Quite a neat little prototype! The graphics and animations are already looking great. I could imagine lots of places you could go with this. Jumping, aerial attacks, monster types, etc. It already feels good. Keep it up!

This looks neat. How is the trading aspect going to work? i.e. How deep of a system of resource management are you going to be doing?

Yes! If you like dungeon crawlers even a LITTLE bit, this is really worth your time.

I'm interested in creating a largely text-based RPG with minimal graphical elements. I'd like the gameplay to be entirely based on touch/click/UI. No true map system, no collisions, no physics, minimal if any animation, etc. I would like to be able to deploy to mobile as I think that would be an ideal platform for this kinda game, OR alternatively web. I have a decent handle on Java and C# but not too much else.

I've looked at the stickies and I'm just a bit overwhelmed by the options. I've used XNA before so Monogame seems like an option, but the documentation is maybe a little light? I've also spent a lot of time with Slick2D but that might be too low-level for what I'm looking for. I'm wondering if there is any kinda tool or framework that is a bit more oriented toward this kind of thing (data-driven games with no collision/physics etc)

Thanks in advance!

As the topic says, has anyone had any experience with monetizing an HTML5 game? eg. either selling it via a web store, getting a sponsorship on an arcade site, etc. There are many more numbers/topics available for Flash games (sales, ad revenue figures, sponsorships, etc.) but since HTML5 is still just taking off, I've had a hard time finding info on it. Basically, the indie studio I'm a part of (http://www.spacewhalestudios.com/) is working on a fun little HTML5 platformer and we're wondering how realistic it is to expect that we might make, say, $1000 from it if we polish it up and get it on some HTML5 arcade.

Here's an early WIP of the game:

http://www.spacewhalestudios.com/duckit/

Thanks in advance!

More updates! I have hotbar item and skill usage/dragging totally working now. It's pretty cool. NPC dialog is coming along - the text now "types in", which is done using a set of timers and text speed variables. I also have it set up for branching dialog (multiple sub-menus) plus I can store arrays of text within each response option. Nice for variety.

Since this is my first time programming a fully-featured, graphical game by myself, I have to say that I had no idea there was such a huge amount of work involved. Even relatively simple things can take days or weeks, like the revamped UI system. It took awhile to get proper drag-and-drop working; even now, my only solution to differentiate between a single click and a drag was to assign the drag function to the RIGHT mouse button instead of left! Yikes. All things that need to be figured out.

Hotbar is now about 95% functioning! You can drag and drop items/skills there, and use items. It doesn't work with skills yet, but it will shortly (I just need to make icons for my test skills.)

I also have the dialog system partially implemented. It's going to be complex, as per above descriptions. Whereas many RPGs have a simple "jerk/saint" type of system, you'll instead have a variety of tones/emotions for your generic responses. Making a subtle threat might carry +3 threatening, +1 aggressive, +1 sly. Different NPCs will interpret those numbers in different ways. A shady NPC that looks on you favorably may appreciate your brazeness and coy in making a subtle threat. Or maybe not. Fun stuff!

I've been taking a vacation for the last 5 days or so and getting a lot of new code done here. I think my new UI system is all done, finally.

* Dragging and dropping (eg. rearranging your inventory, dragging abilities to the hotbar)
* UI container objects that can have a number of 'children' elements.
* Text elements, clickable and otherwise

It's really, really boring to describe but it took an immense amount of work and I think it's pretty flexible now.

I've also started on the dialog system, which will be another huge challenge. There's not much to it yet but I'm proud that I have character-by-character rendering done. In other words, with variable speed, dialog with NPCs shows up like old-school RPGs ("typed in"). Woo.

Thanks!

I know I haven't updated much lately, been busy with my day job (music composition and sound design). However, we have a thorough outline now for the beginning (prototype) area of the game. The UI has also been much-improved. As I'm learning, it's important to define solid systems early on so you don't have to rewrite stuff later. I now have some pretty good UI code, but there's lots more to do to make sure it's future-proofed.

Been working hard on UI stuff for the last week or so. I think the interface here is going to be incredibly important, which is why I'm not trusting myself entirely with it, and instead am collaborating with our artist, Andrew Luers, to develop something intuitive and pretty. I'm now almost done converting the whole system to something very modular, where new icons, overlays, clickable text and buttons can be dropped in quite easily. This has involved basically tearing down all the old UI code and creating an entirely new UIManager class which takes care of all the UI elements (interactable and otherwise) along with rendering and UI-related input.

