Hi Lionel, many thanks for your feedback. There will be a follow up.

Many thanks Aubrey! W/r to anycast, my preferred analogy is calling the emergency number- you always call the same number but you would land in a different call center depending on the network you place the call from and the time of the day ;-)

He's right though, there's is no TCPv6 or UDPv6. What you're referencing is the field that occurs in the IPv4 and IPv6 header that indicates *what type* of layer 4 header *follows* after the IPv4/v6 header. TCP and UDP are the same for either one.

Absolutely! I don’t contradict ;-) The only reason I emphasize on udpv6 and tcpv6 is to build context for the importance of icmp for the ipv6 protocol ;-)

Personally I feel like the addressing part is not that complicated and not that different from IPv4. IP address of a certain number of bits with a subnet mask of a certain number of bits (cidr) is all the same just longer addresses. The explanation of scopes is good though and important but also this is not too dissimilar to rc1918 addresses. The part I think that a lot of people get confused about is routing, how it happens and why as well as different configuration setups besides SLAAC, like DHCPv6, RA/router advertisements, prefix delegation and interface tracking. A lot of people are used to thinking in terms of NAT which is why I don't disagree with this approach but even without comparison or even mentioning it might be important to stress especially for home users who are used to having to forward ports that this isn't a thing. No forwarding ports, instead you block ports with a firewall on the router or local machine. I'm not sure a lot of people fully understand with IPv6 and a typical home router that their pc is now directly accessible on the public internet instead of hiding behind a NAT (which isn't a firewall - it just doesn't know where to send incoming traffic unless you tell it).. While I understand this is an introduction, I had to understand all of that when setting up my own home network even. Though I still don't understand why I can't seem to track or delegate to my 2nd WAN interface, only get one /56 over dhcp and I can't ask for several smaller prefixes or seem to delegate a /64 or something to secondary wan..works fine for lan interfaces though.. not a huge deal but my ocd doesn't like seeing an adapter with unassigned address.

@@forbiddenera > with IPv6 and a typical home router that their pc is now directly accessible on the public internet that obviously depends on the router, I'm not sure what you mean by "a typical home router" but all the ones I've encountered do not allow incoming traffic by default, as should be true for any network device that isn't garbage.

Hi Manoj, the topics that you mention are definitely the right ones - I'd add to these: Unicast, Anycast, Multicast (Router solicitation, Router Advertisement, Duplicate Address detection), NDP ;-) Thanks for the feedback !!!

Many thanks Remy!

Hi Ben, I am honestly and fundamentally happy each time someone tells me that my videos could help them to achieve something. Thanks a lot for your friendly feedback.

Hi Ray, first off - thank you very much for being a patron!!! I am so glad that so many people seem to like the video (at least from what I read in the comments) - it looks like disconnecting from IPv4 is really the key for everyone here. Many thanks for the friendly feedback!

Hi, that's great feedback, many thanks. I experienced exactly the same. It was a great unknown until I took all my courage and stepped into that course. And after all now I realize that it is fun. I'll definitely do follow ups.

@@OneMarcFifty thanks

Hi Thomas - yes it is after all. Just people (including myself of course) tend to be "mentally stuck" in NAT. WE tend to forget that NAT is a workaround and that it is not Firewall ;-)

Hi David, thank you very much- yes, it does require preparation.

Thank you very much! I'll need to plan this episode - i've got some ideas but I'll probably set up a whole IPv6 test lab for that. I'll be back ;-)

Many thanks Volker!

Thank you very much ;-)

Hi Alex, thank you!

Hi Aaron - I went through exactly the same experience here. When I saw the many IPv6 addresses on my laptop the first time it totally freaked me out. And - as you say - the size of the address space is just overwhelming ;-)

"Of course we could always have multiple addresses in IPv4 as well, but it was typically the exception rather than the rule." Not entirely true. Your host always has 127.0.0.1 - though that's a bit cheeky. What's less cheeky is if a NIC is configured with address 192.0.2.1/24, it can receive/listen for messages destined to 192.0.2.255 (local broadcast). While not particularly obvious, your host has multiple IP addresses. It just doesn't really send with the broadcast address as source. Your IPv4 host also uses another address - 0.0.0.0. This one comes into play specifically for DHCP. If you look at the source address of a DHCP discover packet, you'll see this as the source address. In fact there's nothing stopping you from say, booting up a host on a network without DHCP (or disabling the DHCP client), grabbing an IPv4 link-local address (APIPA), configuring that statically on the interface, and then re-enabling DHCP for an additional address.

@@Jamesaepp "Your host always has 127.0.0.1" It's actually more than you think ^^. Even though IPv4 has such a limited address space, they dedicated a whole class A net to the local host, not just a single address. 127.0.0.1 is just the most commonly used and is usually mapped locally to "localhost". However 127.45.12.42 is also a local loopback address. The entire 127 net is local loopback. So 1/256th of the whole v4 address space is wasted. In IPv6 they decided that a single address is enough. If you need more, you can always assign more, but it's not implicit. Note that (0.0.0.0) is not really an address, it's merely a wildcard for the whole network to indicate no particular address. As you may know, when binding a socket, you can either bind it to a local address that the socket should listen to, or bind it to (0.0.0.0). That means the socket would be bound to all local addresses. In other words if a packet arrives at your NIC, it doesn't really care what the destination address is. Since it arrived at your NIC, it's meant for you. A lot servers by default bind to (0.0.0.0) so they accept any packets, no matter to which IP they were directed to. However explicit binding is needed if you want to run several servers on the same machine on the same port. You just need to bind each it to different local addresses. Locally you can bind servers to different local host addresses and it would work fine, locally. So you can run a server on 127.0.0.1:5555 and one on 127.0.0.2:5555. Of course you can only connect to those two servers locally, so this is just an example. Though If you bind a server to (0.0.0.0), you can't bind another socket to the same port as it 0.0.0.0 binds to all addresses. A single end point (IP + protocol + port) can only be used / bound once.

Hi Robert. Awesome - I am glad it could help.

Hey, thanks so much for your feedback. True, I should have visualized subnetting more. The follow up episode is available on my channel page btw ;-)

Hi Marco, many thanks. Glad you liked it.

Hi Tomás - thank you so much for your feedback. Yes - I am already planning follow ups - just need to make up my mind which one to do first ;-)

Hi Bruce, many thanks for the feedback. Glad to hear that I am not the only person making that experience ;-)

Many thanks Naveen. I really wanted to give viewers something to take away. Also I admit that I wanted to do a bit of publicity for my github repos ;-)

Many thanks.

Hi J B, yes - I totally agree - the fun is limited if you only get it LAN-side ;-(

Many thanks for the feedback, benediktus

Hi Mark, that’s great. There will be episodes on how to use it with routers. At least for OpenWrt and maybe for OPNSense/pfSense

Many thanks Stephan.

Many thanks Jörg - glad you liked it ;-)

Many thanks Roy.

Thank you - glad you liked it

Many thanks for the feedback :-)

Flying under a bridge is to IPv6 what apples are to oranges. Not even oranges. More like dust bin cover screws. Duuuude!

Hi Eric, interesting topic! I would need to run experiments myself for this, haven't tested it yet - this seems to be a good primer: stevessmarthomeguide.com/multicast-dns/

Many thanks Philip.

Hi Euxhenio - great to hear ;-) But don't forget to remember it again in the end ;-)

Hi Adrian, I was not aware of this. Is there any official documentation on this ?

@@OneMarcFifty Hi Marc, yes there is. Feel free to have a look at RFC 5952 section 4.2

Thank you - just checked it. It is explicit and states MUST. It even goes further and says that a..f MUST be written in lowercase!

Thank you - you’re welcome.

Many thanks Torin !

Hi Didi, glad you could use it! I can't remember where the course was listed - but it was long. Really long ;-)

Hi, thanks for the feed back ;-) Yes, that was big (yet simple) discovery for me as well ;-)

Hi luis, thank you ;-)

Hi Alex, i’ll give this a long thought but presumably this is not going to happen soon. There are a bunch of videos that I need to make before. Even though a captive portal is an interesting matter!

Hi Stephan, please do keep in mind that if you want to get a /56 prefix from Deutsche Telekom, then this will not work with their speedport routers (at least the one I had, don't know about the recent ones). I had to replace the Speedport with a DSL Modem ( I am using a VIGOR modem for this) in order to get full IPv6 functionality. Also lowered my ping time by 2 ms.

@@OneMarcFifty hey Marc, Thanks for the hint. No worries, i bought a zyxel Modem to avoid the speedports 😂

Thank you very much - I am glad you liked it.

Hi, many thanks for the feedback - yes, you may see a lot of strong opinions on IPv6 in other comments here - I do agree it is more a human subject than a technological one that prevents adoption ;-) But I don't want to sound smarty-pants here because I felt exactly the same before ;-)

Thank you.

Hi, that's very kind of you - thank you very much. It's only after that video and people leaving comments that I realized I had been saying it wrong for over 30 years ;-)

Hi, thank you very much

Thanks Pavel. Yes - as there is so much info floating around everywhere - that was the reason I decided to take that course. I never had to use IPv6 and could safely ignore it. But that left the "there is space to be discovered" voice in my head ;-)

Hi Stuart, many thanks.

Hi Sandra, many thanks for watching.

Hi Matthew - I think we are all the same ;-) We know we have to do this one day ;-)

We will have a deeper look into a couple of things in a follow up video. I just need to think over if I‘ll do that as a generic video or rather related to OpenWrt

Hi Damalia, I have learnt to love it as well ;-)

Hi Nick, many thanks for the friendly comment. Yes - when I launched my "IPv6 only" testlab I realized how bad the coverage is really. ISPs, content providers, CDNs - there's still a lot of room of improvement left ;-)

Many thanks !

Hi Igor - first and foremost - many thanks for subscribing. Glad you liked the video.

Thank you Arne, glad you could use it!

Awesome, many thanks Richard - glad it helped ;-)

Thank you very much

don't judge an address by its look would be the lesson to me then ;-)

Hi Robert, many thanks for the feedback. I never said it's not similar (or let's rather say comparable) to IPv4. I just said it helped me to forget everything about IPv4. The main reason for this was that I kept looking for a translation mechanism between private and public addresses (I was kind of locked into NAT thinking). IPv6 solves this with prefix delegation and subnetting the global address space which either does not exist in IPv4 or cant be done by end users because they would never get a subnet to play with (because of the smaller address space).

Hey, many thanks ;-)

Hi Jairu, great to hear from you. Looks like the video gets some traction so there will be a follow up. All the best.

Yeah - I felt the same honestly- but let’s not forget that NAT is just a workaround ;-) Firewall will do its job - with or without NAT. Privacy concerns could be addressed by changing the ipv6 frequently- as there are so many possible addresses it could in fact even be an improvement!

@@OneMarcFifty thanks for the awesome video, what I don't understand is if with ipv6 it will still be possible to hide your internal hosts from the outside world without them losing internet connectivity. Maybe i get it wrong but It seems this protocol implies no NAT so every device who needs internet connectivity must have a public ipv6 address. What if I don't want my crappy chinese IOT device to have a public ip address in first place ?

maybe adding the abbreviations like ULA, SLAAC to the table is helpful to ...

Hi Henning, maybe this has already been fixed on the Github - github.com/onemarcfifty/cheat-sheets/blob/main/networking/ipv6.md

Thank you ;-)

You're welcome Nikolai - thank you for watching!

Hi James, I believe that I have seen your name pop up in some of the IPv6 related forums / questions a couple of times. Great to see you here ! Many thanks for the feedback ;-) You are right with everything you say - there are so many IPv6 address spaces - I realize how difficult it is for a human being to grasp exponential models ;-) Also - if I think about people trying to use IPv6 with a single /128 address - that's so disappointing because it makes it impossible for them to discover the real beauty of it ;-(

@@OneMarcFifty Quite so. A lot of people have been misled by the IPv4 address shortage and NAT, to the point they don't understand how things are supposed to work. Incidentally, customers are often provided a /128 for their WAN address. This address cannot be used for routing, as there is only one device supported by it. Even a point to point link must be at least a /127, allowing for 2 devices. A /128 is only used to identify an interface, with routing typically done with a link local address. BTW, If you really want to get into IPv6, I can recommend "IPv6 Essentials", from O'Reilly.

Many thanks for your reply and explanation James. I was thinking about making a video for an IPv6 test lab - enabling viewers to test everything in a sandbox before they go live in the real world. Some are suggesting hurricane electric as 6in4 tunnel. Any opinion on this?

@@OneMarcFifty I started out with a 6in4 tunnel, in May 2010, and it worked well. I wasn't with Hurricane Electric, but I understand they're about the largest tunnel provider. My own ISP started providing native IPv6 almost 7 years ago, but provided 6to4 and 6rd tunnels for a couple of years before that. So, including a 6in4 tunnel would be a good idea, as many are still stuck on an ISP that doesn't provide IPv6.

Hi Michał, you’re welcome.

Thank you!