn't need the inward-extending element it will simply be designed as an EF lens. (I don't think telephotos, long zooms, or Big Whites will need the inward extending element, so they will be EF lenses now and in the future). (If an RF lens is mountable at all on a regular EF mount camera it wouldn't hurt anything; the inward-extending element would not deploy, and would not be harmed.)

Guess 1 adds the cost of handling RF lenses to the mount on every camera sold, and I'm thinking the unit cost might be a problematic, but may be workable. Guess 2 shifts the cost of accommodating close approach to the sensor to each lens that needs it, but the cost may be even higher, and flexibility of lens design may be hindered somewhat.

My guess now is Guess 2, the in-lens approach.

Final note: Forget mounting APS-c EF-M lenses on the new full frame RF camera. Not workable. Not gonna happen (without an adapter, which won't be coming from Canon).

I guess we'll know next Wednesday!

Oooooooooh - I am excited again!!!!!!

Exactly. However, I feel like both Sony and now Nikon are still living off the fascination about the size of the first generation A7 cameras. The third generation A7 cameras as well as the Z6/7 are significantly beefier than the first gen A7 cameras (which had abysmal battery life and - at best - mediocre ergonomics). So while a thin camera body is nice in theory, recent camera models show that too thin isn't ideal either. Especially if you want to have a nicely sized grip, you're not gonna get a tiny camera.

So long as the camera body itself is lighter I think that's enough. It won't be as convenient as an EOS M but even 1lb off the weight would be great for carrying it around on a trip. Presumably there'd then be a slow migration to the (presumably) smaller/lighter RF lenses.