Why can't he care about this?
Any user that concerned with micromanaging TCP/IP traffic should at least have somewhat better knowledge about diagnosing issues of this sort. Managing bodies and services like ICANN, IANA, DNS, and WHOIS exist for getting to the bottom of this. Doing a WHOIS lookup to confirm that it's a Valve address was the first thing I did.
It wasn't showing up before one of the patches,
Unqualified hearsay, actually. And was it a Terraria patch, or a Steam client patch? The last Terraria update was 1.3.1.1 on 31 May, nearly a whole month ago. So how recent is "recent"? Are we talking days, weeks, or months? In comparison, the most recent Steam client update was 15 June. It seems far more likely that he only just noticed the traffic for whatever reason and assigned causation where none exists. Based on the initial inquiry we know nothing about the discovery method, the protocol (TCP or UDP -- the difference is important), the analytic process, the firewall used to control traffic, or the method of remediation used for this probable non-problem. And based on the follow up replies, we still know nothing about any of this.
Also unqualified. I found analogous behavior with my Steam client intermittently talking to the Steam network, and I don't see anything about it that makes it particularly "strange". Likewise, I found absolutely no indication that Terraria.exe makes any remote connections on startup (using
TCPView shows the process opening a grand total of zero sockets, inbound or outbound). I was even able to confirm that cloud synchronization is handled by Steam.exe (Terraria reads/writes a cache folder while running). I'm also fairly familiar with the game code and haven't found anything anywhere that substantiates the claim that Terraria is making unsolicited outgoing connections. The only part that appears to be even vaguely capable of doing something like that is the embedded Steamworks NET.dll, which has been the same version included since 1.3.0.2 (released 30 June 2015) -- which sort of goes against the whole "recent" claim.
and goes unexplained unless you know specifically what the IP belongs to,
Already addressed above and in previous posts.
so why can't he ask a question?
He's well within rights to ask a question. It would be an interesting question
if it could be demonstrated that it's actually happening. If it's happening, it should be reproducible, and it's not -- or at least not in the way the question is stated. But the question doesn't get to exist in a vacuum and doesn't get immunity from inspection, criticism, or refutation. Now, if substantial proof of the claim arises, I'd be completely willing to follow that line of evidence to try to get to the bottom of it. But right now, "Yuh huh, because I said so!" isn't exactly actionable information.