Large scale RPG in Construct 2

[rosareven]

Imagine a RPG made with RPG Maker [select your version here], with a world full of details, an indepth story, dialogues and interactions with NPC, and may be 48 hours worth of main story play time. Many people have made that with RPG Makers on HBGames so those kind of games are what I'm referring to.

Now imagine that very same game being ported to Construct 2, which will be played over the Internet on a web browser.

How would you go about making it happen in Construct 2? All of a sudden a game that's been traditionally played from an executable package in your computer, is now played through the internet cables. Do you ask the players to download the whole HTML package to play locally? Develop server scripts to interact with the HTML5 game so that you only load resources when it's needed and allow saving over the internet? Wait for the whole game to load in the browser?

 

[coyotecraft]

The whole point of html5 is to be played online. But it would be annoying if it had to load the entire game world every time you wanted to play it.
Maybe if you separated areas into different html pages and had Player information cached so it could carry over. You exit an area and it loads a new html page for the next area. Or selectively load resources by layout for shorter load times.
Maybe I'm confused and the game already loads content on-demand? I haven't seen a C2 game stop and load, so i'm assuming it loads everything from the start except for music that it streams.

 

[BlueScope]

In general, you need to choose the platform for the game you want to do. If you have an RPG that is single-player, has no online interaction at all and in general has high-quality graphics and whatnot, I don't see a reason to use C2 instead of RPG Maker. Despite my very own doubts regarding the Makers, if you want to make an RPG with a framework, that's the way to go.

In general, after working with it on Power Shift (that's halfway out of C2's comfort zone, I'd say), I'd say C2 isn't made for complex setups or object/scene links, and should work better with small things than with big ones.

Now, if you have a small-scale RPG with simple mechanics that uses a lot of online connections, and you feel pestered by the restrictions of RMVXPA (how I like those abbreviations... you have no idea :D ), I guess you'd be better off with C2 than with a Maker. Still, none of them is really made for something like that, so going another approach and maybe use plain HTML5 or even Flash crazy might turn out better in the long run.

Last but not least: If you actually start developing it with C2, your game will automatically adjust to C2's internal structure. You will end up with a game that's working well with C2's mechanics, and it might even be something as complex as an RPG. You might prove me entirely wrong and make the best game there ever was. There's no way to tell until you actually give it a shot

 

[Venetia]

I agree with BlueScope ... Instead of reinventing the wheel, it will be easier to just use the right tool for the job you want to get done.

It seems to me that an RPG that is not ridiculously complicated in stats or overburdened with hi-res graphics would be totally possible.

There are a lot of useful links in this Scirra forum topic about making & using tile-based maps in C2: http://www.scirra.com/forum/array-tile- ... 49692.html

Since your resources are not self-contained in an exe, but have to load in a browser, you have to be especially careful about how you use images.

However, when comparing C2 and RMXP, for instance, you would be able to pull off much more interesting physics dynamics with far less effort in C2, than you would have to exert into perfecting RGSS scripts in RM. So turn-based RPGs wouldn't be worth it to attempt, but more action-oriented ones, like Zelda, would be more up your alley.


One interesting thing is that some people are also using C2 for making Adventure (point & click) games.

So, say, if you wanted to make a game with a rich story, that uses interesting physics or dynamics for puzzles, you can potentially pull that off much more nicely in C2, than you would in RM*.

All in all, C2 is better suited towards action-oriented dynamics, so that's the engine you'd want to use for that.
If you wanted to make a typical tile-based & turn-based RPG with no focus on physics/action-oriented gameplay, you'd probably be better off with RM* or something similar.

 

[Jason]

I think you COULD make a large scale RPG, but you'd have to cut down on all the flashy effects and stuff to make it small enough that people won't be waiting an eternity for it to load... unless you follow whay coyotecroft said and break it up into multiple games, and have a way to transfer the player data and stats through your cache... someone should try it and see if you CAN transfer data through cache, because if so, that'd be brilliant!

 

[Venetia]

Maybe I'm confused and the game already loads content on-demand? I haven't seen a C2 game stop and load, so i'm assuming it loads everything from the start except for music that it streams.

They're still working on adding layout-by-layout-loading functionality. It will probably be in a future update.

Sounds are streamed dynamically, so the biggest concern in pre-loading would be in the graphics department. If everything is 8-bit or 16-bit then there shouldn't be any problem, however for larger-sized graphics, a developer would want to package their game as downloadable instead of being played in-browser.
Graphics-loading times will be mostly negligible if loading from a hard drive. Functionality for adding update support is already in the engine.

They're also working on making it possible to package your project in an EXE. I think this may be a couple of months off, though.

 

[Jason]

Ooh, packaging it as an exe so we can open it from the desktop and play it locally as opposed to the net? Does this mean it wouldn't need us to preload or stream the music since it'll be included? I'd LOVE that!

