POLL: What do Cocos2d-x users care about the most?

Announcements
#13

Unity/Unreal are not free and I hardly believe normal indie developers spend that money as start up.

But I do agree that migrating project to new release is always a pain. That’s simply because there is no correct documentation on migration. If something needed to be removed, Cocos2d-x team should give us a guide to migrate.
Keeping old/unused materials exist does not help much but just making project complex. We need a lightweight engine and not a monster project.

Back to topic, what Cocos2d-x users care about the most is “how to develop their apps effectively”. Providing better documentation is a way to improve that.

#14

My own opinion on the topic : If users thinks we need more/better Documentation, it means the projects/code is not understandable enough.

Either because it s not organized as people expect, because it doesn’t build on users platform out of the box, or because the cocos doesn’t behave as user expect it to.

So more documentation might not be the solution to the “Not enough documentation” problem.

I personally think that cocos needs a lot of cleanup / reorganizing effort instead of “more stuff” that doesn’t work as user wants it to work.

1 Like
#15

Unity actually has a free version and you can release with the free version, it isn’t even as limited as you think (please look it up elsewhere).

I meet with 250+ Vancouver indie devs, nearly all are using Unity, none apart from me are using Cocos2d-x (some used to). Even the students at the game dev college are learning Unity, they even get deals on Pro licenses. Game industry jobs, games at all the shows, indie devs, dev companies, … 90%+ Unity.

The proof is out there.

Unity as a company are doing everything right.

Back on topic…

Please realise I’m not bashing Cocos2d, I love Cocos2d-x. I would actually like to see it re-gain its popularity in the industry, hence all this.

Why should we care about this when we are talking about users and improving Cocos2d-x?
It is going to be a lot harder to attract new users if Cocos2d-x isn’t used in the industry. Any semi-serious new developer is going to pick a system that is used in the industry, this was one of the main reasons for me choosing Cocos2d-x back in 2010, a good system to learn, potential to get a job at the end of it; Unfortunately this is not the case now, not in the USA/Canada/UK at least.

I believe it could be again, but we (or Chuckong) need to set some serious goals. To be the equivalent of the Unity system for 2D games, along with the sort of support Unity has.

Anyway, I will stop here. I have to work the rest of the week so probably won’t be annoying you for a while. Sorry if you all think I’m nuts or you think I’m talking crap, maybe I am…

#16

I’m jumping in the discussion. Of course documentation is the #1 problem with the current release. But what would really make a difference, at least for me, is a tighter content pipeline. Further developing CocosStudio and bringing the OSX version on par with the Windows one would be a good starting point.

And to add to what has been said about the cocos2d-x is free while Unity costs money, please notice that you can work with the free version of Unity and then subscribe to the Pro version for just 4,500USD or 225USD/month (including iOS and Android Pro). If you’re serious about publishing a game, 225 USD to be able to put your game on the major stores out there is pretty much nothing. So yes, Unity is a good contender to cocos2d-x but since we all love cocos2d-x, I’d like to know what the future roadmap is.

Cheers!

#17

I’d say better documentation and fix bugs / extend COCOSTUDIO!!

#18
#19

What I mean “free” also refers to whether you are free to break down any limitations.
Cocos2d-x are free, providing that you have enough programming skills. For Unity, most features are available. But you cannot strip the code for Android app, meaning that the file size is big. Unless you app is AAA game and file size is not concern, no one will try to download any mini game app with large package size.

btw, did anyone also evaluate another library such as HaxeFlixel or game development tools such as Stencyl? Quite a lot of indie developer works with them. For example, in Lundum Dare 30, there are quite a lot of entries created by HaxeFlixel. I rarely found a project created by Cocos2d-x
https://groups.google.com/forum/embed/?place=forum/haxeflixel&showsearch=true&showpopout=true&showtabs=true&parenturl=http%3A%2F%2Fhaxeflixel.com%2Fforum%2F#!topic/haxeflixel/B_QUHMi5dyM

#20

I tried HaxeFlixel and it’s very nice to prototype with. And if you come from an AS/Flash background it’s very easy to get started with. The plus is that it’s already multiplatform. The cons are mainly the same as cocos2d-x… it lacks integrated tools to get your game assets together. On top of that it’s also kind of hard to add native features to your game if you need to, compared to cocos2d-x.

#21

Generally speaking, for a version number x.y.z:

  • 1 year for a major version x. e.g. 2013 for v2, 2014 for v3, and 2015 for v4.
  • we have 4~6 week for each minor version y, e.g., 3.1, 3.2
  • and use z only for fixing bugs
#22

@GMTDev, thanks for your kind feedback. I reply on your titles.

Documentation / Sample Code / Tutorials

What we’re doing:

  • Improve the docs, and even refactor the docs system on this website.
  • I’ve updated lots cocos2d-x wiki pages to v3.x in the last week

Problems:

  • 3 different documentation system in different places:
  • Another big problem is that, most of cocos2d-x maintainers are not native English speaker. So docs which were not written by @slackmoehrle, reads weird.

