I think this is a wonderful idea --- companies selling in-house tech they no longer need or use. I work in corporate IT ( finance) and we have mountains of internal apps that we built and deprecated over the years. I wish some forward thinking CIO's would realize the value of stuff we throw away !
In case you’re curious, I’ll be starting another business with the proceeds. We spent a long time getting this group video chat to work really well - ours is the only tech which actually works and doesn’t get choppy when you add more people. We are very proud of this tech and it would be sad to see it rot.
Hat tip to Justin Kan for the inspiration with his sale of Kiko (http://areallybadidea.com/selling-kiko). If we sell this I'll be sure to write a post about it too :)
What about webrtc solutions? We currently have native single party calls working that interface with either webrtc in the browser or another ios device with very little if any chop. This runs very well on iphone 4s+ and ipads. We are implementing multicall this month and very optimistic.
Are there any special problems you ran into doing multi call over ios?
Our tech implements server-mediated video chat for up to 6 participants. It requires some processing power on the server, but it is designed with a "scale-out" architecture so you should be able to get it running cheaply at scale without major rewrites.
Custom server or public? ( fms/wowza/etc. ) How many streams per server can it handle? Why bother with a server doing relay if its only 6 people?
It's currently iOS only, but it should work to copy the iPhone code's structure to implement clients for all relevant platforms -- Android, WinPhone, Blackberry, web, and/or desktop.
It is my understanding that developing video chat for android is a major clusterfuck. Not only does each O/S version create crazy complications, the differences and nuances required just across phone models is a nightmare. For instance, on a Evo you need to flip the image. On a newer droid the f/b camera is reversed from the API. Its enough that developing video chat for android is a nightmare in disguise
Performance will be critical for slower devices, so you'll need to implement it carefully.
The video codec is h.264 at 144x144. (We experimented with up to 320x320 and found that with current connection quality on phones, especially on urban wifi and 3G, sticking with a low-resolution default tends to produce the best user experience.)
Don't things look like shit in 144x144 for non-group calls? Do you scale larger when its just a 1 to 1 call?
The audio codec is Opus at 16 KHz.
Are you muxing connections together on the server? Do you lose individual volume control per feed?
I think the most important question is related to server technology. If you can only get a small amount of users on each server, its not going to be possible to be profitable moving forward.
We created a custom server. We wrote our own server code, that's just how it was designed.
Android
If you think it will be hard, you're probably right. Thanks for the info.
Resolution
Yeah, 144x144 is definitely a bit blurry. With a bit more work you could upgrade the resolution for 1-on-1 chats. We didn't put that work in yet.
Audio
You could control the volume per feed -- but that said, when you use the Voice Processing I/O unit, Apple does a pretty good job figuring out a decent microphone level, so it was not necessary.
I can provide more detailed answers but you should contact me privately.
The limiting factor is concurrent video chatters. We only got into the four figure userbase, so we didn't have even enough to push a single AWS C3.large instance to its limits. Contact me privately for more info.
currently 4.1 (jelly bean) and above and we "officially" only support S3 although many of our partners have gotten the suite of samsung and LG devices to work with the SDK
Would you consider running a self-starter (as well) to open source it? I could see contributing $100 if it meant the world got a new, quality solution. Maybe there are lots of people like me?
"My startup, Wave, is the easiest way for immigrants working in the US to send money to their families. We're launching for Kenya and Tanzania soon. Sign up on the site to stay informed."
That's the new business. The old business was group video chat targeted at middle schoolers. We are putting the cash towards the quite heavy regulatory burden on money transmitters.
Hey! Don't hesitate to contact me when you launch. We have a comparison website for money transfer services. We love to feature new services (specially around mobile money)- www.tawipay.com
I just wanted to point out an alternative business model, not necessarily for the OP in particular, since the OP may have decided to sell because he wants to work on something else. An alternative business model is to open source the video chat software, being free software gives you viral distribution if the software is good. You can then sell your expertise in the following:
"...Getting this right required expertise in video and audio encoding, deep systems-level understanding of the iPhone platform, years of networking and distributed systems experience, and a hefty dose of creativity..."
Yes, great idea. We considered this - and maybe a different company would do very well doing that path. Unfortunately it seems like too big of a distraction from the business we wanted to start.
This is professional services if I get your point right, that can be a profitable business but it doesn't scale well. It all depends on what you want to achieve.
Looking at the Ebay page, it says "may not ship to the UK", and that the product only ships to the US. Is that intentional? It seemed from the description that there was no restriction on who/where you sold it to?
