E1.31 sACN Public Review.

[multicast]

The E1.31 sACN Standard has gone to its 2nd Public review. This revision of the standard is primarily about including ipv6 into the standard, and fixing a few minor errors that had arisen.

You, as users of the standard are welcome to submit your comments on the standard. It closes 2 October 2017.
http://tsp.esta.org/tsp/documents/public_review_docs.php

Additionally the TSP ( technical Standards program ) would welcome additional membership from the Christmas Blinky community who are not represented, and i believe would be one of the biggest users of the sACN standard. Please talk to me about this, if you are interested.

 

[multicast]

I have a serious concern about the mandate in the standard that says all devices SHALL ( ie MUST ) support both ipv4 and ipv6.

Taking the approach of mandating ipv4 in the E1.31 protocol, is effectively making E1.31 a grand-father protocol, which will be unusable at some point in the future, or a change will need to be made in the future to the standard.

The standard precludes the use of systems/hardware/software that are Ipv6 only, which are becoming more common place. This mandate will stop development of compliant new technology. There are already ipv6 only technologies such s 6LowPAN. We also The Mobile device market is rapdilly moving to Ipv6 Only, and that phones/tablets will be delivered in ipv6 only.. This means that an application that was intended to run on a mobile/tablet will not be able to be compliant ). As IOT devices and technology advances it will be adopted by the Lighting industry.. ( because it will be cheap and pervasive ), this problem will only become amplified.

The impact of this mandate, is that there likely immediately will be devices developed by manufacturers ( including myself ) who will immediately be none compliant.
I have made exhaustive searches, and have not found a single other protocol across any industry that has made such a mandate, when adopting ipv6. I do not believe that the lighting industry is so special that it needs to make this assumption either.

I do hear the argument about compatibility but I strongly disagree that it is going to be any more than a rare occurrence, and one that any competent person would avoid in the normal course of design/ installing / using a lighting system.

 

[David_AVD]

I think I'd like to see IPv4 left as an option. There's going to be a lot of IPv4 E1.31 hardware around for a long time.

 

[multicast]

I think I'd like to see IPv4 left as an option. There's going to be a lot of IPv4 E1.31 hardware around for a long time.

Its not about removing IPv4 at all. that woudl not be good.. Its about providing the option for an ipv6 ONLY device. The way the standard as propsed written means that you can either have a IPv4 Only device or an IPv4/IPv6 device. The option to have an Ipv6 Only device is not permitted under the standard.

 

[damona]

You should be able to make a IPv4/IPv6 and have an option to be able to disable the IPv4 or IPv6 (not at the same time), if the customer wishes too.

 

[multicast]

You should be able to make a IPv4/IPv6 and have an option to be able to disable the IPv4 or IPv6 (not at the same time), if the customer wishes too.

Some new modern devices and stacks do not have ipv4 capability. These particually are devices that are intended for IoT type applications.. If the standard progress's as written, these devices will never be able to be compliant with the sACN standard..

At some point ipv4 will not exisit, and thats becoming more of a reality faster and faster and faster.

 

[dkulp]

I don't know... there are professional/commercial security gateways that are being sold right now that don't support IPV6 (Ubiquiti/UniFI). Not having IPv4 on a product would be a major fail. Plus, for *THIS* community, it's even more important as the drivers we use (FPP, xLights, etc...) don't currently support output of e1.31 on IPv6. The current "release" versions of FPP explicitly disable IPv6 at the OS level. My latest BBB images leave it alone so it will grab an IPv6 address, but I'm pretty sure the e1.31 input or output won't work with it.

Anyway, I think they want to make sure the e1.31 products that people may buy will work on their networks. I'm willing to be a large percentage of the folks in our community would have no idea if an IPv6 only device would even work on their network. The only way to make sure of that would be to mandate the IPv4, at least for the next several years.

 

[multicast]

Plus, for *THIS* community, it's even more important as the drivers we use (FPP, xLights, etc...) don't currently support output of e1.31 on IPv6.

e131 does not support ipv6 so thats probably a good reason why.

The only way to make sure of that would be to mandate the IPv4, at least for the next several years.

What this does is grandfathers e1.31 meaning that it stays where it is. Another protocol will get developed along with new devices that are only ipv6. ANd then we have yet another standard.

 

[damona]

At the end of the day if devices are made with IPv6, if the company buying them only has an IPv4 network they will not purchase them. Market will decide. Companies are still disabling IPv6 on internal networks.

In fact its not that much different to my suggest that 512 limit should be dropped for E.131 on a per universe. i.e. if all the devices on a Universe know how to handle channels > 512 then you should be able to send 800+ channels in a single packet.

If all the devices on a Universe are IPv6 only it should not be an issue.
i.e. maybe the standard should read all devices on the same Multicast Group should support the same IP standard. i.e. IPv4 or IPv6 Then the broadcaster of E1.31 only needs to send out one packet per universe and its an IPv4 packet or an IPv6 packet.

https://www.google.com/intl/en/ipv6/statistics.html

 

[damona]

Maybe devices to be compliant with the standard
From 1-Jan-2025 all new devices sold/manufactured must support IPv6 and optional support IPv4
i.e. more of a moving forward than a moving backwards.

The market will determine how many still support IPv4.

 

[multicast]

At the end of the day if devices are made with IPv6, if the company buying them only has an IPv4 network they will not purchase them. Market will decide. Companies are still disabling IPv6 on internal networks.

And many now ( google / apple / large universitys ) are now disabling IPV4. We are in that messy in between time.

In fact its not that much different to my suggest that 512 limit should be dropped for E.131 on a per universe. i.e. if all the devices on a Universe know how to handle channels > 512 then you should be able to send 800+ channels in a single packet.

Yes. You shoudl come and join us at the TSP!

 

[Kitman]

IPv4 and IPv6 discussion has been going on for as long as I can remember, we were running out of IPv4 ranges about 8 years ago and yet we continue to use IPv4 for lots of stuff. My ISP allows me to disable IPv6 and only use IPv4 and when I do Netflix traffic is free (Go figure) however a lot of small internal networks will continue to use IPv4 just because it works and there is no benefit to them switching over to IPv6.

Really IPv6 needs to be used for all internet / WAN traffic and IPv4 can be reserved for local networks, yes there are lots of backbone infrastructure that is using solely IPv6 now however converting everyone over seems to not be at the top of companies priorities anymore and they just keep using what works.

 

[dkulp]

At the end of the day if devices are made with IPv6, if the company buying them only has an IPv4 network they will not purchase them. Market will decide.

The assumes the market knows (or cares) enough about it to know the difference and make a choice. Maybe in the professional lighting world, sure. But in our community, I'm willing to be a large percent of the Christmas light folks would have no idea if IPv6 even will work on their home network at all. They just want to buy a "e1.31 controller", plug it in, and it works. Extra specs and stuff thrown around are just extra noise that cause confusion.

 

[multicast]

The 'market' for the pro lighting world and the hobby market is not that different. In fact, i'd say in both a majority dont' know much about networks in general. And they just want to plug it in, an make it go. I do get this. This is essentially the reason the mandate for IPv4 is there.

It is the unintended consequence that i'm concerned about. - E1.31 will essentially be grandfathered.