What, Precisely, Is A 'build Server' Used For? Why Can't That Work Be Done On My Machine?

Getting Started Newbie Questions
#21

Hi Jay, Ken,

Just to clarify, I want to make it clear that the removed post (in which you responded to, Jay) was not removed because it was “anti-Corona”… that is not our policy in any way whatsoever. If there’s any confusion about our policy on that, please read last week’s blog post about the updated forum rules and guidelines.

http://www.coronalabs.com/blog/2013/04/02/corona-forum-rules-and-guidelines/

The reason it was removed is because there is clear evidence that a single person (or perhaps two) were using 7 or more aliases/accounts to post comments that were disruptive and added nothing to the discussions or the community. The individual(s) in question have been doing this for months now, and it was tolerated. Recently, it devolved into personal attacks, among other things; he/she/they were fairly warned, cautioned, and yet still persisted, so all accounts were banned and any new accounts by the same individual(s) will be as well.

Anyway, that aside… Ken, of course you are free to use whatever platform suits your needs. I am just a little confused why a “build server” is a potential deal breaker for you. This has always been the method behind Corona. The process takes (usually) between 20-60 seconds, and it’s not a process that you will perform repeatedly every time you make a change to your code. Many developers code and test in the Corona Simulator and actually build a project a few times per day. Now, I won’t claim that’s the case for all developers, but in general, building is not required every few minutes just to test minor changes in your project.

As for “easing” you into the SDK, this is why Corona Starter was opened up as totally free to everybody, with a solid foundation and a solid feature set. We’re not trying to “reach into your pocket before you’re ready”… in fact, we would welcome you to try out the SDK before you pay. Put an app into market. See how you like the SDK. Determine if it meets your needs. If it does, then continue using it, and at some point if you wish to upgrade to Pro and have the need to, then I would say “you’re ready”. :slight_smile:

Again, it’s up to you. If native coding is your desire/need, so be it. If you wish to tackle the steep learning curve of Moai or Unity, then that is your prerogative. I just wanted to clarify the stance of Corona and the pricing changes.

Take care,

Brent Sorrentino

#22

Hi Jay, 

I must have touched a nerve.  No offence intended. Best of luck with things.

Brent, thanks for chiming in.

As I said, I won’t have my development process dictated by a dodgy internet connection/needing to be connected to the internet, unless there’s a clear reason that benefits me. It’s simply not stable enough, no matter whether I want to compile once a week or every 5 minutes. Since there’s no /real/ reason its necessary it’s simply a sales requirement, and that clearly rubs me the wrong way.  

Adding the ‘upgrade to not have to wait x seconds’ is just an insult.

If it’s really for piracy, use the ‘activate once before use’ model. Don’t put the sting in the end of the tail and get me every time I want to compile for the device (where I have to wait x seconds). 

I appreciate the free version of the SDK.  It’s what has allowed me to try Corona in the first place.

Unfortunately for me, the way your organisation has structured things that’s simply hooking me into an SDK before demanding money from me.  

The way apps are built these days, you do need to have In App Purchase, and I’ve already explained my resistance to the build-server concept and my desire to have the ability for native code.

That means that when I would need to make the inevitable switch, I’d be at the $999 version (and I’m not clear that that removes the build server requirement).  As that simply doesn’t make any sense, all I can do is wish you guys the best of luck.

If you get rid of the build server requirement, and drop your prices a tier, then please do give me a call.

-Ken

#23

@Ken: I believe that if you are interested in Enterprise you can test it for a month or so. Furthermore my internet connection is very bad and sometimes unstable too. But I still manage to get some daily builds up and running in time.

Even though my upload speed is 40kb/s at max - and this in enough to finish my builds in less than a minute(!).

I would suggest you to wait until Corona has finished and released project Gluon and Graphics 2.0 (Pro/Enterprise exclusive!). Gluon might give you what you are missing right now :wink:

Just my 89 (euro-) cents :slight_smile:

  • Max
#24

Hi Max,

It’s not a question of speed.  I’ve got an acceptable 8MB internet line, when it’s working. Once the mandatory wait of 18-19 seconds of up-sell passed, then the build happened quite quickly for both Android and iOS. I didn’t time it but certainly less than 15-20 seconds.

