Ok. I feel it's a bit cheeky to do an "official feature request". Battlegroups Instead of hotkeying a unit/group of units by pressing ctrl+1, we create a battlegroup by pressing ctrl+1. Creating a battlegroup essentially works like fleets from Sins of a Solar Empire. It allows you to set the build destination for factories to be the different battlegroups, which use an arbitarily selected battlegroup leader as the destination. UNLIKE fleets from SoaSE, when the battlegroup leader dies, a replacement is automatically assigned, and the build destination changes accordingly. SubGroups I played a LOT of Rise of Nations. A key issue I had was this - you'd create an Army. The Army would be of mixed unit composition. You'd tell your army to attack a target. So your siege units would attack the target, and your cavalry would charge in and attack the target, and your infantry would charge in and attack, and the enemy AI (yes, AI, not even human) would attack your army from behind while you were setting up another attack somewhere else, starting with your siege engines. So in practice it was better to create two armies for each single army. One was the army itself. The second was just the siege units. A third group was worth creating, to place on Protect the Siege Units, but you often wouldn't hotkey them. So subgroups would be to select groups of units within a main army. I'm not sure what you'd use for this. Perhaps Num keys? Perhaps triple taps of the number key, (one to select, one to center camera, one to select the first subgroup?) I think actual implementation would be messy. So perhaps a great idea in principal, a less than ideal idea in reality. The Advantages: I feel that using these Battlegroups would make a lot of control systems more fluid and allow for better gameplay. I assume there is some smart AI in the game. . So setting up your commander in a battlegroup would allow you to allocate fabbers to assist him. The members of the battlegroup would repair one another, and assist one another with construction projects. You can then assign the battlegroup an escort (of type 1 stingers and doxes) Also, new units joining the battlegroup would automatically assume the formation of the battlegroup.
I'd love to have some sort of grouping system too, but my guess is that Uber is already planning to add that with the new UI.
you can add this to the GUI suggestions list thread : https://forums.uberent.com/threads/community-requested-ui-features-0-1.53170/
I wouldn't say this precisely, but something to help with mass army management. At the very least, rally-to-assist for fabbers will be in, or I'll eat my boot.
Any sort of battle groups would require plenty of automation from the game. - Keeping formations - Equalizing unit movement speeds to keep coherent group - Units protecting each other across the types and keeping correct layout (eg. artillery in rear, brawlers on a front, fighters protecting bombers, and so on) - Automated Repairs and retreats to 2nd line - Giving player a control over automated targeting priorities - Area attack / defend orders (instead of just unit-based) It's by far more complicated functionality than anything we have in the game right now or anything I heard of in Uber plans for PA.
They are reworking the UI and I know they have said they want to add many of those features so it's just a matter of time.
Picking a target priority per group would be pretty cool. I guess this would have to be simplified down to maybe choices based on which factory it comes from (bots, bots Adv, defense, buildings etc) to make the drop down list short per group. That way when your units clash, they immediately take out any Adv units if in range or defences etc. Based on preference. Generally in agreement on all of this. Iffy on the automated repairs. i could go both ways in that argument so I wont argue..
Hey guys sorry I haven't forgotten this thread - I've got stuff to type up, but I need to revise for exams.
One hundred percent support smart groups. PA's increase in scale means that there is a greatly heightened burden placed on the player, unless low-level automation also reduces the amount of micromanagement required. Basic actions like move and attack commands should be abstracted out using low-level automation. Players should generally give orders to groups of units, and make decisions about groups of units, not individuals, unless the player wishes to assume low-level control. However I think you have a fairly narrow conception of what might be possible in this area, stormingkiwi. For one, I see no reason to strictly limit the player to selection groups 1 through 9. The creation of a group of units for logical purposes should be separate from selection recall. Platoons At the lowest level, I think players would find it useful to be able to bind a group of units together into a unitary whole. Once bound, they will stick together, be selected together, and execute orders together, until killed or unbound. A squad of a few raiders or assault units might be bound together using a key combination that is always used to bind a platoon, such as ctrl+q. Clicking or box selecting any member will then select the entire platoon, and and orders given, such as move, will be obeyed by the entire platoon. This is also potentially useful for grouping unlike units into a hard object, such as grouping artillery with a group of guards, or ensuring certain units stick together in specified mixed groups. Abstract Groups Impromptu platoons created on the fly are useful, but their flexible definition means you lose a lot of simple but useful functionality. An abstract group which has a definition can be much more powerful. Defining a group of units allows a group to be ordered to be constructed easily. Instead of manually enqueueing X of unit A, Y of unit B, etc. etc., you enqueue one Armored Battalion or whatever you named the group. Potentially, a group of factories will then intelligently assign build orders to create the desired group of units as quickly as possible. Abstract groups also enable quite a bit of very