My 3¢ about c2 and c3

Discussion and feedback on Construct 2

Post » Sun Sep 27, 2015 11:47 pm

Hi,

I've never really had a vibe to learn programming (even thought I did in the end learn some js/jquery and php because of work ), so the premise of none-programming game making software was always appealing to me. First there was an obscure gamemaker on atari st, then click and play on pc, followed by rpg maker and then games factory. I started when I was 7 years old. Then my friend showed me construct - back then alpha 0.18 with that pseudo top-down gta demo - and I've been learning and using it, and observing its development ever since. It always had and still has marvelous potential.

However I was never fully satisfied with it and never felt like it is delivering on its premise. I always waited for improvements hoping that certain things that I couldn't pull off now, will be made available at some point, and that technical issues will be resolver. The latter I'll leave for some other time.

At this moment, I'm certainly finding construct limiting, both creatively and productively. There always has been some obstacle getting in the way of realizing the vision. And then there is always a talk of walk around's to compensate of lack of many standards. Stuff like units pushing out of object, to be used in games like rts or hack and slash etc. Or input delays to simulate different action rates. Or even such simple thing like scrolling text and default menu plugins, resolution change etc. Many ui and level building improvements. Those need to be within the vanilla software from the start. Thought many things can be done trough events, they need incredible attention, most often feeling more like hacks and rarely ending up working precisely, tightly in action, making often game feel and look awkward. In the event style environment, behaviors, plugins be doing most of work, streamlined but with many options, ready to take on many different ideas, and taking care of mundane, but giving a lot of possibilities. Same goes for UI and level tools. This will free up time needed for development in other, more important areas of gameplay; areas that will make or break the success of the game. I'm talking about original ideas and fine-tuning, testing and experimenting.

So now c3 is on its way and I hope that this time around it will be made truly with the developer in mind.

Thanks for reading
My professional Royalty Free Music at Scirra Assets Store
--------------------------------
Specs: i5 2500, 16gb of ram, gtx 770, win 7, Focusrite Scarlett 8i6, Mackie mr8mk2, Alesis 320, browsing the net on chrome.
B
89
S
30
G
22
Posts: 1,985
Reputation: 20,099

Post » Mon Sep 28, 2015 12:40 am

From my point of view I think it's because C2 was designed and made long time ago (well, not really but a lot time have passed since) And for that time all was great! But times have changed, standards have changed and C2 popularity sky rocket quite high with a lot of new users. And I truly believe that C3 will catch up with all of that.
Can't find any other reason to justify things like
- not being able to manipulate group of objects at runtime without making long and sometimes complex events.
- doubling up every single object in the layout just to add some nice bump mapping effects - which again makes everything harder to work on
- not being able to use more than one lightsource on above effect !
- not being able to use more than one shadow caster on the layout without getting stupid results.
and so on...

Also i think C3 should go full on WebGL. Sorry to say this, but forget about canvas2d, there is really no point to have it. It makes everything more complicated than it should be.
ImageImageImageImage
B
157
S
66
G
41
Posts: 2,598
Reputation: 34,823

Post » Mon Sep 28, 2015 1:11 am

I don't mind work arounds or hacks to get things to work, but I do mind some major technical limitations.

I'd be happy with some memory controls, load sprites into vram, unload from vram, giving the developer more control (optional) over this is required for very large games to run well on weaker hardware.

The biggest limitation I see currently is single-thread logic. JS/HTML5 has higher CPU overhead and we're limited to 1 thread when modern CPUs have 4-8 is quite backwards. Yes, multi-thread programming is very difficult. But other game engines have MT functional, even Unity these days are moving to DX12 and multi-thread native.
B
70
S
24
G
19
Posts: 1,757
Reputation: 17,614

Post » Mon Sep 28, 2015 1:30 am

Think these concerns have been made a few times before and agree with what you are saying. C2 can do a lot of things which are great, but as mentioned you always have the feeling that only the basics have been implemented and all the details needed to really complete it are skipped. The examples that you give are good examples of that. But think one of the most obvious examples of this are something as simple as text object and the fact that its not possible to change alignment during runtime even though you can do it in the editor. I mean its a very common functionality in any form of application, where you work with text. Still after using C2 for such a long time, its still a huge mystery why such details are not seen as important to add. And there are loads of such issues which again underline what you are also mentioning.

So my hope with C3 is that Scirra release a much more polished application from the beginning, where all these details are added. So rather than spending so much time having to adding new functionality that should be there from the release, they spend a lot more energy listening to the community of dedicated users, that in my opinion comes up with a lot of very good ideas that could really improve the program and that they spend time adding those things instead. Most of the people that come up with these ideas are not newcomers to the program, but suggest stuff because they actually have used the program a lot and see gaps in it, that should be fixed. So really hope they keep a list of suggestion made through the years of what people are really missing in C2 so they can make sure its not the same issues C3.
B
44
S
11
G
2
Posts: 1,182
Reputation: 6,838

Post » Mon Sep 28, 2015 1:44 am

@shinkan many of the things in the event system go back to CC and even earlier to gf.

