It certainly doesn't seem as promising a territory as it used to. Not so much because it's more dangerous as because Twitter hasn't turned out to be a "platform" in the same sense as say iOS has.

Does the same apply to Facebook? And if not, is it likely to within the next 12 months?

It seems like a lot of developers/startups put their eggs in the FB basket. I wonder if, from a funding perspective, that behaviour would set off alarm bells today?

Facebook seems somewhere between Twitter and iOS.

Sounds like we need a standardized definition of what "Platform" means.

Would you consider iOS and Android similar in terms of risk and developer potential?

To most startups we fund, iOS is way more important. Nearly all build for iOS first and then maybe one day port to Android. There are a few exceptions like Kyte (http://kytephone.com) who use Android to do things you can't do on iOS. And of course Apportable (http://apportable.com) has been very successful auto-porting iOS apps to Android.

> Nearly all build for iOS first and then maybe one day port to Android.

It's unfortunate that this persists even though Android now has twice the smartphone marketshare iOS has. It doesn't seem like things will be changing any time soon either.

And China has more than twice as many internet users as the United States. However, I'm pretty sure the American users spend more money per capita and that it's not an accident that most YC startups target them first.

I'd argue the fact YC is an American incubator also has a lot to do with who YC companies target. A lot of internet VC-backed startups don't particularly depend on their users spending money.

My point was (fairly obviously) an analogy. However, almost all startups have one of two business models: either they sell things to their customers, or they sell advertising space to people who want to sell things to their customers. Both require customers who spend money.

Remember, if you're selling ad space, the advertiser is the customer. The user is the product. Nitpicky, but important. Your explanation works if you replace "customers" with "users".

From what I hear, a ton of those Android phones are really crappy hardware that's given away free by cell phone providers. Android's 70% marketshare isn't all (or even mostly) the Nexus. So it's not even clear whether many of those phones could even run a lot of apps, let alone run them well.

Additionally, for various reasons, iOS seems to be both "hipper" and more profitable for developers. This could be related to the point above, if most of the Android phones are owned by people who don't use apps much, let alone buy them.

> Android's 70% marketshare isn't all (or even mostly) the Nexus.

That's very true. However, even amongst those in the know (e.g., those in the HN community), almost no one uses an Android phone, even the Nexus, as their daily driver. Even the best Android phone is intrinsically viewed as inferior to the iPhone, which is why I said I don't see things changing any time soon.

So I wonder how much of that decision not to build for Android is based on personal bias. We see similar things on the web when it comes to supporting WebKit-based browsers first, with Firefox and IE having secondary importance. For example, many of the "hip" webdevs use experimental features that aren't available in non-WebKit renderers.

> almost no one uses an Android phone, even the Nexus, as their daily driver

Out of curiosity, what field are you in?

At my company in SF doing heavy systems work, we're 80% Android

I'm technically in biotech, but it involves significant software development. Pretty much everyone uses a Mac/iPhone, although the software (for running molecular simulations) itself is run on Linux clusters.

I agree with you. I use iPhone but amongst my colleagues and tech friends I'd say about 1/2 are android.

This is my experience, but I rarely meet an Android user who doesn't openly express that they wish they had an iPhone. There are a few moralists and uber hackers that I meet that are contrarians, but they are few and far between.

I have been witnessing the exact opposite over the last few months.

Well, I'm seeing high uptake of Android in the Java community, even by programmers who don't create Android apps (or prefer OSX). I think it's a combination of the potential to create apps and an uneasiness concerning Apple's lock-in policies.

also price

Android has twice the smartphone marketshare than iOS. But, it is so fragmented that building for the Android market requires you to have at least few different devices to test out your app, not to mention that there are multiple versions of Android and modified Android from the network providers. It definitely requires more effort than building for iOS.

App distribution on Android is actually more tricky than the App Store on iOS. There are actually a few "App Stores" for the Android, Google Play vs Amazon App Store.

I spoke to a lead developer from one of the biggest Android camera apps that sold to a big company. They said that they realised the Android market isn't nearly as big as it's made out to be because most Android users don't even install apps. They have an Android phone as it was running on the cheapest phone they could get and just use it for calls. Many don't even have a data plan.

They were featured on the Android market multiple times but couldn't get the traction of an Instagram, etc. They believe in real terms iOS and Android are closer in terms of actual users that would want to install apps.

I don't understand why the market hasn't shifted to placing more importance on Android. Android has a significantly higher marketshare in the US and abroad it is almost 3 times as prevalent (59% to 23% according to IDC). Why would a startup with a mobile app want to limit themselves to two thirds less potential users? The only real argument I can understand is for paid apps which in some case studies have been shown to have higher revenues on ios.

Two things I can think of that may be factors, whether or not they make complete logical sense.

