Straight question: Is Construct suitable for a big project?

Hi guys,

First of all, I have to say that I am with love with the program. It is really easy to get and understand.

We are about to start a big commercial project and we are looking for some sort of 2D engine. We think that Construct could fulfill development needs at 99% but things I've reading about bugs, memory leaks, crashes and such makes me think otherwise.

We have not encountered any issues, but would be deadly if they appear later on, that is, on mid/final stages of development.

Questions are then if those problems were solved and if Construct Classic could be safetly used. Of course we will donate to support the software and we think that Construct 2 would come handy, but is not finished and we wish to start now.

Any honest reply would be highly appreciated. Many thanks in advance,

Well, you'd need quite a bit of Construct know-how to pull a stable large project.

Depends on the type of game you're making probably.

Well, you'd need quite a bit of Construct know-how to pull a stable large project.

This a thousand times. We're on our fourth rewrite of our most ambitious project so far and we've narrowed the scope every iteration but finally we're starting to get the results we want. Some plugins can cause crippling flaws that can end the development in an instant but as long as you separate your test builds from your actual project builds you're on a safe territory.

Experiment experiment experiment and make separate test builds, figure out a good development cycle and you'll be able to get the most out of Construct. You'll quickly notice that working with a team might become complicated because of the nature of the binary project files so I encourage you to get Dropbox or a proper piece of project management software to dodge that issue.

Edit: fixed a minor typo

It's possible, but it won't be easy.

Personally I'm working on smaller stuff for now, with my big project on hold until C2 is usable.

Construct Classic has its flaws. Construct 2 is being designed to fix them - not just bugs, but things like collaboration. Working on Construct 2 projects in teams should be much, much easier - Classic uses a big single binary file which is impossible to merge, but C2 uses project folders with XML files that can be merged, versioned on SVN and so on.

I don't want to annoy the Classic developers but I personally would recommend waiting for C2 to come to some level of maturity before starting a large project. Alternatively, scale down your project for Classic - you might run in to difficulties with a large project.

I'm working on what's probably one of the biggest construct projects out there - a sort of 'mini' RPG with almost 500 objects, thousands of animation frames and 10,000 events, and it is possible. The main problem I'm encountering is working with animations at this point - for some reason I can't track down the IDE is sometimes crashing after working with them for a bit. All I can tell is it has something to do with either how many objects I have, how many animations/animation frames I have, or both.

Even then, I wouldn't call my game a big project - more of a medium sized one. I suppose it depends what you define as a big project, though, what did you have in mind?

Regardless, I echo what's been said here and recommend waiting for C2 to start a large project if you can. It can be frustrating at times working around Construct Classic's flaws.

> Well, you'd need quite a bit of Construct know-how to pull a stable large project.

>

This a thousand times. We're on our fourth rewrite of our most ambitious project so far and we've narrowed the scope every iteration but finally we're starting to get the results we want. Some plugins can cause crippling flaws that can end the development in an instant but as long as you separate your test builds from your actual project builds you're on a safe territory.

Thanks mate. It's always good to know that some issues are plug-in related and can be isolated.

[quote:1akj67u0]Experiment experiment experiment and make separate test builds, figure out a good development cycle and you'll be able to get the most out of Construct. You'll quickly notice that working with a team might become complicated because of the nature of the binary project files so I encourage you to get Dropbox or a proper piece of project management software to dodge that issue.

Oh, perhaps I forgot to say that I do have a extended background on games. I worked for many companies and as a team we worked on many projects already on different programming languages. Thanks for your spot on tips.

I'm working on what's probably one of the biggest construct projects out there - a sort of 'mini' RPG with almost 500 objects, thousands of animation frames and 10,000 events, and it is possible. The main problem I'm encountering is working with animations at this point - for some reason I can't track down the IDE is sometimes crashing after working with them for a bit. All I can tell is it has something to do with either how many objects I have, how many animations/animation frames I have, or both.

Even then, I wouldn't call my game a big project - more of a medium sized one. I suppose it depends what you define as a big project, though, what did you have in mind?

Regardless, I echo what's been said here and recommend waiting for C2 to start a large project if you can. It can be frustrating at times working around Construct Classic's flaws.

Thanks for sharing you experience. My idea for a game is a 2D, 2.5D one, platformer. I can't share ideas for known reasons but it would be a Braid-like and would have some complex features that would need many objects and complex programming. But would be 'big' in terms of development time, features, people involved, etc.