I'm also talking about stuff that we all would benefit from like plugins for in game ui/menu that can be extended if needed trough events, more collision options including those that would work together with pathfinding between units and option for units to avoid each other. Simpler paring of same types of objects. More input actions, like possibility to cancel input at any point. Adding pathfinding to platform games. Basically expanding wherever it can be expanded by taking under consideration different game scenarios. As @nimos100 says, adding details, streamlining where it is possible by having chained actions and conditions on top of new functionalities. etc
My professional Royalty Free Music at Scirra Assets Store
--------------------------------
Specs: i5 2500, 16gb of ram, gtx 770, win 7, Focusrite Scarlett 8i6, Mackie mr8mk2, Alesis 320, browsing the net on chrome.
B
89
S
30
G
22
Posts: 1,985
Reputation: 20,099

Post » Mon Sep 28, 2015 1:50 am

@megatronx
Your desire for better pathfinding or smarter pathfinding is in fact seeking the holygrail of RTS AI. Many many big-time RTS still struggle with smart pathfinding. What they achieve, it's all logic from "hacks and work-arounds", thinking outside the box.

I doubt we will ever get a smart pathfind AI out of the box. It's gonna require a lot of logic.
B
70
S
24
G
19
Posts: 1,757
Reputation: 17,614

Post » Mon Sep 28, 2015 2:03 am

I'm not talking about events. I'm saying that core features of the editor should get a massive update.
ImageImageImageImage
B
157
S
66
G
41
Posts: 2,598
Reputation: 34,823

Post » Mon Sep 28, 2015 3:16 am

What really needs to be discussed is how to better implement the plug system.
As is, third party developers shy away from sticking around for a few different reasons, backwards compatibility, no support, etc.
Then those that do attempt to try to make plugs wind up using existing libs rather than implement their own mechanics. That is something else that could be talked about more often.

Regular users need more plugs, and plug developers need more support, enough said.
Image ImageImage
B
168
S
50
G
169
Posts: 8,283
Reputation: 108,195

Post » Mon Sep 28, 2015 10:14 am

Silverforce wrote:Your desire for better pathfinding or smarter pathfinding is in fact seeking the holygrail of RTS AI. Many many big-time RTS still struggle with smart pathfinding. What they achieve, it's all logic from "hacks and work-arounds", thinking outside the box.

I doubt we will ever get a smart pathfind AI out of the box. It's gonna require a lot of logic.

Think you have a point there, but also in defence of Scirra, its not easy to meet every ones demand on which path finding method should be used, they use A* which is fine I think but of course if they can improve it somehow it would be good.
But now that they have added a path finding behaviour and you start working with it, you quickly realize that its very basic and lots of details are missing in it. Like the single collision mesh, which of course is a design issue, but also there could be an option to make the path finding good for grid based games. And an option to make it move as close to an object as possible if you try to path find to a closed off area and that it would be able to path find correctly when in comes to diagonal movement etc. Besides the collision mesh, these things are details that should have been added to the behaviour to increase its functionality. So having to make work around is fine, but it shouldn't be that you have to spend 80-90% of the time doing this, because its obvious that its stuff that should have been in C2 by default, but aint.
B
44
S
11
G
2
Posts: 1,182
Reputation: 6,838

Post » Mon Sep 28, 2015 11:17 am

newt wrote:What really needs to be discussed is how to better implement the plug system.
As is, third party developers shy away from sticking around for a few different reasons, backwards compatibility, no support, etc.
Then those that do attempt to try to make plugs wind up using existing libs rather than implement their own mechanics. That is something else that could be talked about more often.

Regular users need more plugs, and plug developers need more support, enough said.


I was also thinking about something like build n plugin editor, where you can build your own plugins with events.

@Silverforce I would like more functionality n those plugins, more versatile options. And speaking of pathfning, having units avoid each other is such a small but significant addition. And it's been done already in early dos days, i don't see a reason why it would be so difficult have now. But pathfinding is just a tip of the iceberg. There is way too much fiddling round with the mundane, and there is no need for simple things to be made so complex, that they need a lot of events to be made, as I've explained in op. x
My professional Royalty Free Music at Scirra Assets Store
--------------------------------
Specs: i5 2500, 16gb of ram, gtx 770, win 7, Focusrite Scarlett 8i6, Mackie mr8mk2, Alesis 320, browsing the net on chrome.
B
89
S
30
G
22
Posts: 1,985
Reputation: 20,099

Next

Return to Construct 2 General

Who is online

Users browsing this forum: No registered users and 4 guests