Anticept

BI, if I may give you a suggestion :) The TAGES DRM that you use is a bit annoying in the regard that you can run out of activations. Can we get an interface through your website to deactivate old computers to activate new ones? And no, I have not run out of activations. Just a thought for the future, I really don't want to have to go through support tickets if I ever need to.

I'll march to the beat of this thread's drum, and ask for focus to be -entirely- on path-finding until it is fixed.

On this note, the barrier panels have bugged on my game. They are both active but the gate is not down. I've resorted to trying to fling walruses over the damn wall now.... sigh. BI, the save states aren't doing too well...

This is outdated thinking. Unless some huge gamebreaking problem comes up, you only need a little bit of testing, then release it. Any problems that have cropped up since the patch, or those which were accidentally created during patching, can also rapidly be fixed by the next short release cycle. This way no bugs are left un-squashed for any length of time, and less time can be spent on QA and more time on fixing more bugs. It is FEATURES that must be thoroughly tested.

I have suggestions for the future. Every week, have a "Meet the Devs" session in IRC, where people can directly ask you guys about issues and plans. I strongly, STRONGLY believe you would see an -overwhelming- support of getting the path finding fixed, which is the most game breaking issue above all else right now in pretty much all of your games. Next, being an alpha, REGULARLY build and release. Even if it's small fixes. People love to see progress, even if it's slow. Hell, when there isn't much contact with the Dev team, people *love* patch notes and lots of patches, it's that feeling that SOMETHING is being done. This alone would have made the P&C go much smoother for us. On that note, your Arma release cycles are too slow; we would have regular beta releases every couple of weeks, but why wait so long for a new official version after 6 months? Many of us use mods and the game built on day one of beta releases anyways, so what's the harm in more rapid release cycles? I also felt like the bug tracker was not being updated enough. There are those of us out here that really want to help, you could hand out some direct lines to the dev team, be it through a skype conversation, etc, so that those moderators can get feedback and help keep the tracker clean. Let the devs tell the mods what they are thinking of doing, and the moderators can put it in writing, acting as their liaison. tl;dr More contact with the community, even if it's brief. You need MUCH faster release cycles, even with tiny bug fixes, and let some of the community help moderate the bug trackers and forums so you guys can focus more on your work.

What they've done already is well ahead of most games. However, I'd like to point you to Battlezone II. It too uses open terrain and obstacles like BI's games, but even with tracked vehicles, they can navigate strange terrain and places with heavy obstacles. Still gets hung up from time to time, but as it stands, for a game from 1999 and can navigate better than BI's games, perhaps pioneering isn't entirely the term that should be used. Definitely ahead of the pack, but it's already been done better than what we are seeing. They chose to make carrier command, and they also chose the terrain layout. I applaud them for their ambition, but bit of a reality check, that AI pathfinding needs to be fixed. EDIT: Just make walruses hovertanks. I'll miss the wheels (feels cool) but at least navigation won't suck.

