So how does the full version of C3 compare to C2? Is it the same or better? Should I wait to subscribe?

C3 is an amazing tour-de-force of browser programming. I have used it to sketch out a few ideas, and it works very similarly to C2. Everything is super-fast, and it runs on my mac.

I don't like the way saving and export works -- I am a dropbox subscriber, and not being able to export directly to the public Dropbox folder I use for testing is a pain. The current export option requires export -> save .zip to hard drive -> move to dropbox -> unzip -> test on server. Under C2, this all occurred in a single step.

Although C3 has the potential to be better than C2 in the long run, I can't use C3 for either of my current projects, which both rely on a plugin that will probably never be ported to C3. I am really struggling with whether or not I should subscribe, even though I know I will be using C2 almost exclusively over the next 12 months.

Here is a short video showing gameplay.

In its current iteration, the game has become a "strategy first person shooter". You can switch between the characters on your team, but they will act according to their autopilot. Right now, the goal is to be the last team standing. It is very difficult, as you are going up against three other teams.

Each team consists of:

A laser tank - long range firepower

A creep - melee only, but fast and can jump on top of obstacles

An electric tank - medium range, but continuously hits and does damage

You are playing the role of the red team.

Keyboard Controls:

<arrow keys> forward/backward/turn left/turn right

<space bar> reload layout

<enter> switch to a different teammate (you might have to press it a few times)

Touchscreen Controls:

<no buttons> drive forward

<left button> drive forward and turn left

<right button> drive forward and turn right

<left button + right button> reverse

<recycle button> switch to a different teammate (you might have to press it a few times)

Here is a playable demo of the game in its current form. Note that the 2P and 4P selection screen at the beginning is just a placeholder. Tapping on either option will start the game.

https://dl.dropboxusercontent.com/u/55106174/rrfpsbattlearenatest/index.html

Why does everyone think construct is worth $100 a year? Sole credential being a game the creator had to abandon due to an awful runtime and 0 chance of platform ports Scirra promised were built-in (Last Penelope).

There have been other threads to cover this. To many it is a creative tool that people enjoy using. If you are an artist, there is a good chance that you spend more than $100/year on art supplies. If you play sports, there is a good chance that you spend WAY more than $100/year to join a club or attend games. If you love watching movies, chances are you spend more than $100/year watching Netflix, Hulu, and going to theaters. If you love to eat out, chances are you spend more than $100/year at restaurants.

For me, I am willing to support Scirra partly because I love to tinker in Construct, and partially because it is financially self-sustaining. I am very lucky in this regard, as I have been able to sell my creations and (partially) offset the cost of the time I spend. Most of my colleagues don't make any income from their hobbies -- I live in Chicago, and there are several members of my department who have spent several thousand dollars for tickets to watch the Chicago Cubs play baseball a few times a year. Watching the Cubs play is not how I would choose to spend money, but they get such great joy from the experience that it I understand why they would do it.

If I used Construct 2 purely for financial gain and not because I enjoyed it, I might stop using it and move on to other activities. I tried that approach, but keep coming back to Construct 2 because I would rather spend thousands of hours doing something I enjoy rather than something that is going to make me 50% more money for the time I spend.

>

>

> Drat. Now you have me worried, as I am getting close to starting the level design process with Q3D. Using block art and a top-down level layout I hadn't anticipated too much of an issue.

>

> What specifically were you seeing that was giving you trouble? Was it doing things like trying to match up Q3D models with their shadows?

>

> Also, I concur that the 640x360 looks a little better.

>

It's just that i have sooooo many objects of different kinds in different sizes from small pebbles, mushrooms, twigs, flower, to large houses, and trees. Placing all that on the map without any accurate representation of what it would look like until I press preview was my major turn off. If i had the patience I could probably pull it off. If I had a lot less assets and a flat level (instead of height variations as i was doing) it would be a lot easier. If you're game is pretty simple, and most assets are on the same plane, and similar size i guess it's okay. I guess my game is just too complex in terms of the details

That's my only issue... visual representation in the editor.

I just didn't have the patience to place something, preview, tweak position and scale, and test again for thousands of objects..

Got it. I wonder if a lot of the really small stuff could be taken care of algorithmically?

