May not be minimal to maintain and patch legacy C++ code against various security vulnerabilities. The risk/benefit of doing imperfect patching might obviously indicate shutting it down.

Meanwhile, open-sourcing is usually inhibited by 1) licenses of middleware in the tech stack and 2) sketchy code that's not ready for public sharing. It truly is a ton of work to open source commercial code!

If they opensource the older server code, someone else can build a new compatible version with say, golang or rust, that can end up leaner & faster but also a bit more secure as it can be updated.

I’m sure that’s a big consideration, but they really could just dump whatever code they own the rights to, in whatever state it’s in, and people would be thrilled.

Although, now that I write that out, I bet determining the code they have rights to is, sadly, more work than they’re willing to put in.

Edit: also, if some of the code is used elsewhere in current games, this could inadvertently reveal security bugs in them.

The community will figure it out. If it still exists, that is. If not, perhaps it deserves to die. Would be a pity though, I really liked UT2004, better than UT3.