Powerful UI or Not?

This thread is intended for polling simple radio button opinions only. I know people have wildly differing opinions...

Clarification of poll choices:
If you know your preference for your own UI and you do care about what UI others use, choose from the first set of 3 choices.
If you know your preference for your own UI but don't really care about what UI others use, choose from the second set of 3 choices.

By UI power I mean how high a level of player decision is the UI able to carry out. Powerful does NOT mean "decide for me" or automation. Let's not worry about how complex it will be as the goal will always be to make it as simple as possible. Similarly, don't imagine some worst-case scenario where the UI was not implemented properly, choose what you would want in an ideal UI.

Examples exaggerated for illustration purposes:
A UI that could carry out a decision such as 'execute my intricately-crafted plan to build and maintain a base on a new planet' would be considered powerful. A UI that does not let you queue orders in your factories would not be very powerful in that regard.

Hell yeah super powerfull UI for the win!

While the question seems loaded, you should absolutely count me in for high level UI control. Repetitive low-level commands aren't that fun to execute after the millionth time.

[X]I want for myself UI power EQUAL TO TA/SupCom; I don't care for everyone else

SupCom + a few ui mods (better hotkeys to build stuff quickly!) was pretty much perfect for me. I dont want any kind of weird automations, I am pretty happy with placing mex myself or splitting my army myself, cause only that way I have full control over what is happening. And thats needed.

Also I dont care much, what type of gui my opponent is using, as long as it doesnt involve cheating through more information being display ofc.

Why is this even a poll? Obviously Uber is going to make the UI as versatile as they can.

Let me get this straight, you're calling queuing the User Interface? I can't help but feel "UI" is not the best choice of terms for this, but I really can't come up with anything better, so carry on.

Regardless, queuing and designing patterns like in FA is very important to me. If we could add macros to the mix and allow us to design as your first example in the first post suggests, things could get VERY interesting.

As for the actual interface, I'm a big fan of customization. For my work as a naval architect, I use AutoCAD and Solidworks extensively and have thoroughly customized the UI of each to suit my particular style. This I've done by adding buttons here, moving some there, and setting up mouse context menu options that I use very frequently, while binding keys for other things.

If we had a UI where we could pick and choose what is displayed and where certain controls or info is shown, many advanced players would benefit greatly. There may even be a possibility for trading UI builds between players, which would allow new players to download the UI build of advanced players and start off with a nice advantage.

I am aware of the stated opinions from Uber. Their vision for PA's UI is probably not going to be swayed by a single poll. However, there has been much debate on these forums about how much power the players think the game should provide. I became interested in knowing in rough numerical terms how the population felt. I'm sure Uber is interested in this feedback too.

I tried to word it as best as I could to differentiate between a player's preference for how they wanted to play the game and also if they had feelings about how others were playing the game. As a disclaimer, I prefer a powerful UI and I believe other players should be allowed to use as little or as much of a powerful UI as they want. In other words, not forcing anyone to do high-level or low-level, but it is all there available to you.

I'm not sure what most people would think of when they read "user interface" in the context of a game. For me it is the system that facilitates the user giving information to the game and the user getting information out of the game. In my original description above I suppose I have focused more on the side of the user inputting information.

I am in total agreement with you. UI customization is an excellent way for everyone to tailor the game to their tastes. If PA had it, I would use it to the fullest. A downside which I fear may end up limiting the customizability is that of fairness. Some people may not like having to fiddle around with options and settings and possibly looking around for mods; and it is possible that they would be gameplay-disadvantaged for not playing the 'customization metagame'. I'd bet that everyone has experienced the significant drop in productivity when switching from one's own optimized UI to a friend's computer having everything on default settings.

The vocabulary used in this thread threw me off a bit. The OP and the example in it are talking more about a user experience, instead of the user interface imo, e.g., what the game does for the player in a nice helpful way.. much like communication

If the thread is talking about the user interface, e.g. the HUD and its scale-ability, then I would definitely say that I liked FA's HUD a lot, but I am all for being able to customizing a hud. For example, if you are working with multiple planets, it would be cool to be able to "pin" the planet's resource income, expenses, and current level of enemy threat to it, to your hud. That way, you can tell if it becoming more of an expense than a contribution and compare it with the other planets, and easily see alerts to when an enemy has landed on one of your planets. However, you might not want to see all the planets--thats where the user's customizations come in.

