We need a friendly ENGINE, not a dysfunction EDITOR

well for me the main question is ‘not a dysfunction EDITOR’. And than whole mess starts. One simple thing, coco studio shoudn’t be discontinued because people came to familiar here from cocos2dx framework to cocos2dx engine just because c++. So obvious, developers are expecting a good editor for c++ support. Not the side support.Developer needs some serious core support. Let’s take a good example of overlap2d and Viseditor. At this point there is only 1 editor for c++ and its seriously improving fast. And its better support with cocos2dx using old CSLoader. Atleast it seems better than creator c++ plugin. check this out and see if any c++ guys are interested to make that editor even better.
http://git.oschina.net/halx99/x-studio365-pub/raw/master/releases/x-studio365_10.0.2200.3.exe

1 Like

Editor should be open source, cross platform, and I think written using Qt.

1 Like

Agree with @gplayers.
About web game engine, we had lots of choices, so we don’t discuss it anymore.
Cocos Creator itself had been a problem already, I don’t understand why ChuKong close its source? Do they want to get money from it or something else? No matter what it is, it is ridiculous though.

I know C++ support is good. All the efforts of dev team are precious. That’s why I really disagree what they do til now. If the use those resources to build a plugin/editor on Unity, I think they will make more space and time to build other great things well.

Agree with @anon98020523, too.
QT is great for cross platforms, and it uses C++,too.
Developers who familiar with the engine will feel free on editor maybe.

1 Like

Thats true. And I think we like should move in this direction… but creator is moving into opposite way, also into “china way”.

if you guys want to dev your own editor, I suggest https://github.com/c0i/imguix :wink:

You probably don’t know what it is written with. Most likely core is dictated by engine (Frostbite), but rest could be anything.

In my opinion Cocos team is in the right direction (although when I have seen first versions of Creator I was very skeptical if they make it good)…

Before saying something, I have a lot of experience in C++, Javascript, Typescript, C#, Java, etc… I do not stick to one favorite language, but love to study new thing. I programmed for many platform, except iOS/macOS (although my code is able to be compiled to these platforms, customers did needed NFS and iOS did not provide that; moreover I do not want to use Object-C because of hellish annotation forms - it kills all code readability).

My observations:

  1. I would never write my own IDE in Qt today. Best IDE I know are written in Java (PhpStorm) or some own framework (Visual Studio). I have been working with Telegram Desktop source code (written in Qt) and some older experience in Qt, and I do not want to go there. It reminded me when I was making my first windows apps in 99. So old school and heavy.

  2. If today I am going to write and IDE or any other crossplatform app I will without any doubt use Electron. Moreover, I will use Typescript + Angular + Electron. Example: Visual Studio Code was written with pure JS and Electron.

  3. Looking at changes that Google and others make, I can say that ECMAScript 2015-2017/Typescript is language of the future. First of all, huge development in progressive web apps (PWAs) – check out how much of Google IO is devoted to web programming and native using webtools. Huge increase in Angular/React frameworks, many cross-platform stacks are formed: Ionic, NativeScript, Cordova.

  4. If you say that JS is slow, then Battlefield 2 and Eve online was written in Python (Python, Carl!) How much it slower JS? Even my company writes IoT devices with Python… and that’s more than enough from performance point of view.
    Thus normally, JS is more than enough for any game, given that most time critical tasks are done with native API calls, so outside of any script language. But returning to cross platform Ionic is more than enough to write any mobile app in pure JS without any native implementation, if that is not enough for you you can use NativeScript and it will produce native performance because the app is not anymore on top of simply webview but behind the scene render is native and all your calls are bridged to native calls. So writing natively with JS for any platform? That’s where Google, Microsoft and other big players going.

  5. Node.JS. – just one thing that makes JS development great (and thanks Cocos went this direction). You can find any package there install and use. Recently in cocos creator I created typescipt project and want to work with Dates… OK… npm install --save moment… next I want reactive programming… ok … npm install --save rxjs, etc… This things make development start really fast. One grand benefit.

  6. I still believe that Cocos Creator needs more, but in current condition it is great tool already. They use bindings for JS calls, but if they could choose same approach as NativeScript it benefit more (also I miss dependency injections in constructor, that part of Angular is more powerful).
    They need more smart way of working with animations (currently it is only one way to create animations and it is not straightforward).

