Jump to content

[TOPIC: topicViewTemplate]
[GLOBAL: userSmallPhoto]
Photo

Corona for Business Apps
Started by russcampbell Nov 27 2018 08:23 AM

- - - - -
36 replies to this topic
[TOPIC CONTROLS]
Page 2 of 2 1 2
This topic has been archived. This means that you cannot reply to this topic.
[/TOPIC CONTROLS]
[modOptionsDropdown]
[/modOptionsDropdown]
[reputationFilter]
[TOPIC: post.html]
#26

Michael Flad

[GLOBAL: userInfoPane.html]
Michael Flad
  • Contributor

  • 268 posts
  • Corona SDK

Because it's a huge task to do it in a way it's actually worth the efforts (a very basic version is just useless) and CoronaLabs is a small company so they have simply to decide what to implement and what not.

 

Also the needs are very different - I, f.i. don't care even the slightliest about an editor as it would simply be useless to me and I am much more interested in Corona engineers to work on improving the audio API, improve the mesh API, support new targets, implement a Metal backen etc. (knowing that I also won't get all of what I'd love to, for the same reason).



[TOPIC: post.html]
#27

nick_sherman

[GLOBAL: userInfoPane.html]
nick_sherman
  • Corona Geek

  • 1,960 posts
  • Corona SDK

I'm not sure there's any point adding features something like Defold or Unity already has.

Corona can't compete with these huge
companies, they need to find ways to differentiate.

[TOPIC: post.html]
#28

Rob Miracle

[GLOBAL: userInfoPane.html]
Rob Miracle
  • Moderator

  • 26,613 posts
  • Enterprise

@Abdo23, there are couple reasons.

 

1. There are two existing "Layout" tools out there that do this:  Tiled (https://mapeditor.org) and Level Director X (http://www.retrofitproductions.com/level-director-x/). While these may not do everything you need, they exist and they work for many use cases. We don't feel we need to re-invent the wheel.

 

2. We have tried twice and both times we ended up killing the project, so we have some experience with this and it's not good. We have a pretty small engineering team and building a fully functional GUI would require multiple engineers working in tandem for a year or more. We would pretty much have to put a significant amount of our engineering capacity on hold to build this. 

 

But with Corona moving towards open source, perhaps an enterprising community developer may want to take this on.

 

Rob



[TOPIC: post.html]
#29

Abdo23

[GLOBAL: userInfoPane.html]
Abdo23
  • Contributor

  • 120 posts
  • Corona SDK

@Abdo23, there are couple reasons.

 

1. There are two existing "Layout" tools out there that do this:  Tiled (https://mapeditor.org) and Level Director X (http://www.retrofitproductions.com/level-director-x/). While these may not do everything you need, they exist and they work for many use cases. We don't feel we need to re-invent the wheel.

 

2. We have tried twice and both times we ended up killing the project, so we have some experience with this and it's not good. We have a pretty small engineering team and building a fully functional GUI would require multiple engineers working in tandem for a year or more. We would pretty much have to put a significant amount of our engineering capacity on hold to build this. 

 

But with Corona moving towards open source, perhaps an enterprising community developer may want to take this on.

 

Rob

 

I appreciate you taking the time to respond. I'm not sure if what I'm about to say would be worth anything to anyone, but in any case, here is my own experience:

I'm not a pro programmer, I'm a designer, but I've been using Corona for the past two years to create my own projects. I keep fiddling around with Unity and other stuff from time to time but I always come back to Corona. Lately, I got an idea of a new game that requires some sort of a visual editor to design the levels. Knowing that Corona doesn't have that I decided to go with Unity. Not to bother you with boring details my experience there wasn't very good, yes it's powerful and easy to get stuff up and running, but I found myself struggling with several issues mostly regarding developing 2D games especially for mobile devices. I also had a look at Godot but despite its huge spike in popularity lately, I found it not to be very mobile friendly.

 

I then decided to go with Defold. While it's very powerful and keeps rapidly improving over time and already have a very active community for both providing help or making useful assets, I found it to be not mature enough and still needs some time to get some basics right. But I'm sure it's only a matter of time before it reaches that point. People there are working hard on it and keep improving it all the time. There might be some workarounds to overcome the issues I encountered there but as I mentioned I'm not a pro programmer and I didn't want to spend a lot of time and effort( and frustration) deciding how to come up with workarounds for these problems.