There's nothing wrong with the islands, or the carrier layout. BI's AI has always had issues with navigation. Always. From OFP to Arma II, it never could navigate once the land started getting a little tight, even with infantry it has difficulties. I strongly believe the problem with their AI is it doesn't think ahead, isn't good at all with understanding vehicle limitations, and doesn't consider what other AI threads are doing with other vehicles. This leads to avoidance routines which are crude (put distance between self and another vehicle, even if that means putting itself in a difficult to maneuver position), and tries to navigate in strange routes. A good example is a forest. A person who has no problem getting through a cluster of trees will get the AI hung up. If you watch how it navigates, it will drive between two trees, and it's as if that's all it was concerned about at that exact moment. It doesn't seem to consider the path ahead when it's doing avoidance with those two trees, causing it to run too close to a tree in front that is easily avoided if it had planned ahead a little bit. Then it kicks in those shitty avoidance routines again, where it will try to back up to get away from that tree it just about ran into... then it seems like it just drops what it's doing to get to the next waypoint as though it gave up, run it's ass into another tree, causing this whole cycle to spin out of control. There's no grace in it's navigation. No planning. It's like the waypoint system just drops a few nodes for them to get to and the movement routines just sit there dumb-faced going "OKAY :3" with no idea what it's really doing. Another symptom showing is that when you watch the AI do turns, you can hear the engine RPM drop at each node. Again, seems to me it only evaluates what it's doing after it achieves it's one-tracked goal of GET THERE NOW, SCREW THE OBSTACLES. Once it reaches that point, then it recalculates it's route to the next waypoint, and there's that split second it takes to do so. Finally, if you have a vehicle in front which is turning onto the road ahead just as another vehicle is pulling up to it, it will do a silly avoidance steering off the road or something like that trying to avoid hitting it. It doesn't seem to consider that the other vehicle that is turning onto the road is going to be going in the same direction; it act's like it just sees an obstacle and it needs to veer to avoid. The AI seriously needs to be planning a few steps ahead as it moves, and figure out ways to smoothly connect each path so that it doesn't get hung up in silly ways.

Well you have 25 years to make up for then, get to it!

I think the demo we are seeing is a fraction of what BI has planned and already implemented. A lot of things seem very incomplete, or feel like there are gaps that were filled at one point. I get the feeling this P&C is to see how the game performs at large on a lot of people's systems, and to see what bugs us the most. Right now, seriously, the AI is annoying the hell out of me. Always been a real problem in ARMA and the cause of phenomenal amounts of frustration... it's basically functional, but JUST that. For CC:GM to really hit home with me, I'd like to see the AI take top priority for fixing, even if it takes all of the beta time to do it.

http://forums.bistudio.com/showthread.php?133472-The-problem-with-Multi-Player&p=2135309&viewfull=1#post2135309

I'd be happy with co-op only (at least on release) and have it so that each person controls from the same carrier. They can take control of any unit or gun that they wish (hell, imagine a driver and a gunner on a tank). If their unit is destroyed, their view goes back to the carrier and they can control the next one (if available). Basically, give them the same abilities that single player has. If a unit is already being controlled and someone wants to take over, then it will flash on their screen that someone wants control, and they can switch back to carrier or another unit (think Trine 2 MP). This game probably should not have a multiplayer lobby. AI War: Fleet Command did not use a lobby because they felt that games with randoms just won't work. They did have an IRC channel for setting up games though, if people wanted to put in a little effort. It worked out really well in the end, because when it takes effort to set up a game, you filter out the people who don't want to put an effort into teamwork. One person, not necessarily the host, should be voted in as commander to oversee operations as well. Again, if it takes a little effort to start a game, then voting a commander will be insignificant. This eliminates most, if not all, problems which would be encountered in building MP support for co-op. Plus, you can have pick-up games with friends, and having them join in on the middle of an island assault wouldn't be a strange break in immersion.

Go to stockpile island (not sure if this works anywhere else connected to it) EDIT: I've had the carrier restock (I think) by itself after going to a new island.

Because I want to knock them over on my snow island, drag them up to my command center, and invite my enemy over for a christmas party. Then back to shooting.

The Carrier AI Turrets do autofire, but it's heavily bugged. It will work, but only randomly. This also seems to be the source of crashes. Summary: Reason: Access Violation. Illegal write by 119610fc at 1 Class: WeaponManager Fuction: Autofire It seems that any unit attacking the enemy carrier will cause this crash. (Carrier not really there to the code? Or maybe targeting data is overflowing?) EDIT And on the enemy carrier game-stopper bit: http://www.carriercommand.com/feedback/view.php?id=35, fifth note.

Yeah I have his domi posts. Issue is, he deleted all domi related material on dev-haven, and if there is anything in that thread of important, it will take me a week to sort through all of the posts. A lot of what I know about domination is by experimentation, as I never found any thorough documentation on it.