It’s not even clear what’s being done when the “Build server” is invoked.  I suspect that there’s no a lot of  “build” in that “server”.  There’s presumably a lot of us (and a lot of computing power on our end) compared with a limited amount of Corona Labs (and limited computer power comparatively speaking on their end).  From a technical standpoint, it would make sense to do all the heavy lifting on the customer’s computer, and only give final blessing on the server side.

Of course, since it’s not clear what happens during that time…(I’m sure it’s not any of these things, but just as an example of what I could imagine) the source code could be posted on a web server, saved as proof of what you’re up to for later law enforcement scrutiny, the source could be shown to the design department and examined for good ideas that could be lifted and put into their own games, etc.  Who knows?

Again, I’m sure NONE of these are the case!  A company would have to be mad to try any of those…this is me truly stretching possibility into farce to make a point: we simply don’t know what’s going on there.

In fact, if I am correct (by no means a guarantee) that nothing happens from a “build server” except a sales effort and a piracy defence mechanism, then Corona would never want to admit in public.  Because then they’d have to admit they are actually wasting the developer’s time each and every time the developer compiles for no other reason than to try to up-sell them.

So, when working in this way, I’ve constantly got the sales team from Corona jumping in front of me every few seconds saying “UPSELL!”  Every time I start the simulator in the free version, I get the “upgrade now” message.  Easy enough to dismiss, but when I’m dreaming of the architecture of my app, and trying to learn the SDK, it’s jarring.  The “build server” message simply gives them another chance to get in my way to up-sell me again.

I find that irritating, frustrating and insulting, and the fact that they will soon want $600 per year for the privilege, in an environment that doesn’t allow native code, unacceptable.

Even if they were the only game in town, I’d be unhappy with it, but might still go along.  

The fact that there are other games in town for rapid cross platform mobile development (moai, appcelerator, LiveCode, MoSync, PhoneGap, Monkey) make it ridiculous.  Considering almost all of these can target more than just iOS and Android, I wonder if they’ve really thought about their value proposition?

-Ken

#25

Hey Ken,

I do unterstand your points. As far as I know the build servers are required in order to “convert” the lua code into objective-c. This might be done on our machines as well but as you already said it could be just an anti-piracy thing.

A few weeks earlier my “Indie” subscription was about to expire and I got an “Renew your license now” (or something like that) alert everytime I started the simulator even though I had upgraded to “Pro” already.

Things like this are really annoying but not deal-breaking (for me). I would like to see an official statement about the build-servers even though I remember having read such a (blog-) post in early 2011…

At this moment the only difference between “Starter” and “Pro” are IAP and the little advertising when you try to build for a device. I do not need IAP right now and I just make a few builds per day so this would not disturb me.

There are a few features missing in comparison to other engines and I hope to see more information about graphics 2.0 in the very near future.

But what I can say from my experience is that Corona´s real strength is the community. There are other 2D frameworks, but even also LUA-based ones are not as easy and beginner-friendly as Corona.

There are many features that has been announced a long time ago but are still missing or not prioritized right now, you are right. But I hope that this time is over :wink:

  • Max
#26

Hey Ken - thanks for all the comments. Here are some responses/comments:

  1. Server builds - let me be clear, this is not a “sales tactic” or a “piracy prevention scheme”. There is real work happening on our servers. This architecture actually has let us be more nimble over the years as we don’t always have to make client changes when something changes in the target build platforms. So there are real reasons behind why it was built this way.

Furthermore, we have always been very clear about what gets sent to our servers. We convert your code into bytecode and send that up. We do not send up any assets (e.g., images) or raw code. Once the building takes place on the server, everything is deleted and we do not keep anything up there. We never look at your raw code. Period. No need to make up scenarios here :slight_smile:

You’ve made it clear you find this part of Corona annoying and not to your liking. That is fine. We hear this from time to time, probably about 1 person per month roughly. We think it gives us a number of advantages, so the benefits vastly outweigh the negatives. Now, there are companies that have compliance requirements that require offline builds and for them we have Corona Enterprise.