 

I then did some research trying to figure out where should I go next, and that's when I stumbled upon Tiled. I thought to my self "Great, Corona with a visual editor, I couldn't dream of anything more!" but boy was I mistaken. Working with Tiled itself was easy, the problems started when I tried to link everything back to Corona. The huge problem here is there is no official support between them by either of them. There is only a 3rd party library which had its own issues. When I decided to write my own code to overcome the weakness of that library I found that the exported file you get from Tiled is a complete mess beyond belief, I then realized that Tiled's UI itself is poorly done and gives false assumptions to its user and I've already explained those issues at their forums. 

 

I've finally managed to ditch the 3rd party library altogether and wrote my own code that deals with Tiled's exported file to make it work with my needs. While some people might find this to be a fun process trying to figure out how to solve programming problems and such, to me it was the opposite, it was a necessary evil. 

 

I love Corona, it does so many things right. It's powerful, easy, lightweight, easy documentation and the only thing out there that actually knows how to properly work with mobile devices. But at this day and age, the lack of a visual editor is just a deal breaker. Even though I like working with it, I still find that making stuff appear on the screen, positioning and organising them using code to be very redundant, especially when it comes to UI stuff. 

 

I've decided that this will be my last project with Corona and I will definitely be switching to Defold as it does basically everything corona does (including the use of a very easy scripting language) plus the visual editor. But I would still strongly urge the people behind Corona to re-consider adding a visual editor to the platform,  especially from a business point of view. Not only would it keep current users from migrating somewhere else, but would also attract newcomers and help grow the business. Just my two cents.



[TOPIC: post.html]
#30

Rob Miracle

[GLOBAL: userInfoPane.html]
Rob Miracle
  • Moderator

  • 26,613 posts
  • Enterprise

@Abdo23, I appreciate your candidness and professionalism in your response. You, of course, have to do what is right for you.

 

We wish we had the resources that King has (It's a subsidiary of Activision-Blizzard.) We really appreciate the effort that community developers have put into Tiled map handlers.

 

But it simply comes down to available engineering resources vs. the engineering expense of doing so.

 

Rob



[TOPIC: post.html]
#31

richard11

[GLOBAL: userInfoPane.html]
richard11
  • Contributor

  • 539 posts
  • Corona SDK

Ironically, being a programmer isn't a binary thing. We all work in different ways and feel comfortable in different environments. Therefore we all want something different from our toolset. Additionally, what we want at any one time is somewhat subject to a particular project. There simply isn't a one-size-fits-all toolset, and actually, you shouldn't be constricting yourself to any one single system. I feel that your struggle to settle down with a framework is simply that you're trying too hard to find a framework that offers absolutely everything you need for everything you're wanting to build.

I've been programming for 22 years now. I was doing this before people had the Internet, and at a time when schools, in the UK at least, pretty much just had a few Acorn computers to teach word processing software on. I'm therefore entirely self taught and although I'm a developer by career now and pretty damned senior, some things simply baffle me. Procedural code is where I'm at home. I can work with OOP when forced to, but I wouldn't say I'm competent with it. Drag and drop WYSIWYG coding? Nope, no chance. I have literally zero understanding of that. I have to actually draw an entity and drop it into my world to be able to attach events to it? What if I want that entity to be generated as a result of a series of other events that means it's truly unique to the one player? What if I want the world itself to be built procedurally, and infinitely, like with No Man's Sky?

My point being, I'm old-skool. I love Corona because it ISN'T some modern drag-and-drop WYSIWYG crud. I'm never going to be a Unity developer and I'm never going to wish Corona was more like Unity, or more like Godot, or more like any other system. Those systems already exist and already serve their purpose. They're neither inferior nor superior to Corona, or to any other platform, they're simply built for different people.

That though, is not to say I'd use Corona for absolutely everything. An experienced programmer isn't somebody who can code anything you ask of them in one particular tool, it's somebody who can look at a project spec, identify the most suitable tools for building that project, and then do a solid job of it, using those tools to produce efficient code and a stable product.

Would I use Corona for business apps? Yes. I'm currently building a woodland management app for a client, in Corona, because despite being game-centric, Corona offers a number of features that benefit the build process, such as the way content scaling works and how I can interface with the various sensors using simple Lua code without having to worry about the different platforms and their APIs.

Would I use Corona for 3D games? Probably. I'm planning to at least. Sure there are 3D-centric frameworks out there, like Unity or Unreal, but again I'm a procedural programmer and to me, dragging things around and attaching pre-coded events to them just isn't programming. I can't make sense of it, that's not how I work. I'd rather build my own renderer and know what's going on under the hood when I pan my camera around a 3D model, and therefore how to go about generating, placing, and animating those models on the fly, without pre-created assets.