The next step is creating the skill/item hotbar, a la games like Diablo 2, Guild Wars or World of Warcraft. You'll be able to drag basically any 'action' to a 6-slot bar for easy access, which I think will help the game's playability and learning curve.

Wow, very cool. I'm using Java to develop my own 2D RPG (a Roguelike-inspired game, more or less) and so I'm interested to see how you're approaching some of these challenges. First of all, are you using something like LWJGL or Slick2D here, or is it all from scratch? Also, have you ever checked out Tiled for comparison?

I've been continuing my work with AI. NPCs prioritize actions/targets differently in combat, and they can now use skills (albeit a bit stupidly.) It's refined enough for now: next up, I'm going to implement the dialog system next. As described in the OP, it will be based on emotions/delivery styles as opposed to having lots of filler text. We'll see where that goes. I haven't really sketched out any pseudocode for it yet...

On the art front, here's a quick screenshot of some of the updated tiles we're using. FYI this is absolutely a work-in-progress. Stuff needs to be cleaned up, made more seamless, etc. Still, I'm very happy with the direction:



Our writer, Steve G. (from these forums!) has been doing a great job with creating an interesting world, characters and locations. I only had vague ideas initially but things are getting more fleshed out now.

The game's setting is roughly steampunk. There's some magic involved, but not too much - it's rare and even treated as heretical/demonic by some. Technology is mainly limited to guns and simple machines as opposed to complex devices, computers, cars, and so on. The first vertical slice of the game will take place in a coastal region set next to an ancient ruin, where the player will have the opportunity to make an impact on the village (and its inhabitants.) I'll elaborate more in future posts :-)

That makes a lot of sense. Thanks guys. With regards to how the game is going to "feel", it's hard to tell. Being that it's a deep, open-world RPG (or at least, that's the goal) it's a bit hard to test the 'fun' or challenge factor of something since right now I'm developing the foundation of the game now. To test NPC interaction, I'm just throwing them in a little pen and testing what happens when I do X, Y or Z. But once I start really structuring maps and making some semblance of a 'game', I'll absolutely be doing further testing.

Here's a followup question. Let's say an NPC has decided he wants to attack something. He might have a variety of skills at his disposal. I figured I'd flag 'offensive' skills with a simple boolean (it's hard to dynamically test whether a skill is truly offensive or not, since a given skill can have many component.) How should the NPC decide to use, say, Poison instead of Fire? Or Thunder instead of Ice?

[How would you approach this?]

I'm working on a 2D, turn-based RPG in the vein of Roguelikes, eg. every time the player moves/does something, all entities take a turn simultaneously (well, there's a priority queue, but close enough.) One key feature I would like the game to have in its finished form is genuinely interesting AI that provides a good tactical challenge for the player in combat.

Here's a quick overview of the system I'm using right now, most of which is implemented on some level and working.

* All NPCs have an array of personality traits. These traits will influence their behavior in different situations. For example, a cautious NPC is more likely to flee than fight (but still might fight, given the chance.) A friendly NPC will be more forgiving of hostile actions. A paranoid NPC will act unpredictably and without provocation, etc. By combining traits, I can use one general AI subroutine with parameters.

* All NPCs can belong to multiple factions. These are generally not things as broad as a race, but rather things like "member of town x", or "member of organization y". The purpose of factions is obvious enough.

* Here's where it gets a bit complicated. All NPCs have a set of "Attitudes" that are updated every turn, and every action, for every other NPC (and the player.) For example, NPC X (a shopkeeper) may have a generally positive outlook on the player, but a negative outlook on a thief mulling around outside. But this is a very simple way of describing it.

Attitudes are actually comprised of several 'scales'. Right now, the two scales I'm using are "Like" and "Fear". The "Like" scale simply represents whether the NPC likes (X) or not, and how strongly that feeling is. The "Fear" scale starts from a 0 position and determines how much fear the NPC has for (X).

* Attitudes are influenced or modified in several ways. One is by observation. Every time any entity takes an action, all other nearby entities will observe it. Thus, if the player attacks NPC X, who is of the same faction as NPC Y, NPC Y will dislike the player. If the player kills a creature in one blow and NPC X sees, NPC X may become afraid of the player. These observations are then tempered through personalities. Some NPCs may be able to tolerate a great amount of dislike for others before acting on it, while others could behave irrationally. Monsters/beasts act in more primal ways, generally responding only to fear (they won't care that you donated money to an orphan.)