Support

I’m building a tech support group for paid services in China. If it works, it can be ran in South Korea, Japan, US and other countries. The idea is, if I have enough income from paid tech supports, then we can have, e.g. 20% support engineers always supporting forum for free. I still don’t know if this way can work.

Install App - Getting Started

We did this in cocos.com. And the some solution can be applied to cocos2d-x framework. If we want things go more smoothly, we can have prebuilt binaries in the packages, and suggest coding on scripts (Lua or JavaScript). This will cut down the learning curve for using cocos2d-x. How about your opinion on this?

No More Breaking Backward Compatibility

This is the large complain in cocos2d-x community. My previous idea was that, coco2d-x is still too young, if we keep backward compatibility, it will slow down our iteration. I prefer to have a “better” version, and keep the compatibility later. Seems v3 API set is good enough.

Personally speaking, if we dare to change 80% API in cocos2d-x v4, just like v2 to v3, Ricardo and me would probably be fired by Chukong :wink:

Chukong’s Plan

Chukong’s plan is to have an unified development platform on cocos.com, with Cocos Studio, Code IDE and many commercial services. Our challenges are:

  • Cocos Studio and Code IDE is not very popular so far. Their qualities are not good enough as expected.
  • Chukong regard cocos2d-x and cocos platform as its major business. That’s why Ricardo and me join in Chukong. But until now, we still have no significant income from cocos2d-x.
  • Not all VP-level guys in Chukong has the vision of globalisation. So you can see cocos2d-x is more popular in the east Asian then in the western.
  • As an open source community, we have too much technical fragmentations to merge, to integrate.

I mentioned the future plan in this post. Generally speaking:

  • Unified framework, instead of technical fragmentations in cocos2d-x, cocos2d-js, quick-x
  • Better integrated workflow, instead of separated Cocos Studio and Code IDE
  • Easy to use services. AnySDK is a success in China, and it will bring to the world in 2015.
  • Commercial 3D

BTW, all Chukong’s products will be free for developers, the framework level (cocos2d-x, -js) will keep open source. The plan is to make money from: (1) making games and game publishing (2) revenue share with commercial services such as ads, analytics, push, payment etc.

#23

I don’t like anysdk. I don’t want to upload my APK to some service which is located in China which does some integration magic…

If I am wrong pls correct me.

#25

For AnySDK, you don’t need to upload your SDK to our server. Instead, it’s an offline tool with GUI, you can finish your work with this offline tool, which supports windows and Mac OS X. This tool is based on Qt.

#26

Oh, sorry! The screenshots on en.anysdk.com looks like it is webbased!

And I think you are right. Cocos needs to merge the docs in an easy and intuitive to use docs system. Sadly I can’t name you one :confused:

#27

along with the release, updated documents and tutorials on how to use must be available

#28

@walzer it seems that the AnySDK is great solution for all our needs about social/payment/ad … systems, so i hope that it will be available as soon as possible

#29

Sadly, I review all the whole team and resources today, and we still have only one full-time staff working on documentations in English.
Yes, everyone know him in the forum: @slackmoehrle.

#30

@sillyatom - @walzer is has mentioned that my job is to write documentation. Currently I am writing the programmers guide.

#31

#1 Bug fixes: New features and documentation are great, but if existing functionality is causing glitches or worse crashes, then it makes the system un-usable. This carries over in to some backwards compatibility or maintaining older versions, like making 2.2.6 to support iOS 8 64 bit reqs (thanks).

#2 Documentation: There are a lot of features that I would love to take advantage of, but have trouble finding instructions as to how to implement or modify, two that I am thinking about are iOS IAP through plugin-X, and Sprite3D FBX generation (what are the limitations on the models?). Thank you @slackmoehrle for taking on the Programmer’s Guide, it must be a daunting task.

#3 Features:

  • Cocos Studio: Better Armature Editor (for Mac too): Many of our contractors use Flash to do animation, which my team then runs through DragonBones, then through version 1.6 of Cocos Studio to get a usable JSON file that plays nice with multi-resolution. We could try to get them to learn Spine (I hear it’s lovely) but the licensing requirement for my company as too expensive, and the editor in 1.6 is mediocre at best so I doubt we could get them to use it. I just checked out Spriter, and it seems better that the old editor, but it either it would need to export JSON for Armature/Animation, or ArmatureDataReader would need to read the SCML file that Spriter uses.
  • Multi-store IAP support integrated: Freemium is the current and future wave so making it easy for us to make money makes us more likely to want to spend money on services that Cocos can charges for (support system?). As mentioned before, better documentation for how to use Plugin-X would help, but even better would be an integrated class for store management. For now I’ll investigate Soomla as I think they cover the stores I care about, iOS, Amazon, and Google Play.

By the way, I would be happy to pay for a “fix my project” service to solve crashes or quirks that I cannot find a solution to, especially on Android. Maybe this service could also be a way to find and add new features that are frequently manually added by developers (of course with their consent to use the code).

#32

Having Spriter integrated into the cocos studio would be great indeed.
On a side note, being new here, i wouldn’t mind paying 10-20$ a month if it meant the engine would improve faster(this is in reference to the unreal engine model).

#33

yes spriter integration would be much better.