 

[BlueScope]

Seriously, using C2 to make a game and then compile it to an executeable file seems rather odd. Not to mention you can already do that by exporting it as HTML5 and just run the .html file from your local host - will give you instant loading time as well, as well as compatibility on other than Windows OS's

C2 is made for games played from the browser online, and that's what it should be used for. While HTML5 isn't unsuitable for desktop applications in general, C2 certainly isn't perfected into that direction and you'll likely find using Game Maker or anything similar easier to use or at least more performant or more capable in the end (namely, adressable renderers, ability to read from the local filesystem, ...)

 

[Jason]

Yeah but I mean so it opens up in its own window sort of like RPG Maker games lol. I know it's made mainly for HTML5, but there's no harm with them branching out and giving us more options, is there now? :wink:

 

[BlueScope]

That depends... it's the old Photoshop problem: Right now, it's a very useable, easy to grasp and easy to use program for beginners and professionals alike. The more features it will get, the more complicated to use it will be, and even though the basic useability isn't vastly improved (with the exception of some cases), the program gets way more complex. In the end, you have a swiss pocket knife where you take a minute to find the tool you're looking for, while at the current state, you'd pick the knife that you know is on the upper side and do whatever you want with it.
That's always a problem for software developers and interface designers like me as well: You want to pack 1000 functions in a program, all intuitively operateable and immediately understandable. The only problem: It's not possible.

But that's maybe a bit too in-general... another export function might not hurt, but it will likely not be done with that.

 

[Venetia]

While it's true that d/l'ing the HTML file and playing it locally would work fine, I think devs were asking for EXE exports as well so they can market a game that isn't inside a browser, but rather is entirely self-contained. Also it would probably add another layer of protection as far as images & sounds or source scripts are concerned. Less easy for people to decompile and mess with.

In the end, though it's not totally necessary: It couldn't hurt.

I understand perfectly well what you mean about the Photoshop problem. C2 is a far more approachable engine than GM, any day of the week (perhaps mostly, or even entirely due to the fact that there are less features). Shooting from the hip on this analogy, but C2 is to GM as GIMP is to Photoshop. Each have slightly different strengths; Photoshop has more features, but GIMP is easier to just pick up and use.

Plus, this is entirely sentimental, but YoYo Games is a fairly large, well-funded software developer (not the biggest, but definitely not 'indie'). C2's software dev team is about as indie as it gets: Two game-loving guys in London. HBGames is an indie-oriented site, so it seems like a no-brainer that we would want to back a fellow indie.

(I realize that GM's beginnings were independent, but it's been under YoYo for ~5 years now.)

 

[Venetia]

oops

 

[rosareven]

Wow thanks for the discussions guys, this thread just sort of imploded with all things interesting.

Indeed action-style RPG was what I had in mind (sorry I'm a bigger fan for action RPG and ignored the FF style turn-based department), and Construct 2 really works out of the box for the action combat department. So very dead easy to implement. I was mainly concerned with large image content if a large world is involved. As mentioned by coyotecraft and Jason, perhaps right now we'll just have to cut it up into chunks of different games, may be episodic.

 

[BlueScope]

@reven: I think that sounds very artificial. Why make a game you know of that it'll make problems on C2 with that framework, and try to find a workaround for that?
Think about the player here: You're providing an experience, and the player will be able to dive into it when he gets a single game, not little chunks full of redundancy.

If anything, the right way to go here is work with a C2 function or plugin that allows specific preloading of images. That way, you could determine the time of load, and maybe directly set in the project properties that you want to load images for each layout individually instead of globally at the beginning. Saving data to cookies is something someone definately made a plugin for yet, so that would work as well.

@Ven: I agree... still, there's a reason why 50 developers use 45 different engines: Eddiciency. And there is no efficiency in sight with a html5 game compiled as an executable file. Definately a nice feature for the requesters, but just because a lot of people want something doesn't mean it's the smartest thing to do.
And I'm definately with you as far as supporting the developers go. That can't mean that we all should do game design exclusively in C2 now, no matter the concept §^

 

[rosareven]

Ah I didn't know there's a plugin for cookie data saving. Will have to look into that.

 

[BlueScope]

I'm not sure there is, I'm just massively convinced

 

[rosareven]

I got a topic about the possibility of on-demand image loading in Scirra forum. A member said the plugin "Webstorage" is really only for text storage, and there isn't one he knew of that's for images. Although he believes that PNG 8-bit compression with image original size being the same as their display size in game (aka if an image is going to be shown as 32x32 in the game but the actual image is 128x128, shrink the actual image to 32x32) will really speed up the loading without major loss in quality and retains transparency.

At this rate the only option is to get hands dirty and try it out!

 

[Brewmeister]

A tidbit I sumbled on... http://www.rpgcreator.net/

 

[rosareven]

Wow nice find Brewmeister. Coming up in June, no wonder it never came up in searches before when I researched extensively for game makers.

 

[Brewmeister]

Another stumble-upon. (allegedly government funded development. (not verified)) http://tiltfactor.org/layoff/play.html