If the thread is talking more about what the game will build/automate for you based on your style i.e. user experience, then I would say, in the same manner, more customization is better. Let's say for any particular planet you have have forces deployed on, there is a planet management window-- in which you could customize things like: queue build on this planet only when you have +X amount of metal and energy flowing, and not <+X or -X. Your build queues there will pause until your metal/energy goes >+X (i.e. you have excess and you can't spend it fast enough by yourself). Then, from this particular (let say the first that you conquer after your original) planet, the game will apply this general setting to future planets, and/or you have a global setting check box which allows you to have one across all and uncheck it in the planets where you do not want these options applied.

This is a really dumb poll.

veleiro's post helped me understand better the sources of confusion. Thanks

So -- I will try to clarify more on the words and concepts I'm using. I don't expect everyone to read all of this. It is really for those interested in more technical and abstract things.

The actual screen elements such as the hud, icons, cursor, buttons, text displays, graphs, do form part of the user interface. They are the very top layer that players interact with. They give players information (mostly visually), and allow the player to input information into the game by pressing keys and clicking and dragging with the cursor.

Explaining UI power
To start with a low-level example, take 'selection'. Suppose a player has decided to move several units all to one location. If multiple-unit selections was not supported, they would have to work around this by translating their plan into "I want to move this unit, I want to move that unit, I want to move yet another unit", which the UI has the power to understand. Thus, the player is still able to make their plan happen and no gameplay ability is lost, but it wasn't due to any power the UI had to offer.

Explaining UI efficiency
Contrast this with efficiency: suppose multiple-unit-selection was supported. One UI only allowed shift-clicking to add to a selection and another UI also included dragging a box around units. They would both be more powerful than the UI from the previous example and be equally powerful with each other; but the first UI would be less efficient. The player's desire, "I want to move these units", requires many user actions (multiple clicks) to execute their plan versus one drag and one click. Both UIs are able to understand and carry out the player's wishes; no gameplay ability is gained or lost. The UI that included bandboxing is more efficient but also slightly more complex (having two forms of multi-select).

UI power and complexity put in context
Now, the lowest-level actions in the context of a typical RTS unit would be moving, building/repairing, and attacking. You could say these are 'atomic' actions. Every possible gameplay decision a player could have can be broken down into these lowest-level actions. A UI supporting only these commands would capture all the gameplay potential but have incredibly low UI power or efficiency. On the other hand, the UI complexity is very low. It does not understand queueing, but you can effectively perform a 'queue' simply by translating "queue command1,command2,etc" into "execute command1", wait for it to finish, "execute command2", wait for it to finish, etc. The UI lacks the power to understand your higher-level command and is unable to translate it for you; you have to do it yourself. 'Patrol' is simply repeating a 'move but if you see something in range, interact with it, then go back to move'.

So I hope that explains what I mean by UI power, efficiency, and complexity. It is not about asking the computer to make decisions for you or about what gameplay possibilities exist or don't exist, but how the human and computer exchanges information through an abstract layer we call an interface. An analogy would be trying to explain a task to someone so that they can perform it precisely. Consider how productivity would change if the task were more simple or complex, or if the other person's capacity to understand your ideas or language were higher or lower.

The Importance of UI power in large-scale RTSes
Now I will try to explain why, for an RTS to scale up, UI power is important. The key is that it is in realtime, unlike in turn-based strategy for which UI power is more of a convenience factor. The player's resource of real time per game time unit is fixed and limited. In turn-based, you have arbitrary amounts of real time per unit of game time. In an RTS, a player is given exactly one second of real time for every second of game time. A very complex set of actions such as 'build a base' would take many time units to execute if done at the lowest-level. This would mean, over the course of a game, not many of these complex sets of actions could take place because there just isn't enough time for players to execute all the commands. A faster player could do more, but those players are rarer and there is still a pretty low limit on the number of complex actions they could perform over the course of a game. Furthermore, PA wishes to focus on strategy, and being unable to carry out some strategies for lack of UI input time limits the game's strategic potential.