FWIW, now that we aren't sure what is going on with X3M, I've been looking into what it would take to do a C3 port of babylon.js. It is a little overwhelming, but I think it might be easier to create the necessary plugins for babylon.js rather than three.js. I'm definitely not the ideal person to start the process of creating a set of 3D plugins for C3, but if no one else volunteers I might consider taking on the project.

I think it would be wise to wait until we have z-ordering capability in the editor though.

May 15 update.

Playing around with Q3D, everything seems to be working fine, but been struggling with level editing. I've decided to go back to 2D isometric, as any type of level editing with lots of small props and items placed all over the map at various height and sizes is too tedious to work with, since C2 doesn't support any 3D viewport. All of the benefits of Q3D, doesn't add upp when it comes to editing levels. As I can't get a clear view of what I'm doing, as the visual representation in the editor is lacking.

Hopefully at some point C3 will open up possibilities for plugin developers to make a proper viewport..... so ...

Drat. Now you have me worried, as I am getting close to starting the level design process with Q3D. Using block art and a top-down level layout I hadn't anticipated too much of an issue.

What specifically were you seeing that was giving you trouble? Was it doing things like trying to match up Q3D models with their shadows?

Also, I concur that the 640x360 looks a little better.

[quote:q2qd00am]Can please Tom clarify what happened there?

Yes, look at his profile he was hacked. The hacker who hijacked his account also attempted to hack our store which caused problems for our sellers (now resolved). The store hack damage was limited to items being taken off sale for a period of 1 or 2 days. No other damage was done as other safeguards worked fine.

Thanks for letting us know what happened with — I was planning to use the as-yet-unreleased version 1.0 of B3D for my project, but it looks like I will need to fall back to Q3D until development on B3D resumes.

I still have my fingers crossed for someone to port one of the 3D plugins to C3. In their current forms, Q3D and and B3D add a lot of functionality to C2, but losing X3M is a big blow to my hopes of a C3 port.

What are you practicing towards by making lots of games if you are not aiming towards an audience?

Currently it seems like 'idle' is the main draw card, along with killing stuff. I have a game with 'slayer' in the title on kongregate which is an idle game with 5000 plays, and an idle game about growing lettuce with 3000 plays.

My newer small games I'm trying out on Scirra arcade, which is a smaller userbase I grant, but I thought I'd see a touch more traction. Then again 'Khloro Slaying' on the Scirra arcade is mine and got 62 players, 85 plays - the range I'm expecting (slightly better, actually).

I guess having 'you're killin' stuff' in the title probably is the main draw card. I've been trying to avoid the topic of violence and killing.

I guess I should have figured that already. Maybe I was trying to avoid acknowledging it.

I figure I have spent at least 300 hours on C2 projects whose only real value was in developing my skills. Over the course of the past three years I have accumulated 153 exported C2 projects, 50 of which are now published on my website. As a teacher, my target audience is my own students first, then other students around the world. I typically come up with an idea a few days before we need it in class, get something working within a few hours, then spend many more hours polishing it so that it is good enough to publish. Out of the 50 simulations, only about 10 see the lions share of users.

Looking back, every simulation I published was a passion project that I felt needed to be built, but I started a lot of projects before I first stumbled on one that was truly novel. Once I realized that I had something great, I polished it to the limits of my ability and published to the web, iTunes, and Google Play, where it has been pretty successful on all three platforms.

I suppose that I haven't really addressed your original question. I don't think that attracting a significant number of users has much to do with the theme of the game (zombies vs idle game vs puzzle game), aside from the fact that a game should match the player's needs on the website where it is published. Instead, I think attracting players has a lot more to do with how polished your game is, and how effectively you let the world know about your awesome game. If "kill-kill-kill" doesn't interest you, maybe you could try a different website to show off your creative work? Perhaps one that caters to players of the games that you want to build?

While it is true that a master might be able to make a beautiful game quickly, it is highly likely that the master spent thousands of hours honing their craft to get to the point where they can do it quickly.

Thats the big advantage of doing lots and lots of small games for practice. While you are gaining skill, don't expect any of the hundreds of projects you will create along the way to be popular or successful. There will be a few gems, but most will not be worth publishing.

