Implement IPv6 support for server and direct connect

[jrb0001]

No description provided.

[pchote]

A quick initial review as requested. It will be a while before I can put time into thinking about the potential implications of the changes here.

[teinarss]

Why not use dual mode? http://www.lybecker.com/blog/2018/08/23/supporting-ipv4-and-ipv6-dual-mode-network-with-one-socket/

[jrb0001]

Not all OSes support it. Also multiple sockets are much more convenient if you want explicit bind addresses. I decided against adding an option for overriding the bind addresses in this PR because that would need fundamental announcement system changes which are out of scope.

[pchote]

I spent some time poking at this tonight, sorry for the long delay.

A few crashes and minor style nits:

[pchote]

The connection failed dialog now repeats itself:

[pchote]

Everything else seems to work as expected, so I expect 👍 once the issues above are resolved.

[jrb0001]

Fixed everything except the error dialog:

• I don't see a way to remove the redundancy without removing a line from the dialog. @pchote doesn't want that (discussion on IRC).
• Reverting back to the old error handling isn't an option for me. The lines on the dialog are still redundant and the generic "Server is not responding" is just wrong in many different cases.

[pchote]

As I said on IRC, my issue with changing the error dialog in this PR is:

• The error messages is that the screenshots above, which cover the two most common cases that players see, is a step backwards in terms of polish.
• The connection dialog implementation is old and fragile, and I am not prepared to risk regressions given the short timescale we have to branch and ship the playtest.

It seems silly to delay IPv6 support over an unrelated scope creep, but if you're not willing for us to change this I don't see much other choice. Merging such a big feature without time to bake on bleed would probably not have been wise anyway.

Moving to next + 1 gives this more time to bake and resolves my second point so please go ahead.

[pchote]

ping @jrb0001.

[pchote]

@jrb0001 do you mind if I take over this and push to your branch to update this PR?

I would like to get this merged before it becomes too late in the Next + 1 cycle.

[jrb0001]

@pchote I don't think I will find time and motivation to work on it in the near future so go ahead.

[pchote]

@jrb0001 can you please check the "Allow edits by maintainers" checkbox at the bottom of the right-hand sidebar? Your repo is currently rejecting my ability to update this PR.

[jrb0001]

@pchote That checkbox is enabled, no idea why it would reject your push.

[pchote]

paul@Ragnarok:OpenRA+ipv6-server$ git push jrb0001 +ipv6-server 
ERROR: Permission to jrb0001/OpenRA.git denied to pchote.
fatal: Could not read from remote repository.

Probably easiest at this point to file a new pr, so i'll close this one.