DEMO: Combining text adventure and 3D graphics

I’ve been planning on creating a demo of a game that combines the best of the text-adventure and 3D graphics genres for a few years now. I finally found a good quality level online for the 3D part of things and have been working on adding a solid text interface to it. Being given a level and then having to create a plot and puzzles that fit into it has been challenging, but i’m almost at the stage where I can release a playable demo.

My goal is to use this demo to put a team together to make a real game with the same mechanic.

I was hoping to get some feedback from the community about my progress so far. At this stage I have a brief video of the introductory portion of the level that demonstrates the basic mechanics. The game itself is currently fully playable, however in the video I haven’t solved any of the puzzles so there are no spoilers. Here’s the link:

I’m very interested in feedback, so if anyone has any comments please let me know!

I’ll be honest, I was entirely sceptical when I read your thread title, and I wasn’t very hopeful clicking on the link. Based on 3D text adventure hybrid experiments I’ve seen in the past, I was expecting horrible non-moving graphics and not much text adventure. But this looks awesome! Seriously! I wanna play!

Wow, I’m certainly impressed! Is the parser pure C#/Unity, or are you using Zifmia or similar to let another program parse the commands?

Good lord, it looks extraordinary.

Do want!

In the past, this is what I imagined adventure games will evolve into. It’s so annoying that everyone wanted to get rid of the parser.

I suppose it’s because once you have such a gorgeous 3D setup like that, it becomes easier and less time consuming to get rid of the parser and introduce a P&C interface. I mean, let’s not kid ourselves here - proper 3D is hard, and a proper parser is hard even for us, who use pre-existing parsers honed over time. Who has the time and patience to combine both?

Auspaco does, apparently. Joy!

EDIT - Hey, it’s funny. We’re embracing a sort of IF hybrid where we do away completely with the “We’d rather give you the 1000 words” paradigm, which has long been one of the main selling points of IF, and where we keep the parser, which comes not too long after a long discussion of whether or not the parser is worth keeping.

Food for through.

Vespers 3D is still in active development:


I’m looking at a display showing a youtube video showing a 3D game showing display panels on the wall and showing > look display on the text panel…

I feel this is too deep even for DiCaprio…

And it’s also very cool. It’s just that development, although active, is so quiet and beind the scenes, one simply tends to forget about it… for shame, to be sure.

For the past few whiles, a changelog has gotten posted to the blog more or less weekly, so it’s actually louder than most WIPs in the IF world… But it has been in visible development for a much longer time than most projects.

That already happened with 2D. Nothing new here. It is a shame; the parser does add a huge amount of depth to any adventure game. It is the only way you can really express to the game what it is you actually want to do other than some generic “use” action.

It wasn’t the expressiveness of the parser that struck me in that demo video, but the characterization provided by the AI. You could almost do that with a voice-over but text has advantages over that I think.

So, how does the Unreal Engine (that’s what the menu bar indicates, I think) interface with the parser? This is quite fascinating.

The usage of the parser would have to see some modification for a serious first-person 3D text adventure. For one thing, the EXAMINE command should be irrelevant. We see what everything looks like. For another thing, the 3D engine is in real time, and the parser will probably have to be in real time too.

Not necessarily. There’s a lot more information that the protagonist could have about an object than is conveyed visually. Gods know I’ve played plenty of graphic adventures in which the image of something doesn’t really let me figure out its significance, even though the protagonist obviously understands what it means.

So: does the protagonist know anything about this object? Does it evoke any memories, emotions, aesthetic responses in them? What is its normal use? Is it in its expected context? What does this show us about the protagonist, or imply about the world? And so on. That’s IF Item Description 101.

Wow! Thanks for all of the great feedback!

The parser is implemented in unrealscript (UDK 3). I did think about writing a separate parser in C and then interfacing it but the advantages in having an integrated system were too good to pass up. For instance there are lots of things in the game that are a “bin”. The parser has access to where things are on the screen or if they are being blocked by something to help it resolve those conflicts.

The text interface for me was the reason to make the game. I did consider removing some of the redundant commands like EXAMINE. In the end I allowed a small amount of automation - like looking with the mouse button, and auto-examining anything you pick up. It all comes down to personal taste, and maybe it’s pure nostalgia for the sierra games but I really love typing things in!

If anyone is interested in beta-testing the demo (Windows 7 , approx 120mb) let me know! I’d be keen to iron out as many bugs as I can before letting it go.

Doesn’t detract from my point.

Hi. I’m new here but the video piqued my interest so I thought I’d throw in my 2 cents …

My first thoughts are that this is really cool, but it might be difficult to find the right “compromises” between the 3D and the text. Both mediums will need to be justified in order to rationalise the effort that goes into building a game like this.

