I'd say that it would be impossible for a game to release a dedicated Multiplayer mode like this without having rock-solid code, but there have been two major multiplayer releases (1.2 Wings and 1.4 CQC) that proved otherwise. They've been building on a foundation that was cracked from the start. Regardless if Multicrew is fun or useful, I'll be singing hallelujah if 2.3 finally fixes the underlying multiplayer mechanics and networking.
I hear ya...
I wonder what percentage of ED players actually play in a wing.
Not many, I'm afraid.
It's easy not to notice that you are seeing different ships than your wingmate in open space until you try to lock on to the same target and it's not there for one of you. The CMDRs trying to scan ruins together got a taste of this when they all saw different things on their screens, for example. But when pilots are in the same exact ship together and see different targets on their scanners, or different points of interest outside their windows, or their ship becomes suddenly uncrewed because their crewmate's client desyncs and drops them in a different location, there are going to be a lot more people screaming about it.
Consistency is one of the major problem areas for Elite. That, and the fact that P2P instances don't lend themselves well to getting all members of a wing into the same instance. My guess is that they're hoping that Multicrew limited to 2 players per ship will make problems less severe than they'd be with 4 or, god forbid, 8.
And I'd like to ask Brett once again since it got lost in the thread, is there any chance that there will be a hotfix/bugfix release before 2.3 goes live? Or are we going to be struggling with issues like instafail missions for another two months?
We had a fix for NAT traversal in the last patch. I think this did help a lot against the typical "can't see anyone in multiplayer"/"can't join this almost-empty instance" problem so many people have experieced. Heck even Frontier experienced this on their own livestream, when they were the only ones in their instance and were wondering why for the better part of 20 minutes nobody else was joining them. The fact was that a lot of us tried, but we never got into their instance because they were unable to establish connections to any outside peers. I admit it was good to see them suffer from this because for once they got a taste of this bug. I am pretty sure their recent fix was at least in part a consequence of their own experience back there.
There's no way to "fix" the wing cohesion issue when dropping out of SC for example. It might be possible if you and your wingman drop at roughly the same time, but if you have been in an instance for a few minutes and your wingmates join some time after other players came in too, there's no way to guarantee they can meet with you.
There is a way to fix the consistency issue but I'm afraid Frontier wouldn't like that solution. As it stands consistency in a P2P system is hard. Fast consistency (I don't want to call it realtime because that's... something else) can't be guaranteed. Not with peers joining and leaving at any time. It was fascinating to parse their netlog files, watching them try and move ownership of entities to different peers as people left and joined... It gave me new appreciation for Descent, which did something like this in the old DOS version when playing multiplayer via IPX. Ultimately though you're wading through a brown sea of sorrow.
What they could do is finally move to a client/server architecture, with players "owning" an instance and all entities in it. Benefit: You could guarantee wing and crewmate cohesion. Drawback: If you join someone who's network connection is limited, you'll have a bad time.
But you know why I'd still take this over P2P? Because with P2P, if one client with a bad connection joins an instance, everyone in that instance will have a bad time, not just the player with the slow connection.