I do not want to return to C++ after Typescript. And I definitely cannot survive without Node.js/NPM packages.

3 Likes

So… my company had a meeting about cocos2d-x and its future… Discussed all the points in this topic. We came to the following conclusion.

  1. The future of c++ in cocos2d-x is almost dead. There aren’t many cocos dev working on the source anymore.
  2. Why do we switch to cocos creator? its new its buggy and its around 5-6 years behind unity.
  3. We want to make quality games without missing out on the new stuff. VR, AR, apple tv etc. And honestly cocos creator is at such a basic stage that there are plenty of more options(other than unity) 50x better.
  4. Cocos is now basically only targeting Chinese market… well sorta, in my opinion.
  5. I’m fine with having no 3d options. I like 2d games but compared unity and UDK again, we lack the community/dev support.
  6. Marketplace- Templates, asset store, guides and modules. Again we are way behind compared to unity.

Now, I’ve always been a loyal user of cocos2d-x for 5+ years. But everything I loved about cocos2d-x is now changing.

We need a friendly ENGING, not a dysfunction EDITOR.

I’m sorry to say this but explottens will be my last game on this engine. Soon as its complete i’ll be switching to unity or udk.

Thanks all. :slight_smile:

5 Likes

I was the same, however experience with unity from my friends tell me that it’s crap), every day they struggling…

However, I will use cocos2d-x, because I have SpriteBuilderX and that perfect editor for me. However, all these unused lua dumb things and js in cocos2dx makes me sad… Really want a new cocos2d-x without all that… just simple C++ engine with cool editor…

6 Likes

SpriteBuilderX and cocos2d-x is a good pairing. I use it too.

@anon98020523 and @gejza Does SpriteBuilderX support timeline animations? I haven’t used it and me and my team are currently searching for an alternative to Cocos Studio(which had a good potential but…) and we’re not sure yet that Creator can be a viable option.

as I remember it’s supported. You should try.

1 Like

Until now i used only frame animations, but timeline animation should also working.

The documentation in Chinese are also very buggy for we Chinese.
Cocos的中文文档我们也觉得不好.

After that, the engine will not even compile.

Why do you want to use cocos for JS ? I thought there are some better options for JS.

Anyone who appeals for open_source-ness as a plus…did you all really contribute? I mean did you all make some changes which has been tested against many many different platforms / devices and don’t crash something else at the engine? If so, I really happy. Because when I changed damn one method at DrawNode I spend eternity trying to make it work with all the devices (even for android and windows!) and still not succeed. When you say that you can change something at open source, what you really means - that you can tweak some bugs for your concrete case.

I think the same way =( And It feels bad. I spent last 4-5 years with this engine and now its slowly dying.
1 developer, working on the engine? heh. ok.

No, what for example? Not simply JS, but typescript with classes, npm and blackjack (but that is minor) .

Criteria:
Any platform that compiles natively to Android/iOS/WP, plus HTML5 for web/facebook. Same code for everything. If consider analogy, it is not even Cordova, it is close to NativeScript but for games.

2 Likes

@walzer amy news :slight_smile:? All will be as described before? Or maybe you have changed your mind?

1 Like

It would be amazing if the engine capabilities could be documented the way Phaser.io documentation is. The link is an example of their audio documentation and we can see everything their audio engine does (fade in, loop, pause resume, etc…) Audio is an example, they also show buttons, loader, Box2d, camera etc…

Cocos has something similar here (Actions) but I’m afraid it is becoming outdated :frowning: I found this to be the most useful resource when I first started learning about the Cocos engine as it allowed me to use its capabilities in creative ways as you said there are different ways to do something.

I think this kind of documentation would do wonders!

@anon98020523 I think Cocos Creator is the right direction to move to. I am a JavaScript developer and Creator is the only thing I use to make games now.

Wow what an awesome documentation!

Cocos definitely needs a documentation with working demos and more code examples for every section of the pages, but it would need a lot of resources to make it. On the other hand, it would attract a lot of new users, because one of the main reasons you see people abandoning the engine is the frustration when something doesn’t work, you google it and nothing, you look in the documentation and there aren’t examples, you ask in the forum and no one replies, the last step almost always is, checking the source code of the tests, but not many people like to do that, nor know that it was a feasible way of resolving the problem.

1 Like

I agree that this is really needed.