This idea is brilliant though. I hope you find a great new home for the technology with a company who use it well.
Cool. I'm not trying to buy it I'm afraid, but it might put off people who are interested. I know it made me wonder if there were export restrictions or something stupid like that, I know crypto stuff has had issues with that in the past.
Did you guys considering selling it on other platforms besides eBay? (I don't know of any; I'm asking out of curiosity of whether there's any competitors for hosting an auction like this)
P.S. I looked at your site -- I took 252r this semester and loved it, though the topic was different this year.
I would have bought it in a heart beat for $2k-$5k. But 50k is really out of my league. Have you considered selling licenses instead. I don't care if I am the only one who owns the code. You might get more sales as well that way.
Yeah, I considered it. We even had an offer. We just can't afford the support/maintenance burden, it's too big of a distraction from what we want to do with our lives.
(Any entrepreneur with the cash is, of course, welcome to purchase it and resell the licenses yourself! You'll need to be quite tech savvy to maintain it though. I'd spend my integration week however you want, including teaching you everything about how the code works.)
If you ever consider selling a license, please send me an email. You just need to provide basic documentation. Maybe 2 months support. But that's about it.
So 144x144 video, without compression, is about 21KB/frame, with 5 other people is ~102KB/frame download. I'm curious if the low pixel count is where the biggest gain from this comes from....
what are you doing to make remittance easy? As an immigrant (Sierra Leone), my family regularly sends remittance and for years have felt scammed by Western union. Are you going to make it significantly cheaper? Paying $20 to send $100 is a scam. Is this possible without significant fund raising?
and only 0 bids till now! waiting to see who buys for 50K and pays extra 20K (Lincoln's at least two week time for integration and KT) to get the code and <9999 users. Ohh.. btw, you don't get the name!
Very interesting idea, good luck with the sale and your new endeavor. It's definitely not easy supporting a commercial video chat app these days. There are a few things I'd like to point out though, for anyone following the discussion.
On your site you say that ooVoo doesn't work well for group chat on mobile devices. I’d like to say otherwise. Have you ever checked out the all new ooVoo mobile video chat SDK?
We recently soft launched our mobile video chat SDK, so we haven't advertised yet, but I invite everyone on HN to check it out: http://developer.oovoo.com. We've spent quite a bit of time building out a killer SDK (and all the related additions to our backend) for mobile video chat. Right now our SDK supports iOS and Android, and we have a Windows phone version coming in about a month, and yes, even blackberry too is coming soon. We’re working on adding new features like crazy and we also have a native desktop SDK and WebRTC support (with some nice tricks ;)) coming down the pipe which will let you do desktop <—> mobile chat. It’s so easy even a 9 year old can do it: http://readwrite.com/2013/10/07/teen-hackers#awesm=~osusMMqT...
Don’t take my word for it though, check it out. If you want to see how well it works, you can find our sample app on github: https://github.com/oovoodev/iOS-SDK-Sample (NOTE: You still need to download the SDK to compile the sample app, because it needs some of the bits from the SDK bundle.)
In terms of quality, we recently shipped our patented SuperClear technology which greatly improves the quality and experience of desktop AND mobile video. Especially in situations of network congestion/degradation. Users have been pretty happy, we've consistently maintained a 4/4.5 star rating on the Apple App Store. You can feel free to try it out: https://itunes.apple.com/us/app/oovoo-video-call-text-voice/.... We are also consistently rated as having the best overall quality of any consumer video chat product.
Our infrastructure is world class, our A/V cloud is distributed all over the world so that we can support over 90+ Million users doing hundreds of thousand of concurrent calls every month. Thats over a billion minutes!! It’s no easy feat but we do it because we are love providing the best quality video chat experience no matter what platform you’re on. We’re also committed to giving developers the best experience possible. We’ve spent years building out our services and infrastructure and so we are happy to open it all up to developers now. We’re also network agnostic, so no need for you or users of your app to have an account on any social network if you don’t want them too, not even an ooVoo account. It’s also a fair bit cheaper than buying an app outright, it’s just .005 cents a minute (half a cent) to build something kick ass.
We have more than a few hundred developers who have been playing with our SDK, everything from consulting shops, to shopping apps, to large TV networks and the feedback has been great. We’re glad to open it up and we do our best so that you can do your best and we’ll take you as far as your imagination will go. For any questions/comments you can email oovoodev@oovoo.com.
We think maybe! We actually filed a provisional patent app (just did it ourselves on the cheap) covering the technology. We haven't talked with a patent lawyer yet about whether the tech is actually patentable though.