If a client approached me though, with a short deadline and wanted say... A designer tool creating where he could rotate a 3D model of a shirt around and draw his own design, and this only needs to work on desktop, would I still use Corona for that? Nope. A short deadline means no time to mess around creating renderers, and desktop-only means Corona's mobile-first design methodology or cross-platform compiler support isn't useful. There are better tools for the job for this one, including WebGL based kits that I already know work well in the major web browsers. This one could be built as a web service that doesn't even need the user to install any special software, and bonus, that'll mean it works on mobile too, which the spec doesn't even ask for. Nope, Corona wouldn't be the right choice this time.

Don't "switch to Defold" and never look at Corona again. Play around with everything, learn the pros and cons of different tools, and when projects come along, use your experience to determine the best tool for the job. That tool won't always be Defold, same as it won't always be Corona.

[TOPIC: post.html]
#32

nick_sherman

[GLOBAL: userInfoPane.html]
nick_sherman
  • Corona Geek

  • 1,960 posts
  • Corona SDK

Again, not sure I agree with your assessment of Unity. There's no necessity to drag-and-drop anything, 99.99% of things you can do in the editor can be achieved in code at runtime. No reason why No Man's Sky couldn't be done in Unity, the drag-and-drop part of it just saves time which can be used for coding the behaviour of the world, rather than building it.

 

Prison Architect is built in Unity, and all you start with there is a patch of dirt. I've written many games in Unity with procedurally-generated levels, standing in a VR room as the world you programmed is built around you is pretty cool!

 

If you were to build Sonic The Hedgehog in Corona, you wouldn't place each platform and ring in code, you'd use Tiled, Level Director X, or build your own level editor. It just so happens Unity has the level editor built in.



[TOPIC: post.html]
#33

richard11

[GLOBAL: userInfoPane.html]
richard11
  • Contributor

  • 539 posts
  • Corona SDK

Fair enough, that does make more sense to me then. Perhaps one day I'll look into Unity properly. My current understand is based most likely on the marketing videos I've seen, and I think I read through the initial tutorial once which probably put the focus on drag and drop.

Still though, my point above is the same - different tools are built for different developers, and my own personal reason for choosing Corona is exactly the reason @Abdo23 wants to move away from it. Unity is, I believe, OOP driven and if the WYSIWYG methodology isn't necessary, it's at least encouraged. That's not for me, and I wouldn't want Corona to go that route personally.

[TOPIC: post.html]
#34

nick_sherman

[GLOBAL: userInfoPane.html]
nick_sherman
  • Corona Geek

  • 1,960 posts
  • Corona SDK

And the award for most pointless bot post of the year goes to...

Random quote plucked from an answer on Quora.

[TOPIC: post.html]
#35

Rob Miracle

[GLOBAL: userInfoPane.html]
Rob Miracle
  • Moderator

  • 26,613 posts
  • Enterprise

Here's the problem. Here is the original post the spammer used to get buy our first post moderation rule:

 

"Good day, came at the invitation of a friend. I hope to get a lot of information. There are already interesting topics for me to discuss." This was posted to the correct forum for introducing yourself. Yes it's very generic. I suspected this might be a spammer/bot but I can't be so strict that I block these posts. Its very gray area and I have to make a call.

 

So I let this person in, wanting to have some faith in humanity and when they post a clearly bot/spam post, then I nuke them. As a community you're going to have to accept that this is the way I handle these things. Instead of complaining about it in the forum post and hijacking the thread, I would much rather have you click on the "Report" link, so I'll get notified that a spammer got through.  

 

Thanks

Rob



[TOPIC: post.html]
#36

richard11

[GLOBAL: userInfoPane.html]
richard11
  • Contributor

  • 539 posts
  • Corona SDK

Hmm. Fortunately I read the spam post already and know what Nick is referring to, but now it looks a lot like it's a complaint about my last message 🤔😂

Rob, for the record I don't think anybody here would dispute that you're doing a great job. This is a popular forum and all forums are subject to spam. I've seen forums of a similar size be shut down because they're too time consuming to moderate effectively, and those are forums moderated by more than one person!

At the end of the day, the occasional instance of a bot slipping through the net is preferably to real people being denied their post because it looked a bit spammy.

[TOPIC: post.html]
#37

Rob Miracle

[GLOBAL: userInfoPane.html]
Rob Miracle
  • Moderator

  • 26,613 posts
  • Enterprise

This is why that the appropriate action is if you see a spam post, report it through the report feature that way, these threads don't end up being confused.

 

Rob




[topic_controls]
Page 2 of 2 1 2
 
[/topic_controls]