* To complicate things, every entity maintains a record (memory) of anyone who has entered into combat with them, or vice versa. They keep track of things like # of friendly vs. hostile actions, total damage done, total turns in combat and turns since last hostile action. Why is this necessary in addition to the attitude system? Because if Combatant A has been running away for the last 15 turns and has only performed one hostile action, they are clearly a lower priority than Combatant B, who has attacked them every turn. In other words, maintaining this combat data allows for more refinement in combat situations.

* Every turn, every entity considers every other entity and checks what its current goal is. If a current goal exists, the AI will need significant motivation to change it (ie. a pissed-off guard will not stop chasing you after 1 turn.) Here's where my question comes in.

How should I approach AI decision-making when it comes to combat priority? I'm tracking basically all the data, but I'm not sure how to let the AI process it. It's easy enough to determine which entity has dealt the most damage to the AI, as well as which entity has been in combat the longest, or has helped it the most, or is the most injured. But how can I compare Combatant A, who has been wailing on the AI, to Entity X, who is a faction enemy (but non-combatant), to Combatant B, who is critically-wounded?

How would you approach this?

[Play it for free here!]

Play it for free here! (Firefox recommended): http://www.spacewhalestudios.com/duckit

Duck It! was our (Space Whale Studios') entry into the 2011 Philly GameJam (hosted by TooManyGames and our local IGDA chapter.) Being that it was around Mother's Day, the theme was "Mothers." Obviously, we decided to make a game where you're a mother duck running through a creepy black-and-white world and the goal is to make it to your young and vomit pixels on them. It's inspired by games like Canabalt with some action and physics elements thrown in for good measure, not to mention everything is pseudorandomly generated for infinite ducking action! (Complete with "duckstep" intro music and KMFDM ripoff level music)

Trailer:



The game was developed using HTML5 and the Impact engine in 45 hours total. Our team consisted of two artists, two programmers and one musician/sound designer (me!) We managed to win the WTF? award for our efforts. Proof:



We'd like to keep developing the game further, so feel free to give us your feedback. Enjoy!

Thanks for the kind words. I've played some games from both series you've mentioned but not the early M&M games. Any one in particular you would recommend? I'll happily play it.

Still working on AI. Right now I'm working on a way of enabling NPCs to determine priorities in combat. It's one thing to select the most injured ally from a list, or the most threatening enemy. It's another to reconcile the two and determine which is more important (especially when factored in with the aforementioned attitude system.) But it's coming along!

Not only that, I'm now working with a very talented writer who is helping to flesh out the backstory of the world and various other cool things. Should be a fruitful collaboration indeed. The goal within the next few months is to have a very fleshed-out vertical slice of gameplay: a bustling town and an outdoor area to be explored, plus some quests. So, you'll get lots of NPC interaction, explore sets of skills, and fight stuff (if you want.) That should make for a good prototype.

Lots of cool stuff + algorithms.

1. Free targeted skills are in, and they can have rotatable shapes.

2. When using a skill, you get a nice transparent display showing all possible spots you can use the skill.

3. Healing, damaging, and status-affecting skills are all done. Skills can have multiple effects.

4. Basic monster AI is in, and I'm refining it constantly. The AI observes everything around it - every action by the player (that it can see) along with actions of other AIs!) Not only that, but AIs can have goals beyond their current action. If you attack their buddies, they get mad. If you kick people while they're down, they'll hate you (or if they're cruel, maybe they'll respect you.) Cautious NPCs are more inclined to flee, etc. LOTS of work and expansion to do, again, but I think it's pretty exciting and should pave the way for deep AI in this game.

5. Basic A*-based pathfinding. This took me most of the day. An AI that wants to kill you now knows how to maneuver through and around obstacles (of all sorts) to get to you, and will stop when they're in range to attack. I'm very proud of this since I didn't use an existing implementation but wrote one from scratch instead.

6. The gorgeous 16-bit style background art is looking fantastic. Will post more screenshots when the beautiful stuff isn't sullied by my awful MSPaint drawings...

7. Characters can now be displayed as larger than 32x32 (right now, only in the vertical direction). This is great so that we can have classic JRPG-style sprites where the characters are bigger than tiles.

Nothing wrong with puzzle games if they fit the core audience. Granted, most puzzle games don't, but some do.

Does this robots game have a demo?

Sure - we're going to upload it publicly soon (just wanted to get a new YouTube video up and send out some emails to blogs again.) I can send it to you in the meantime, just shoot me a PM with your email.

So... no puzzle games? Would this game qualify:

www.returnallrobots.com

In a world of samey lo-fi 8bit and faux-8bit games, I have to say this is among the most unique art styles I've seen. Brought back instant nostalgia of old-school PC adventure games like King's Quest and Gold Rush. Bravo.