An example of inputting a high-level plan into a powerful UI
My final example will be showing how 'execute my intricately-crafted plan to build and maintain a base on a new planet' can be accomplished with a sufficiently powerful UI and without any decision-making by the computer. For simplicity's sake, this plan will be intended to be carried out by one engineer (you are allowed to assign more or non-engineers, to varying results). The UI supports orders as first-class entities, which means we can create orders and plans 'right on the map' without yet assigning any units to it. The entire set of actions is in a queue:
1. Create a move command somewhere on the new planet.
2. Create a build radar command.
3. Create a build factory command beside the radar and pre-load it with orders to build 2 engineers followed by 10 light attack units and 10 heavy attack units. Pre-load all queued units with orders to guard a circular area centered on the factory; set firestate to attack everything; set 'persistence-state' to not leave the circular area for any reason.
4. Create a build mass extractor area command in a big circle around the factory.
5. Create a build defense towers area command in a ring around the factory. (The newly-built patrolling engineers will eventually assist with this if they aren't too occupied with assisting their mother factory with the attack units.)
6. Our original engineer has now finished, and is given the same circular area guard command as its comrades. Optionally send a message alerting the player of the completion of the plan.

To actually execute this plan, simply find a spare engineer and right-click them onto the first order. They launch themselves off to a new world...


You may think, gosh, the UI could be really complex and it could take me just as long or longer to create this plan and execute it. For a single-use plan it is probably easier and quicker to execute some of it yourself instead of spending time detailing it to the UI first; nothing will stop you from continuing to do things that way. For executing this plan or variations of it many times, however, telling the main details to the UI once will save you much repetition and real time in the long run.

UI complexity is something that can be worked on and improved: look at how RTS UIs have kept their power but simplified over the years. So it is not a given that a powerful UI must be complex. The goal is a powerful UI, and it is the job of the UI designers to try to create the simplest and easiest-to-use UI without sacrificing power. Just because you may not see how to implement it doesn't mean that someone else would not figure out how. It is certainly a difficult challenge.

What if there are obstructions on the new planet, or interruptions from other units?
Well, the plan in the example above did not account for this. So, where obstructions exist, an error alert could be sent or the command in question will be ignored like in existing UIs. If other units interrupt, our strategy did not account for this, so we would have to intervene or the plan could be in jeopardy. The computer does NOT make decisions for you.

What if my plans never work out the way I want them to? The UI is too hard to use!
That is definitely a possibility whenever greater power is presented to people. The UI does not enforce any particular way of doing things. You may make use of all its functions or only a subset. Limiting UI power because not everyone will fully utilize it could satisfy more people, but at the risk of putting an artificial limit on the things people could potentially accomplish with the game.

How is such a UI more powerful?
Because it is able to 'understand' the player's wishes as described at the high-level, akin to telling someone a complex idea and they reply "yep, I completely understand what you mean", instead of "uh, I don't know what you mean by this" or "I don't know how to do that, please tell me in simple terms". Allowing the computer the ability to understand higher-level commands means that, from there, it is only a couple clicks to actually execute it once or many times.

What about the feedback side of a UI?
The feedback side of a UI is important overall, but for the purposes of executing high-level strategy, it is not the primary focus. It certainly benefits the player through saving them time by having such feedback tools as visualizations of your plan so you can see what you are crafting, ETAs and progress indicators where appropriate, able to more easily monitor more things from a super-zoomed-out view, graphical summaries of your empire, alerts presented in a prompt and useful manner, and all the many kinds of GUI goodness that we're all used to.

