By the way, all of this is conjecture on my part and personal opinion. I am not a game designer, at least not of adventure games, and certainly not of anything notable. 
I have one quick question, because from what you wrote I cannot imagine the answer:
Ok, so the only reason is you would lose interactivity. But then, let me reformulate the question:
Why donât we just have: left click to act on an object, right click to look, and no inventory? I mean: when you click an object, the puzzle on that object (if any) is automatically solved. (the character automatically uses the right inventory object). Whatâs the problem? we trust the player, so we know that, if he clicked, he has already solved that puzzle in his mind. And we have also preserved interactivity. So is there nothing wrong with this?
(I am starting to believe that the answer could be âno, thereâs nothing wrong with itâ 
)
1 Like
Yes, you might be on to something, DZ-Jay. If we trust the player to think before acting, it becomes very easy to do a good game. We might even put a big âsolve puzzleâ button that appears when you walk in a certain zone. When you press it, the character solves the puzzle automatically with a cut-scene. (not kidding :))
I mean, if we can trust the player not to press this button until he has solved the puzzle in his mind, whereâs the problem?
So the verbs will only be used to interact, not to solve puzzles. (and then they can all become context-sensitive.)
My 3yo daughter had no problem in understanding that the wrench icon in my game meant âuseâ and the feet meant âgo toâ and the hand with the arrow meant âpick upâ. That made me feel the âcoinâ was ok to use.
However, there were adults who struggled with the interface. I think itâs that children learn everything by trying stuff and figuring out patterns, adult make expectations in advance and want them satisfied.
This has been an interesting topic, but I feel thereâs no âperfectâ solution for everything and everyone.
3 Likes
For example: letâs take one of the best puzzles of all times: Guybrush must saw the peg leg of a pirate (an apparently useless object), because he guesses that then the woodsmith will be called to fix it and leave the shop.
How would this puzzle be implemented with this interface? as soon as you walk near the leg, a button âsolve puzzleâ appears on the bottom of the screen (not on the leg). You click it. The game asks you: âare you sure you have solved the puzzle in your mind? yes / noâ. You press yes. Guybrush takes the saw out of his pocket and saws the leg.
I am not sure if this is crazy or it can work.
Itâs crazy AND it can work 
a âdo stuff just by going nearbyâ model would be perfect, in my opinion, for a âchoose your own adventureâ-style game. Instead of a classical adventure game where you solve puzzles to get to the only (or one of the few) ending(s), the game is not driven by puzzles, but by the choices you made, as if the clicks were just paths you take in the narrative. You click on the peg leg, even if you donât know why and what heâs going to do⌠and from that point on, youâre in the âI cut the peg legâ part of the game.
1 Like
Ok but I think the game should tell you âare you sure you know why you are clickingâ? Because I, as a child, when I played Monkey 1, I did not even understand I was supposed to know that! I thought that the game was just doing what was written in the walkthrough book.
Maybe teams should start making click-and-swipe adventure games to crack a different market.
1 Like
After reading all of this I come to the conclusion that designing an adventure game is hard.
I can understand when game designers say: âAh âbeepâ it, letâs just shoot everything!â
1 Like
What a discussion.
I just want to throw one thing in:
Thatâs the Larry 7 UI and it doesnât work(ed): The most time you can use the coin. So the player is focused on the (frequently) offered verbs. He just forget that he can enter verbs. And if he remember that there is a parser, he donât know what to enter.
btw: A good parser wouldnât be a problem today. You can define synonyms, so you can âliftâ or âpullâ a carpet.
So you are saying that the Larry 7 technique (text parser) could work today? or it wouldnât work anyway because âhe forgets that verbs can be enteredâ?
Yes and no.
The UI in Larry 7 works only if you force the player to use the parser more often or if you make absolutely clear that the player has to enter a verb in that/a situation.
But the UI of the old Sierra adventures could work: The player has to enter text commands and the characters on the screen are executing these commands immediately. That would be similar to the input method via voice commands.
And that by itself would be a spoiler. Jesus.
Not necessarily: It depends on the puzzles and the hints from the other characters.
1 Like
Letâs make a classical example. Two people, A and B, are discussing in front of you. A says to B: âcan we speak in private in my room?â. Then A and B go to the room, and lock the door behind them. You follow them, only to see they shut the door in front of you. Now, the puzzle is to understand you need to âlisten to doorâ or âeavesdropâ.
With the Larry 7 interface, the player must first have the idea, then click on the door, choose âotherâ, and type âlistenâ or âeavesdropâ.
Now, you say this wonât work because the player has forgotten that they can type a verb. Now, how do you suggest to the player that itâs time to type a verb, without spoiling the puzzle?
1 Like
Also you have to face concrete problems due to the âissuesâ of the moment. For example:
When I had Amiga 500 I never had problems with Monkey Island floppy disks, the 4 disks were just loading fine, the game was cool and played and played again. But I had also Larry 5 and oh shit⌠it was something like 5 or 6 disks if I remember correctly and the intro could take 15 minutes to load, then you had to right click to switch actions and the mouse was not responding well⌠then you opened a door and it asked for another disk to load⌠and so on. The result is been that it was so frustrating playing Larry 5 that I never played again. Well, now there arenât these kind of problems anymore but I think that a good interface is 40% of the game: you can make an incredible story but if for some reasons you donât like the way you control the character you end up to quit the game.
yesterday I played Chronicles of Innsmouth, where they have a very nice âlistenâ puzzle, which I solved by chance. I got close to the bush by chance.
Also, in thimbleweed park, I wore the hat by chance. And a couple of other puzzles I solved by chance.
Itâs even worse than that. If your argument âthe player can forgetâ is correct, then this argument applies to non-text interfaces as well. Take the SCI sierra interface, the one with hand-mouth-ear. âearâ is only used once in the game. So the player can forget that it can be used. So, we have to conclude thatâs a flaw in the interface? How on earth could we make a game where the ear is often used?
Canât we just say: âif the player forgets, he lost.â? If he is not able to win the game due to a fault of his own, whereâs the problem?
I really dig this idea for a story-driven kind of game, where the playerâs decision at any number of âdecision pointsâ sort of congeal together and drive the story towards a unique ending (The Stanley Parable comes to mind). You could mix in some more discrete puzzles to keep it varied, as well.
For the interface, the single-click works fine, but there should be some indication as to what your character will do, such as:
on_mouseover(pegleg) give these options:
Saw off the peg leg (if saw in inventory)
Compliment the pirate on his peg leg
But this would be a spoiler. You are not giving the user time to think and solve the puzzle by himself.
it seems better if the verbs that come out on mouseover are only obvious things, that donât solve puzzles.