18 January 2013
Guest post: How we hacked together Snapchat for video in a day
As the co-founder of Lamplighter Games, Andy Minkstein writes on developing this week’s App of the Week winner, Vidburn. When first released, Vidburn was pegged the ‘Snapchat of video’ however the same day of Vidburn’s launch, Snapshat added video functionality to their own app. Check out Andy’s tips for building a video-based app in record time, and his experience working with Corona Enterprise.
Want to know how we built Snapchat in just one day? Ok, ok – it didn’t exactly take just 1 day. However, we did make the core of the app easily within a day. Once we decided the app was worth launching, we spent the next week refining it so we could publish to the App Store.
For those of you who are unfamiliar with Snapchat, it’s a neat little app that lets you send self-destructing pictures to your friends. They are viewable once for a pre-set number of seconds and then disappear forever.
As you can imagine, the launch timing of our app – Vidburn – was both interesting and unfortunate due to the fact that 45 minutes after our launch piece on Gizmodo, Snapchat lifted the veil on a big new update. Snapchat’s update provided a feature similar to what Vidburn was offering – it allowed users to send short, self-destructing video clips.
Either way, a lot was learned while producing Vidburn. To start, we created the alpha version in a self-policed hackathon style coding session where we drank numerous cups of coffee, stayed up late and got a working prototype. Once we realized we had something, we took the time to tie up the loose ends, fix bugs and get it a little closer to production-ready. We would definitely recommend that all developers engage in these types of hackathon coding sprints. It’s not good to do everyday, but once in a while it can not only be fun, but rewarding as well. Developers will often be surprised at their level of output when faced with no other choice but to “get it done.”
One problem we faced from the very beginning was that, despite our love for Corona SDK, the media APIs did not have any methods for capturing video. The obvious solution was to make the leap over to Corona Enterprise and to natively code any of the functionality Corona SDK did not have. We have a few key takeaways from this experience, that we think every Corona developer should consider before making the jump to Enterprise:
1) Native coding is much harder than coding in Lua. We guarantee you that you will appreciate the quick, iterative style of developing apps with Corona when faced with a challenge that can only be solved with Objective C or Java.
2) Corona Enterprise offers unlimited possibilities. Your ideas no longer need to be constricted by what the Corona APIs can do. If you’re willing to roll up your sleeves and dive into native coding then anything is possible. Just remember that any extensions to the platform have to be done twice, once for Apple and once for Android.
Alex M
Posted at 14:41h, 18 JanuaryHere we go again. Of course the app of the week is Enterprise. It starts out so innocuous, with no mention of enterprise being needed in the original AOTW blog. Get people interested, get them asking how it was done, then wham! Enterprise! Of course. It is a great bait and switch tactic.
First was premium support for pressing bug fixes, then the promise of enterprise, now the promise of project gluon… what’s next? It seems like a lot of indies are being shafted.
If nothing else you are on a slippery slope, alienating the customers who got you where you are and pulling bait and switch style tactics with enterprise.
Pelayo
Posted at 15:13h, 18 JanuaryI agree. Absolutely. Looks like the time to switch to another SDK is approaching.
Corona SDK roadmap is abandoned!!!
Ale
Posted at 07:34h, 19 JanuaryAlex M, you have forget the Corona Levels project 🙂
Ale
Posted at 07:37h, 19 January@Alex, you have forget the Corona Levels project 🙂
David Rangel
Posted at 10:46h, 19 JanuaryHey Alex – I’d like to understand how exactly this is a “bait and switch”.
So, would you suggest we just not talk about (let alone name App of the Week) any apps made with Corona Enterprise? Should that be forbidden out of respect for Indie developers somehow?
Dave Baxter
Posted at 14:43h, 18 JanuarySo where is the article telling us how they hacked it together in One day ?
Dave
Andy M
Posted at 15:06h, 18 January@Alex – I don’t know if I would look at Enterprise that way. If anything, its an opportunity to take advantage of the Corona API’s when you can, but also not be limited by them. They didn’t have a video record feature yet, so we built it. I’m sure they will eventually have it as part of their core API’s but we didn’t want to wait.
@Dave – Basically the core of the app including video recording was done in a sort of hackathon coding session. We spent the next week fixing bugs and optimizing the interactions with the backend.
Alex M
Posted at 15:23h, 18 JanuaryHey david baxter, you see what I mean by bait and switch?! There was nothing in the article on how they hacked it in a day other than “we used corona enterprise” 🙁
Dean
Posted at 01:01h, 19 JanuaryWithout fail Alex M is back with some complete delusional nonsense.
Alex for the sake of our sanity, please go and use another SDK.
Why are you still hanging around, are you really that bored, I am sure you have better things to do.
Cao
Posted at 02:36h, 19 JanuaryWaste of time reading this garbage article!
Noe
Posted at 11:21h, 19 JanuaryBasically…buy the enterprise and code native…Whick is not the user profile of corona mainly.
J. A. Whye
Posted at 23:05h, 21 JanuaryFrom what I can tell Lamplighter Games would be considered an indie studio, so I’m not sure how Corona Labs is shafting anyone. And just because you’re an indie developer doesn’t mean spending $1K for a cross-platform development tool is out of reach. For a hobbyist, maybe. For an indie who’s actually in business to create apps, no.
And they did say how they put it together in one day — they created the alpha version in a hackathon-style coding session fueled by caffeine. In other words, they just coded, heads down. It was only after the alpha was done and working that they went back and cleaned and polished.
Also, to do what they wanted, they used the Enterprise version of Corona.
Those two things — code like crazy + Corona Enterprise — is how they did it.
There was no code dump, not technical details, but I still appreciate hearing stories like this. They can either inspire you, or you can pick them apart and find fault with everything.
Jay
Dominic
Posted at 05:16h, 24 JanuaryI agree with Jay as with most of these posts they are for inspiration more than ‘How to…”
Interesting to hear if the revenue covered the Enterprise License cost though 🙂
Dan
Posted at 07:29h, 01 FebruaryIf you’re willing to code in objective-c why use Corona Enterprise and not just the usual iOS SDK ? BTW, the article title is misleading.
Hendrix
Posted at 03:48h, 17 JuneI would just mention the fact that Corona Company needs to make money too like anyone else. The offer much for free in the free version (of course to get more people in but still….) I see it this way: If I choose one SDK for development, I like that company to be stabile and able to support and maintain the whole system for us developers. This is my ONLY concern bout this issue…..”will it be here tomorrow” I dont mind paying a reasonable price for this product.