How would this make the scale of PA larger?
With plans being first-class entities, you can easily save, load, modify, and assign and re-assign to units. In short, they act like templates. To re-use the example plan above, simply load it up, set it down on the planet of choice, and assign an engineer to it. Naturally, someone taking the time to input a complex plan intends to save time over the long run. They may have intended to establish quick cookie-cutter bases in many new locations: the total cost of inputting the plan plus executing it many times is lower than having to manually execute the plan many times over. This allows the player to accomplish more complex sets of actions for a lower real time cost. The rigidness and lack of 'intelligence' in the plan would make it less successful on average than someone micromanaging each and every new base. This is a time-saved tradeoff and I think it is a good balancing measure for an RTS game.
A different kind of plan could be a battle plan. You design and input it once, then use it many times in the sense of sending many units at different times over the course of the game on the same battle plan. Battle plans tend to be dynamic, so it is a valid point to say that the re-use of battle plans without modification has less strategic value. Modification of plans is easy: it is like visually editing a queue. The player will have to weigh the potential costs and benefits of giving their time towards microing a battle or letting battles follow a plan and spending their time elsewhere. If the cost of not microing a battle becomes consistently too high, the choice to spend time elsewhere is effectively lost and with that the game loses potential higher-level strategy.

How would this increase strategy?
Players using less real time to execute plans means more time can be devoted to creating said plans, allowing them to be more complex and able to encapsulate more and grander strategies.

I truly believe PA can be a next generation RTS and this would be one of the ways it can.

Ok now I get it. Yeah, plans are a good idea, but implementation could be a problem.

There were similar discussions for Supreme Commander, or at least in a similar direction. For instance, it was the opinion of some people, that any UI mod, that gives the user a significant advantage is basically cheating, because others might not use said UI mod or even know about it. Examples were Goom's Range Checker and Advanced Groups mods, that I requested from him and which I used extensively in every game.

I have the same opinion as Cola_Colin, "anything" should be possible, as long as it's not "actual" cheating. e.g. displaying information to the user, that he would otherwise be unable to obtain, or rather supposed to be unable to obtain, or executing commands, that he should not be able to do etc.

On the other hand I can understand if the devs want to limit the possibilities of interaction with the game, so that everyone can play on more or less equal ground.

A simple solution to that would be to list the mods that everyone is using so everyone knows what's going on. You could also automatically download the UI mod in the lobby of the game you've just joined by clicking on the list of mods other players have.

Let's make sure the UI is as flexible as possible and that everyone can easily get every mod they need.

My one concern with this, is that it quickly becomes players playing against AI's operating under the supervision of a player through their UI.

Meaning, it is a fine line between the automated functions keeping the massive logistics and unnecessary micro in check, and playing a glorified computer capable depending on if the player is paying attention to your area or not.

I would caution about considering aspects of the game in just the "how does this affect my play" mindset, as it is very directly related to "how I play against this" also.

An example of this is patrol orders. If memory serves, assigning a patrol order in all of the Supcomm line games and in the TA line rolled either an attack or repair command in with the order. The AI wasn't exactly intelligent about what it would assist repairing with, or what it would attack, but it would free up a fair bit of time spent doing micro. This i think is good.

Being able to macro a build order so that say, construction units from x building automatically move to point (x,y) and begin construction of PD is bad. Having factories auto construct units when attacked, also bad. In general, any time the Players AI begins to make "shaping" decisions, is a bad idea. The AI should not be capable of strategic choices, though tactical choices could be fine (Example, coding so that AA towers are always repaired first on a patrol path for engineer units, vice mobile attack units, or that a unit that needs repairs and is about to receive them stops moving so that it can receive said repairs).

The UI should be simple enough to enable to player to very quickly assign these tasks to the AI, without needlessly complicating the players strategy. I.E. I don't want to have to select "repair all Rocko's" from a drop down menu after i select my engineer unit and give it a patrol order. These aspects should be organic and out of sight of the player. Let the modders tinker with it if they so desire.

I don't want Automation, I want powerful tools.

Mike

Automations are always tools when they are controllable.

But it leads to them being subar as tools.

I'm the commander, I make the decisions.

Mike

How is this anything different then what I said?

A game of this scale really needs a powerful UI or it will just get messy and disorganized. I really hope it will support modding.

Just two very important to me things I want to ask for:

1. Every single bit of UI (except for the most main of all main menus and the show/noshow menu itself) can be toggled between show and noshow.
2. Every single bit of UI can be moved around on the screen. Partly why I didn't like FA UI at first after playing the SupCom1 demo some time before was that it was stuck all around in the corners, leaving the view and command area a garbled beyond uneven polygon instead of a neat clearly defined rectangle. Yes it was transparent, yes it was toggleable between show and noshow, but only the minimap was movable and resizable.