  1. Price and value proposition - The value proposition of Corona is clear: you will build great apps quickly and efficiently, especially if you need to go cross-platform (on the platforms that count today). If you value your time highly, each hour counts. And the hours you will save with Corona make it worth it. If you value other things very highly (e.g., a full open source solution) then your priorities may be different, and that’s fine. Or you may enjoy the challenge of other platforms and not mind the extra time. That is fine too. But lots of developers use Corona and have no problem paying for it.

David

#27

Hi David,

Thanks for writing back.  

If it happens once a month, you’re probably tired of answering the same questions.

I guess we’ll have to agree to disagree here.  

Corona sees value in using a build server, where I do not.

Corona feels that the value proposition, even at $600/$999, is competitive because “it’s possible to build solutions so much faster than the competition”. Perhaps it’s time to review the competition.

After reviewing several other packages, it looks like I’m going to be using Monkey.

Using an object-oriented BASIC language, it gets points for simplicity/speed of development.

It translates BASIC into various languages (std-c++(desktops), Html 5, Flash, Objective-c, java (android), c# (XNA (xbox)).

Since it’s native, the speed can be pretty phenomenal, and can be tweaked to take advantage of the development tools offered on the various platforms.

The bulk of it is open source, so if there’s a problem that they aren’t fixing fast enough, I can have a go and tell the world about it.

I still get the source code to the part that is proprietary, I just can’t disseminate it myself.

Building happens on my local machine, so no network hassles.

I costs £85, including VAT here in the UK, for lifetime access to updates.

-Ken

#28

I’m new to Corona (using it for a non-game app) and I generally think it’s pretty good. But honestly, I think they are in trouble once Unity’s new native 2d framework is available. I want to like Corona, but whenever I use it I just wish I’m using Unity. I heard “super fast build times” with Corona, but is it really that fast? I’m used to Unity so maybe this is affecting my judgement, but I feel like I could put together a game prototype much faster with the aide of Unity’s visualization and built-in components.

There are so many benefits to Unity, such as the awesome IDE, local building (I know, I know - I read this thread and the benefits of a build server, but I don’t think they outweigh the cons), high performance, C# (I think it is about 1000x superior to Lua) and much, much better pricing.

I’m not trying to knock on Corona here, I like the product, I just am confused by a lot of things with their direction. 

#29

@Helios: What feature do you miss in Corona´s 2D engine? In my experience Corona is the fastest engine if you are making a prototype for a project. If we have got a cool idea for a mobile app/game, we always use Corona because you do not need very much time to create awesome things. 

There are great Third-Party IDEs out there, I do not miss anything. And to be honest, we have been working with both Corona and Unity… and the performance highly depends on your programming skills, most of the times we´ve got better results with Corona (!) :wink:

The only advantage of Unity is multi-platform export, in my opinion :slight_smile:

#30

Just my 2 cents:

Just a little background, I’ve been using Corona as a Pro Subscriber for about two years now.

We have 9 educational apps published each with a free and paid version and with NO IAP or ads any plugin or native code we just use the core functionality, in fact we could easily use the FREE starter kit and no pay Corona a single dime for it! We are not making huge amounts of money but we do have a steadily growing income.

  1. We choose to be Pro just for the daily builds. We feel this alone is worth the money. 

  2. I certainly understand the reason a build server is needed. server side authentication is the best mean to prevent piracy. Of course a totally free product like MOAI won’t benefit from it…

  3. I feel that Corona has a part in our success, and I estimate it’s contribution is much more than $600 annually. I’ve tried out MOAI, I’ve tried going Cocos2D and even native, NONE of those would enable us to develop our business rapidly enough. 

  4. If you consider other PAID solutions you’ll see the prices are much higher. It’s enough to look at Carlos Icasa’s Platino engine pricing. I would say their product is fairly comparable to Corona, especially with the upcoming 2.0 graphics. Just take a look at their pricing and tell me what’s more fair. Corona is giving for free what they charge $68 MONTHLY! (or $90 for the full bundle).