Characters

The laser tank and creep models and textures were created by Tam Vu for Robot Rumble. They were rendered in Blender with Lambert/Phong shaders. We are planning to use Q3D's "basic" shader in-game (no shadows):

Laser Tank

Level 1 Creep

As a point of comparison, I recently started building our 3D game in Construct 2 using the Q3D plugin just to see how it would compare to building the same game in Unity.

Originally, the game was set to match the resolution of the device. On a 2017 Surface Pro 4 with a hi-DPI screen, this brought frame rates down to below 30 fps, almost unplayable. I have found that by keeping the screen resolution lower (854x480) and disabling all effects, performance is a rock-solid 60 fps/50% CPU load on both a low-end 2013 MacBook Air and the Surface Pro 4, and is 58-60 fps on an iPhone 7 plus.

The game is physics-intensive. I built several prototypes of the game, first with a custom 2D physics engine (built using C2 behaviors) and 2D graphics, then using the standard C2 physics engine with Q3D graphics, and finally with the Q3D Oimo.js 3D physics engine and Q3D graphics. In the end, I found that the OIMO physics engine gave me the best game feel, and by limiting the number of physics bodies I could keep CPU load down. The standard C2 physics engine was a smooth 60 fps as well, but the game really needed the z-ordering and 3D torque of a full 3D engine.

Now that I am a few months into development, I am planning to stick with C2 rather than switching to Unity. I haven't found any technical roadblocks to completion, and I expect the finished game to fall into the "medium-sized" category. I have been a heavy C2 user for the past three years, so I am very careful with the way I structure my event sheets for maintainability. There have been several times that I wished that C2 had true object inheritance, but most of the time I am happy that to work with event sheets rather than C#.

Good points by tunepunk here. If I am looking through a list of games, the first thing that catches my eye is the art. If I am looking through the Scirra arcade, for example, I am more likely to click on artwork that incites an emotional response.

Artwork that incites this response typically takes effort. Concept sketches can be quick, but polishing and refining might take many hours or days for each character.

I wonder if the better approach to see what attracts an audience might be a beautifully crafted character in a convincing pose. Once you have people's attention, then work on the game itself.

Robot Rumble 2.0 - a first-person strategy shooter

Command a team of robots to destroy your opponent's base, but make sure to defend your own!

-Each player controls a team of robots.

-Switch between your robots mid-battle to gain the upper hand!

Playable Demo:

https://dl.dropboxusercontent.com/u/55106174/rrfpsbattlearenatest/index.html

Robot Rumble 2.0 will feature:

-4 new robot types - laser tank, rocket tank, shotgun tank, electric shock tank

-Bug-like creeps

-Amazing professional artwork / UI / level design courtesy of Tam Vu http://www.vutam.weebly.com

Robot Rumble 2.0 (RR2) is a ground-up re-imagining of the successful AirConsole 3D Robot Fighting game "Robot Rumble" for AirConsole: https://www.airconsole.com/#!play=com.nerdislandstudios.robotrumble

Technical Details

-Both the original Robot Rumble and Robot Rumble 2.0 are currently built using the excellent Q3D plugin.

-RR2 will be designed first for local multiplayer on AirConsole, with additional content and levels added later to support a launch on iOS / Android.

-RR2 is intended to be both a fun game and a demonstration of what is possible in 3D using Construct 2 / WebGL.

I have asked myself many times why I am rebuilding Robot Rumble in Construct 2 instead of Unity 3D, and the answer is simply because I want to explore the limits of what can be done using Construct 2.

Current Status

Camera: After about a month of experimentation, we have settled on a 3rd-person over-the-shoulder camera view.

Controls: The over-the-shoulder camera view makes mobile controls simpler. With automatic firing, the player only needs left/right buttons. After a lot of experimentation, we found that playing this way seems the most natural.

Multiplayer: For AirConsole local multiplayer, the over-the-shoulder camera view means that RR2 will be limited to 1-4 players with a "Goldeneye"-style 4-way split screen. By keeping the number of players down, we can amp up the physics and effects. Right now, all robots are using physics-based wheel driving, so that they roll and tumble naturally on top of each other.

Screenshots:

iPhone screenshot showing the results of a successful battle -- the player won using a laser tank. The "ghosts" in the screenshot were all killed: