probably outdated, but probably still valid: http://www.kloepfer.org/ipv6-homenet.html

After reading through this article I have a question, which may be an issue (at least for me). Should I not use an fe80::/64 IPv6 address because it is a link local address, which means that I have to specify the interface each time I try to connect? Since my ultimate objective is to use my C code and LabVIEW to use IPv6 to communicate between my Linux machine and another computer, I have no way of specifying the interface currently when calling C code commands like "bind()" and "listen()" and "connect()" or in my LabVIEW code either.

Is there some other IPv6 address that I can use that would work for my application where I can just specify the IPv6 address only (not the scope ID or the subnet mask prefix)?

FYI I contacted my IT and they are looking into the Firewall settings.

Thanks!

In order to work with computers you need to understand all of the technologies you are working with. If you plan to make a functional program that communicates over the network, it is important that you understand all the implications. In this case you have very little grasp over TCP/IP networking, including ipv4 addressing, ipv6 addressing, the DHCP protocol, and routing. I suggest you do a bit of research. For the entirety of this thread information has been spoon fed to you. You could easily use this information to do a Google search and find your answers. This forum is not here to spoon feed you, please do some research. Any questions should be directed here once you have a better understanding of how TCP/IP networking works. Without this understanding your project is pointless.

Your questions were already answered by mralk3, so I can only repeat myself: it is not a coding problem, but a configuration issue. Your network is still misconfigured and you need to understand it first and fix it to be able to connect.

great information. Thanks

Okay. So, I've read up a lot on IPv6 addressing on the web, and I have a couple questions. Sorry, if they seem trivial. I'm just trying to learn, so that I can program this software for my job. I'll just specify the best source I found for each of my questions to make it easier, instead of linking all the pages I read.


After reading this: https://www.iplocation.net/subnet-mask I came to the following conclusion, which is hopefully correct...

So if a subnet mask performs a bitwise AND operation on the IPv4/IPv6 address and I only need one IPv4/IPv6 address for each of my computers, can I just specify a subnet prefix length of 255.255.255.255 for IPv4 and /64 for IPv6, which should give me all bit 1 for each number, which means the bitwise AND operation would give me just one address i.e. the address that I specify when configuring my network? Then, I could just give one computer an address of fe80::000a/64 and the second one an address of fe80::000b/64



After reading this: https://www.ipv6.com/general/ipv6-addressing/ I have a question.

For the link local addresses (specified in the above link by an IPv6 address of fe80::/10) shouldn't I be able to use any subnet prefix mask I want as long as my address starts with fe80::? Like, I should be able to use fe80::/64 not just fe80::/10 correct? A second question that I have is: Is it necessary to use these special IPv6 addresses or is it just good practice and I can technically use whatever I want (other than these special addresses) and the IPv6 protocal should just work normally?



After reading this: https://linux.die.net/man/8/ping6 I have a final question related to my C code.

The above link says that when using ping6 and a link-local address, I have to specify the interface with the -I flag. Does this mean that when writing C code for IPv6, I have to also specify the interface in the C code, like Eth0 and Eth1, especially if my Linux computer has two ports?