I'm talking at the packet level and you're talking about email gateways. Completely different things.
And about your TV analogy, this is a perfect example of a good interoperability, unlike IPv4 and IPv6. Color stations broadcast a luminance signal and a chrominance signal. B&W stations only broadcast the luminance signal.
B&W station to color TV: only the luminance signal is processed, TV shows a B&W picture
Color station to B&W TV: both the luminance and chrominance signals are received but only the luminance one is processed, TV shows a B&W picture
Stations could immediately switch to color broadcasting and consumers could buy a color TV when they wanted.
If TV was handled like IPv4 and IPv6 we would have the following instead:
B&W station to color TV: no picture
Color station to B&W TV: no picture
But there would be a lot of ideas about signal converter boxes and stations having to broadcast in both B&W and color using 2 separate channels.
And what happens when the destination host (which is IPv4) receives an IPv4 packet with IPv6 "routing info" in the message body? It cannot make anything useful out of the packet which will be corrupted by that "routing info".
What you're probably talking about with your vague definition of "bridge" is a tunnel or proxy. The former needs to be done between the source and destination host and the later needs to keep statefull information and is basically a NAT from an IPv6 network to an IPv4 network and needs its own IPv4 address as well.
Not really transparent is it? Not to mention the troubles trying to put a server using this.
The main problem with an IPv6 host is that it cannot talk to an IPv4 host directly. Why? Because IPv6 is not an extension of IPv4 but a separate address space.
Things would have been so much easier if IPv6 hosts were allowed to talk to IPv4 hosts, because most IPv4 hosts do have IPv6 compatible software already.
Example of what happens currently:
IPv6 host wants to reach an IPv4 address
failure to connect
Example of what should have happened (let's call that protocol IPv7 and consider it as an extension of IPv4)
IPv7 host wants to reach an IPv4 address
the packet is routed because the routers, even though some of them only have IPv4 addresses understand IPv7
the target host receives the packet, this host, although it has an IPv4 address, understands IPv7 and replies to the packet
For the first example to work, you need the target host to have IPv6 compatible software an an IPv6 address as well. Can you imagine pestering every network administrator to configure an additional IPv6 address for every host with no immediate benefit? Certainly not.
But for the second example to work, you just need the target host to have an IPv7 compatible software. Once all software is compatible to IPv7, the whole Internet will be reachable through IPv7 so you can then start ditching IPv4. This scenario is incompatible with how IPv6 currently works.
It already started with floppy disks.
1.44MB where in fact it was 1440KB.
And about your TV analogy, this is a perfect example of a good interoperability, unlike IPv4 and IPv6. Color stations broadcast a luminance signal and a chrominance signal. B&W stations only broadcast the luminance signal.
Stations could immediately switch to color broadcasting and consumers could buy a color TV when they wanted.
If TV was handled like IPv4 and IPv6 we would have the following instead:
But there would be a lot of ideas about signal converter boxes and stations having to broadcast in both B&W and color using 2 separate channels.
And what happens when the destination host (which is IPv4) receives an IPv4 packet with IPv6 "routing info" in the message body? It cannot make anything useful out of the packet which will be corrupted by that "routing info". What you're probably talking about with your vague definition of "bridge" is a tunnel or proxy. The former needs to be done between the source and destination host and the later needs to keep statefull information and is basically a NAT from an IPv6 network to an IPv4 network and needs its own IPv4 address as well. Not really transparent is it? Not to mention the troubles trying to put a server using this.
The main problem with an IPv6 host is that it cannot talk to an IPv4 host directly. Why? Because IPv6 is not an extension of IPv4 but a separate address space.
Things would have been so much easier if IPv6 hosts were allowed to talk to IPv4 hosts, because most IPv4 hosts do have IPv6 compatible software already.
Example of what happens currently:
Example of what should have happened (let's call that protocol IPv7 and consider it as an extension of IPv4)
For the first example to work, you need the target host to have IPv6 compatible software an an IPv6 address as well. Can you imagine pestering every network administrator to configure an additional IPv6 address for every host with no immediate benefit? Certainly not.
But for the second example to work, you just need the target host to have an IPv7 compatible software. Once all software is compatible to IPv7, the whole Internet will be reachable through IPv7 so you can then start ditching IPv4. This scenario is incompatible with how IPv6 currently works.
The whole Internet is reachable through IPv4, only a small portion of it is reachable through IPv6.
Dan J. Bernstein explains the problem perfectly well here as well as suggesting a solution.