For instance, one of the advantages of I.F. seems to be that you can construct any space, of any dimensions, using only words – there are virtually no limits. So a huge underground cavern with pools of lava heating the air and reflecting from the walls is described in a couple of sentences. 3D requires everything to be manually constructed, and then rendered at an acceptable frame rate – that same lava cavern suddenly takes 20+ hours of building work, and still might not appear the way the author intended it to.

Also consider NPCs. I’m sure it’s not easy to write a realistic and fleshed-out NPC into a traditional I.F., but it only becomes more difficult when you want to implement that same NPC into a 3D space. Modelling, textures, animation … Will you have voice acting? Compare the feedback, “The butler smiles nervously at you.” with actually showing that on the 3D model.

On the other side of the coin, the advantages of 3D are obvious. A true sense of place within a 3D space; actual sounds for your ears; the ability to move your PC physically … In fact, a lot of the more simple commands that we might give in a traditional I.F. (“put card in slot”, “open door”, then “go north”) can be well-catered for with the first-person interface (drag the card to the slot, click on the door to open it, and then move your character through it). The question then becomes – why is the text interface there at all?

Clearly text can convey much information that a 3D game can’t – the smell, air temperature, vibrations, how the PC feels (hungry, sick), memories the PC has, etc. It also allows much more complex instructions from the player than a 3D interface alone would – “ask beggar about cheese” or “look through keyhole”, for example. It seems to me that the hard part will not only be combining the 2 mediums into a smooth experience (although it already looks pretty slick in this early demo!), but justifying the huge workload of implementing a full 3D game on top of a text interface.

Having rambled on like that, I hope you understand that these are just some thoughts for your consideration and that I was impressed with the video. I sincerely hope to see the project develop further! :smiley:

  • Gray

I’d like to premise this post with the obvious fact my opinion on this topic means absolutely nothing.

With that said I’m not a fan of adding graphics(3D or 2D) to IF. In my opinion the only thing IF has going for it (aside from being able to have a development team of one) is the fact with the right writing the visuals generated in my head are a million times better than anything a computer can generate.

Written text is significantly better at sucking someone into the world of the story. Look at movies as an example. Have you ever heard someone say the movie was better than the book? And movies have large teams of people working years to develop.

Seeing IF and 3D combined my brain instantly see’s a game that’s not as good as it could be. Either the graphics aren’t going to be good enough or the text looks like a game with a lazy game control. Why not just use voice actors? Why not find a control setup to achieve these functions. Why not let the room describe itself with the visuals? Then i read this comment

This is very true. But then we are talking what IF features can add to a 3D game not what 3d can add to IF. You’re demo displays this very well. But I see it as a 3D game implementing IF techniques. Having the IF basically being a HUD of sorts works very well. The game has significantly more promise than I would of imagined. Though I think tweaking how the interpreter plays would go a long way. Like was said before typing examine seems a bit silly when you should be able to see it right in front of you. Maybe analyse or add to description archive or something like that. Idunno it’s your game to figure out lol.

Also adding in text to voice and vice versa capabilities would go a long way in my opinion.

Either way the fact i started this video with the preconception it was already a total failure and I now see potential is beyond impressive to me :smiley:

no doubt

here’s a 1982 game:

and here’s another from the same year:

in the former, they cut down the bland text to bare basics so they could get space to the jaw-dropping cutting-edge graphics of the time

in the latter, all the screen is taken by fine prose that still reads great today

in other words: IF should cater to readers, not to mindless gaming drones, the kind of which are tired after reading a long twitter post and that are lost if gameplay involves more than mindless button-mashing or link-clicking

I agree – but if we’re not asking should somebody make a hybrid game, but rather how can they make their hybrid game as good as possible, I think the problem becomes quite tricky. The 3D simply has to be great. Anything less than great graphics and movement will detract from the experience, not add to it. As namekuseijin said, the text needs to do what I.F. does best, otherwise you may as well just have the 3D game without the text. So it’s a problem of balancing the two against each other …

Now, the demo looked fantastic. That’s what impressed me immediately. Like Sunkin and others, I expected a lot less when I clicked on the video link. The way the text part is representative of your suit’s A.I. is a great idea!

That’s why I for one am interested in this project :smiley: If nothing else, it’s interesting to examine the concept of it. I was briefly excited for another certain hybrid game recently. I won’t mention the name but it was a cyberpunk game; I think you know what I’m talking about :wink: That game was immediately and completely destroyed by it’s own prose and grammar, but the idea still interests me. Remember, it doesn’t have to replace I.F., haha. But if it works, it will provide another option for our gameplay – not to mention possibly dragging some fresh players to the I.F. scene.