  5. I personally think the API itself is the easiest you’ll find (as long as you do not go to low end solutions like GameSalad).

  6. The simulator is such a great tool that we build for devices once every few days, I won’t say it’s perfect and we still catch device only bugs but they are quite rare…  As for IDEs there are some great free options but I myself prefer a simpler text editor like TextMate. The simulator relaunches every time I change a file and the feedback for my change is immidiate. really zero need for a debugger at least in our class of sub 20,000 lines of code projects.

  7. Unity and Corona are not competitors in IMHO. I feel Corona is better suited for our needs which are small educational apps which requires at most 4 months of our two developers. Bottom line, If you can do what you want with Corona I wouldn’t go Unity. If Corona SDK handicaps you function wize go to the Unity way. I think a lot can be accomplished with Corona.

#31

I’m new to Corona (using it for a non-game app) and I generally think it’s pretty good. But honestly, I think they are in trouble once Unity’s new native 2d framework is available. I want to like Corona, but whenever I use it I just wish I’m using Unity. I heard “super fast build times” with Corona, but is it really that fast? I’m used to Unity so maybe this is affecting my judgement, but I feel like I could put together a game prototype much faster with the aide of Unity’s visualization and built-in components.

There are so many benefits to Unity, such as the awesome IDE, local building (I know, I know - I read this thread and the benefits of a build server, but I don’t think they outweigh the cons), high performance, C# (I think it is about 1000x superior to Lua) and much, much better pricing.

I’m not trying to knock on Corona here, I like the product, I just am confused by a lot of things with their direction. 

#32

@Helios: What feature do you miss in Corona´s 2D engine? In my experience Corona is the fastest engine if you are making a prototype for a project. If we have got a cool idea for a mobile app/game, we always use Corona because you do not need very much time to create awesome things. 

There are great Third-Party IDEs out there, I do not miss anything. And to be honest, we have been working with both Corona and Unity… and the performance highly depends on your programming skills, most of the times we´ve got better results with Corona (!) :wink:

The only advantage of Unity is multi-platform export, in my opinion :slight_smile:

#33

Just my 2 cents:

Just a little background, I’ve been using Corona as a Pro Subscriber for about two years now.

We have 9 educational apps published each with a free and paid version and with NO IAP or ads any plugin or native code we just use the core functionality, in fact we could easily use the FREE starter kit and no pay Corona a single dime for it! We are not making huge amounts of money but we do have a steadily growing income.

  1. We choose to be Pro just for the daily builds. We feel this alone is worth the money. 

  2. I certainly understand the reason a build server is needed. server side authentication is the best mean to prevent piracy. Of course a totally free product like MOAI won’t benefit from it…

  3. I feel that Corona has a part in our success, and I estimate it’s contribution is much more than $600 annually. I’ve tried out MOAI, I’ve tried going Cocos2D and even native, NONE of those would enable us to develop our business rapidly enough. 

  4. If you consider other PAID solutions you’ll see the prices are much higher. It’s enough to look at Carlos Icasa’s Platino engine pricing. I would say their product is fairly comparable to Corona, especially with the upcoming 2.0 graphics. Just take a look at their pricing and tell me what’s more fair. Corona is giving for free what they charge $68 MONTHLY! (or $90 for the full bundle).

  5. I personally think the API itself is the easiest you’ll find (as long as you do not go to low end solutions like GameSalad).

  6. The simulator is such a great tool that we build for devices once every few days, I won’t say it’s perfect and we still catch device only bugs but they are quite rare…  As for IDEs there are some great free options but I myself prefer a simpler text editor like TextMate. The simulator relaunches every time I change a file and the feedback for my change is immidiate. really zero need for a debugger at least in our class of sub 20,000 lines of code projects.

  7. Unity and Corona are not competitors in IMHO. I feel Corona is better suited for our needs which are small educational apps which requires at most 4 months of our two developers. Bottom line, If you can do what you want with Corona I wouldn’t go Unity. If Corona SDK handicaps you function wize go to the Unity way. I think a lot can be accomplished with Corona.