Open margau opened 6 years ago
I guess this implementation step was currently disabled by missing support for IPv6 on dependency "rs/node-netmask", right?
I didn't currently find a netmask library supporting IPv6, but the only missing thing would be finding the broadcast ip, right? If that's the case, why don't simply combine it from given IP address of the controller and highest-possible value for the last ip segment?
Pls give highly appreciated feedback :-).
Hi, the problem is deeper: ArtNet has - as far as I know - no support/spec for IPv6 yet, so we would be kind of "faster than spec". As IPv6 has only multicast, and doesn't support broadcast any longer, that's also something which has to be implemented and considered.
The problem is not with the implementation, but with the strategy behind.
Best regards margau
Hi, checking back the ArtNet 4-specification I come to the same conclusion - IPv6 is not part of the specification at all.
But in fact - after thinking about it: Assuming that ArtNet really wants the first IP address's block to be "2", then there are still 256 x 256 x 256 addresses remaining for random ArtNet-nodes / ArtNet-devices. Even divided by 512 devices per universe, there are still 32,768 universes available, and that's one again what ArtNet spec prompts. Even industry leading manufacturer MA supports with its new console MA3 only 250,000 parameters, so about 500 universes.
So why do you plan to integrate IPv6 at all?
Best regards ulrichrobin
Hi, I'm not wanting IPv6 for numeric reasons, but for protocol reasons. As a network guy I'm trying to shift away from Legacy-IP (also called IPv4) to as much IPv6 as possible.
And: v6 has many interesting features - like autoconfig/SLAAC and larger Multicast-Groups - which could be very interesting for ArtNet too.
Best regards margau
Implement IPv6 (v4 only, v6 only or dualstack)