The problem is not the complex programming, but knowing if for Construct related issues, an obscure bug may appear that could cause us a drawback.

Perhaps a little scared, but your experiences bring some relief. Thanks everyone for the comments. More experiences would be highly apreciated.

I'd say it's definitely possible to create large games in Construct. However, I'd suggest spending good time on a smaller project first to figure out what works and what doesn't. I used to get a lot of crashes and find bugs and stuff, but over time I've learned how to avoid them. Construct hasn't crashed for me once on my latest projects

I think the biggest issues out there now are resource management with hwa, knowing the workarounds for "features", and the less than obvious issues DirectX 9 brings.

A platformer, shouldn't be a big deal, but 2.5d on the other hand....

Well that's more of a complexity issue rather than ability. There is actually quite a lot of built in functionality for that, parallax layers, z height, z index, etc.

I would suggest figuring out what would be the hardest thing to implement, and testing that out first. That should give you a hint as to whether Construct is for you.

Then the forum is always pretty helpful for specific stuff that might evade you.

Honestly, if you're going to do anything 3D, even if its 2.5D, you should probably use something else. Construct's 3D is quite limited. It doesn't even have basic lighting. While technically you could make your own 3D engine with events, it's overly complex and wouldn't run as fast as other options like unity would.

You seem to sound a bit comforted by what we said, and maybe in particular from what I said. Maybe a bit too comforted. I don't mean to leave you with the impression that what I mentioned is the only issue I face with my game - it's simply the only one I haven't managed to find some way to work around. Some of them are easy to avoid, like the or condition which causes crashes, or the layout object which among other things doesn't register left clicks. Some of them are a lot weirder, like I have an old event that crashes the IDE if I try to move it in the event sheet (though I can deactivate it) and a layout I can't use without my game going haywire - easy enough to work around by making another layout and not using the bugged one, but seriously distressing until I figured out how to work around it. One of my event sheets with 5000 events takes 10 minutes to open due to some weirdness (though I have an old processor, which certainly contributes, but it shouldn't take that long regardless as a newer event sheet with 1000 events only takes a few seconds to load).

Then again, I also started this project 2 1/2 years ago, so a lot of the bugs that I face have already been fixed.

I feel somewhat nervous about what I said before actually, because it seems it's kind of hard to predict what quirks might arise for a user, and I really don't want anyone to risk financial ruin based upon my experience alone, because there is one user I know who has encountered a showstopper, though he's the only person I know of who has encountered that particular problem. Because of that, and how unpredictable construct's quirks can be, I don't feel comfortable recommending anyone risk their financial future on the chance they might encounter something they can't work around. Perhaps I'm being overly cautious, but I really, REALLY don't want you to encounter something that forces you to scrap your project and lose a lot of money, especially from my advice.

What I'm trying to say is, can you? Most likely. Will you encounter quirks that you can work around? Almost undoubtedly. Will you encounter a showstopper? Probably not.

But risking a business on a probably not probably isn't the best idea.

Again, I feel much more comfortable recommending you wait for construct 2 instead. It's far, far more stable than construct classic was at this point in its development, therefore there's much, much less risk of quirky weirdness, and it'll also help you in the long run for porting to different platforms. It's a much better business decision.

The problem with waiting for Construct 2 is that you could be waiting years before it gets all the functionality you need from Construct classic.

At the rate they're going I would guess that it will take considerably less time.

Think of CC as just the prototype . Ash and Davo have already done this once, they have a much clearer idea how to do things now.

Depending on your Computer, another thing that gets annoying are the long loading times in big projects. It can sometime take up to 2 Minutes untill I can preview anything. From what I know, once you previewed your game once, the next time It will go alot faster until you change a sprite, then the time needed to preview the project is resetted. (I'm no Construct Wizard, maybe I am doing something wrong, but I witnissed this more than once now)

That being said, I personally think that once you know what you're doing in Construct you are able to make a big project. The mentioned "Showstoppers" occured here aswell (mostly because of my GTX 420) but I always found a workaround and could continue. One just needs to know how live with those quirks. Don't try to work against them, but work with them.

So in short, I think it's stable enough for a big sized project. You just need to prepare for some additional work because of long loading times and possible quirks.

Yeah, a way to reduce the preview times is to load large static images like backgrounds at runtime via the 'load frame from file' sprite action instead. That helps a lot, it reduced my compiling time by like 2/5ths. You're correct, it does cache the graphics, so the second preview is faster until you edit an image again.