I wasn''t addressing those particular options or analyzing them in any way.

I received your point about wanting good unit AI.

I just wanted to make the counterpoint of giving the player control where it is needed. Who knows what will be needed when? A player could wish for the simplest option in a given situation and it might not be available. Anything that helps the player is a good thing, but it is also easy to go overboard with it and overwhelm the player, which is the point of this post.

My latest feel about it is that there will be a base list of options that will help any player conduct their battles effectively. Then the ability to make some custom targetting and command options will be available for those who have more specific ideas. Hopefully that ability won''t become a balance problem.

A nitpick to start with: by defintion, "too many" options is a bad thing - the question then becomes how many are too many...

The most basic elements for player control are "attack target" "attack location" "follow target" "move to location" "[special action] (target)" with those and unit groups (and unit command queues), you can get units to do pretty much anything you could do with more elaborate commands - at the expense of having to micromanage. Better AI or more complex commands are often useful, but if you don''t include the basic commands, you will have players unable to take advantage of superior tactical ability...

So, basic commands such as:

* Stay in formation
* Break to attack
* Attack target (intelligently choosing targets if target is in a group)
* Cover target
* Cover me

And these kinds of commands for the player:
* Target nearest enemy
* Target nearest friendly
* Target nearest enemy threatening your craft
* Target target''s attacker
* Target target''s target

Plus a few special ones for the player:
* Target nearest non-targetted enemy (to keep them all busy)
* Target nearest enemy attacking friendly with little cover

might be a good start.

I''m sure I can come up with more. Thanks for the input!

Right now I''m trying to learn Vi and Emacs on Linux, and let me tell ya, all those options scare the beejezus outta me. It''s almost like I don''t know where to begin. But mny people swear by them, and say it makes productivity more efficient. I''m willing to believe them, but something has to be said about the learning curve.

I remember when I first played X-wing back in the day, it was quite hard to try to get all the hotkeys down right. Sometimes we''d play multiplayer co-op....my friend would man the joystick, and I''d play R2, basically re-assigning the shields, blasters and engines, reinforcing the shields or capacitors, and doing all the nitty gritty stuff while my friend concentrated on flying and shooting. It was pretty damn fun actually.

But eventually you got used to rapidly switching all the hotkeys, and I got to be a pretty good pilot (in Freespace as well). The diffuculty wasn''t so much in the learning curve, but in its application. Easy to learn hard to use would be a good phrase. If you have lots of options, definitely make sure they are at least easy to learn...the easy to use part comes through practice.

When something like this crops up, you need to look deeper. In this case, you''re talking about too many hotkeys/key combos. In other words, too many sets of keys of size 1 or greater that need to be pressed simultaneously.

You''ve already said yourself that there are many similarities between the commands, such that you could build the commands yourself. I''d guess that these commands are used infrequently, but when used need to be inputted with speed.

Thus, I''d move towards a sequential input approach with feedback. Picture this: The enemy has pulled his reserve forces in for a five pronged assault on your defensive line. Many of your troops are holding their position while the prongs carve a hole in your line. You select your entire force and press O, which is a blanket hotkey for building a complex order. A box flies across the top of the screen, where it would sit at the left edge if not for the fact that you quickly press C for cover. Instead, the box flies off the screen as another of different size takes its place. You press N before the box stops, making it accelerate off screen again. A small box zooms past; you''ve already pressed F. What kind of a stupid designer thinks you ever want to cover your enemies? Anyway, a long, skinny box zooms to the edge, but you''ve already pressed enter. You''ve ordered your entire force to Cover Nearest Friendly.

Later in the game, you''ve got three of some sort of terrestrial supermonster guarded by fliers. You select your SAM''s and press OCNF like before. You select your ground attackers and press O. The menu zooms into place, where you see:
(T)arget
(A)ssist
(C)over
(B)reak

You click T, which zooms the old menu off and zooms a new one over that says:
(N)earest
(W)eakest
(S)trongest
(F)astest

You click S, which does the same, and the new menu says:
(F)riendly
(E)nemy
(N)eutral
(T)errain (for trees and boulders and such)

You press E, which brings up:
That is in your (A)rea
That is a mission (C)ritical
None(enter)

You press C, followed by enter, strictly from habit, even though it''s not necessary(there are no more options after that)


The problem with using a hotkey in this situation is there is no immediate feedback(takes a second or two before you see what they''re doing) and a multitude of options and few keys on which to map them. This is technically a Wizard interface pattern.

Other possibilities include default options and modifiers. Such as:
Press T to target most threatening enemy. Press T, followed by M to target nearest mines. Press T, followed by F to target nearest friendlies. Press T, followed by F, followed by W to target weakest friendlies. Press T, followed by F, followed by W, followed by E to target weakest friendly unit on your escort list. Order may or may not be important. Find some way to provide immediate feedback (crosshairs zoom in on new targets), especially if this is a persistent state (crosshairs zoom in on new targets, with decreasing precedence indicated by decreasing opacity, as well as the text or a graphical representation thereof (somewhere near the selection description is the text TFWE)), or the interface will have an needlessly steep learning curve.

Disadvantages of this approach is that they may begin immediately to execute undesirable orders while the modifiers are typed, or, if there''s a delay to wait for the modifiers, there will be a delay between information input completion and execution.

In answer to your question of if there''s such a thing as too many options, I''d say NO, it''s a problem of presenting the options in a way that reflects the importance of the various options while also maximizing efficiency. Lists of hotkeys and a keyboard with an importance-agnostic layout are not the best way to do this, although it often the most efficient possibility.

Finish the order sequence with a click (selecting the "arguments" for that order : targets, and unit selected)

I was actually also thinking more along the lines of assembling a custom targetting or command option, and then binding it to a key sequence, so a player can also custom build hotkey commands instead of having to follow a key sequence every time they want something special. For instance, any key sequence could be assembled and then attached to SHIFT-A.

That''s a pretty simple feature to implement in my scenario. Instead of pressing enter, have the player press K, followed by the key combination. If it collides with some other hotkey, just bring up another menu with options for resolving it.