powerful management functionality. For example, it would be possible to reform a blob of units into as many complete groups as possible by merging multiple groups that have suffered casualties. Such groups could also request reinforcements, possibly from a group of factories. Any units missing from the group will be enqueued and rallied to the group's current location. Transports grouped with the factories might be used to deliver newly produced reinforcements automatically. If the group contains transports, those transports might automatically load and unload when given the order to do so. When moving large armies over longer distances, having each specific group already associated with a specific transport saves a great deal of loading micromanagement. Instead of manually routing units to the front, the player could increase the desired size of an army by an additional Armored Battalion. The next group of that type created nearby that isn't already spoken for by a different army will be automatically rallied to the army, increasing its size. Requisitioning different groups of units would allow the player to easily order new functionality for the army, such as requesting artillery, anti-air, gunships, air support, and so on. And once obtained, that army will automatically request reinforcements to maintain the player's specified size/strength, including any new groups. Automate Non-Decisions The basic idea is to ask the player to make any strategic or tactical decision during the game. However any mindless, tedious, and/or repetitive task that is not a recurring decision should be automated. Setting up the process may require a decision, but its implementation can be automated until the player decides its operation should be modified. After deciding what kind of units to build, the player shouldn't have to repeatedly enqueue more. Solution: infinite queue. After ordering an army to move to a particular area, the player shouldn't have to manually give each unit type (different speed) different orders to make them move together intelligently. Solution: groups. After ordering that a battlegroup be kept at a certain size, the player shouldn't have to manually rally every newly produced unit to the front lines. Solution: automatic reinforcements. And so on. The potential for advanced command and control in RTS games is absolutely immense. Moving past giving primitive move and attack orders should really be a priority in RTS these days. If Uber doesn't implement advanced command and control UI and automation features, I hope someone will step up to the plate to revolutionize how we play RTS games.
Ok dokey! First off, I was wrong about Rise of Nations. Rise of Nations did allow you to set an army, and requisition troops directly to that army. It also allowed you to use "alt"+"right-click" to make forces siege a location. The rest of the army would protect your siege units. I do know there was some reason I wasn't happy with it though. I'll have to play the game again to find out what that was. I'll just cover automated repairs first - automated repairs is a combat behaviour. I have seen it work very well in Sins of a Solar Empire and Earth 2150. I understand how it could be HORRIBLE (100+ fabricators supporting the enemy defenses). On scalability - Earth 2150 and Rise of Nations both did platoon wide formations. Earth 2150 was rather simpler (square vs line), Rise of Nations allowed you to select a line, a chevron, a few other shapes. Those two games were released in 2000 and 2003, respectively. Earth 2150 also did retreats based on a range of conditions. (health %, out of ammo, etc., as well as targeting priorities In addition, Rise of Nations did organise troops in someway. Pikemen would be in the front of armies by default. There was no equalization of unit movement unfortunately, so when you moved your troops it was always invariable that your cavalry would arrive first, your troops second, your siege engines much later. I think that's what annoyed me - the lack of unit movement as a cohesive unit over a long distance. Personally I found this series of articles very interesting: http://christophermpark.blogspot.co.nz/2009/06/designing-emergent-ai-part-1.html He's an indie developer, but gives quite an insight into the AI. In one of those I think he talks from a technical point of view, talking about system requirements. Essentially, the graphical intensive role is the He talks about this idea of an AI Commander, AI subcommanders, and AI individual units. I can see how sub-commander routines could be integrated with the idea of the battlegroup, to manage the battlegroup dynamically. I don't think this would be hard to engineer so that it was technically feasible (For me it would, because I'm still learning how to program). It would have to be thrown onto one of the threads of your CPU. I think Christopher talks about the AI during his article, bearing in mind that that AI is actually more complicated. Anyway, the article by Christopher is especially interesting for me to read. Apparently Sorian is one of the best in the business at video game AIs, so I believe that if he is able to integrate such a system into the game so that it works, he will. For me personally, I enjoy using other hotkeys, mostly to do with commands that are important but tend to glitch out, or to do with build order etc.. I can see other key commands that would be desirable (you just need two additional combinations, one to allocate and one to recall, and you have 100 possible selection groups. Which I think would be more than adequate). There are many ways of implementing that, but you don't want the UI of the controls to be too complicated. Infinite queue is such a basic RTS component that I fully expect it to be in the game For the factory in Rise of Nations, so you could queue up one heavy infantry one archer, one light infantry. That's all for today folks. Back to learning how to program in C[89]
I think it's very relevant, especially since we have a larger unit roster. It would be awesome to be implemented in one way or another.