1. Fragmentation

2. Prevalence of Mac Book Pro as dev machines.

1. fragmentation is very real but not actually a problem and very manageable for 90% of apps.

2. The Android development sdk works just fine on Mac OS. I would argue that certain parts of the Eclipse+ADB combination are actually more developer friendly than Xcode.

Let me share with you an example. I have lots of friends using iphone and lots of friends using android phones. One thing is common for iphone users is that they really don't care about spending $1.99/$2.99 for a small application which they might not use in 1 month. But for android phone users it's a different story. Then i can easily understand the sense that iOS is targeted first when building application. :o)

Market share is not always a good indicator of what your user base will be.

We build music apps for high end stereo's, so it's a mobile device application, but not mobile in the sense of taking a device with you. Most of our market is iPad, with a large share iPhone and iPod touch. Android is kind of small so far (always about to take off!), but I think the Nexus 7 and maybe Kindle Fire will change this. The size is good for browsing music and the price is significantly below the iPad, basically the same as an iPod touch. We have desktop apps too, but they are maybe 10% as popular as the mobile ones. A lot of our users would consider purchasing a device especially for browsing music, iPad is a first choice, but the Nexus 7 is a good option.

The other difference for us is that people tend to keep their existing devices, so we have lots of old hardware to support. Phones get updated a lot more than tablets and iPod touches.

Everything I've seen points to iOS users being a) more likely to use apps/"smart phone capabilities" b) being willing to pay money for things c) interested in new technologies as they emerge. When you ae building a business that involves getting people to install your new app and somehow get money to you in the process, that makes for a much more attractive platform. Sheer number of handsets out there is just the tip of the iceberg (and is always a dubious number since many Android vendors report units shipped rather than units sold to end users, which is what Apple does)

To generalize: people who get Android smartphones don't use them as smartphones. iOS is where the money is, not to mention where the mindshare is.

Android has market share, but iOS has the profit (money and interest).

iPhone users still spend much more on apps than Android users do: http://gigaom.com/2011/12/13/ios-enjoys-3-1-advantage-over-a...

If anybody has suggestions for a good Android outsourcing company, please add them here: http://news.ycombinator.com/item?id=4332873

do you see an API becoming a "platform" in the same sense as iOS/Android or will they remain services that provide content and facilitate specific features?

They seem to have become more of a data aggregator and provider than a platform.

Does the explosion of results-by-tweet at the London Olympics, including first-hand from athletes themselves, shows that as an underlying platform for real time news twitter has strong disruptive potential?

Yes, Twitter itself is clearly having a big effect on the world.

So the conclusion is that Twitter itself will be a significant force for a long time, but principally to their own benefit and building a company entirely dependent on their platform (or any single platform) will continue to be a dangerous game to play.

@pg - I suspect politics might prevent you from answering, but do you have a view on whether this direction of Twitter's, is smart or beneficial to them in the long run?

They seem to be doing pretty well. They head-faked a lot of developers, but I don't think that was deliberate.

>They head-faked a lot of developers, but I don't think that was deliberate.

Yeah I'd agree with that.

Can't say either if this will have major repucussions in the future.

Users probably don't care about these issues and dev's will suck it up, because they still want the userbases.

That said, a social network can potentially become 'uncool', pretty quickly, especially if they piss on the influencers.

Do you feel like Dalton's App.net could grow into a viable alternative?

My how things have changed so quickly. I applied to YC for W10 with RFS3, got in, quickly abandoned the idea since the market didn't seem big enough, did something else (Notifo) on mobile platforms which seemed more promising, watched Twitter quickly destroy it's 3rd party ecosystem, decided never to build on top of Twitter again. All within about 18 months.

Is open sourcing Notifo still in your plans?

I recall that you mentioned that quite some time ago.

And here I am building on Twitter again...

Can you share what you did around RFS3?

See @pg's answer here http://news.ycombinator.com/item?id=2341567

There's actually a generally interesting question here:

When is an RFS considered "closed?"

If one cannot answer this question themselves, it's most certainly closed for them.

I think there is still a lot of potential in twitter as a platform, specifically for the data they hold. There is a lot of information shared on twitter which previously didn't exist on the internet. It would be shared by people chatting to each other, maybe msning colleagues or emailing links to friends etc. That information suddenly being available is what I think is interesting, and still worthwhile building on.

Back in the day, building a twitter client may have seemed reasonable but ultimately turned out to to be 'snatching coins from in front of a steamroller'. I think if you're building something which is a replacement or improvement on the twitter browser UI you're clearly taking a risk. I don't know that other uses of the data they hold are such a risk. Obviously I could be wrong but for me it's at least on the opposite end of the risk spectrum from twitter clients.

[disclosure, my current side project uses twitter data. It's in my profile if you want a look]

RFS3 is clearly "Things built on Twitter", and just that. Read RFS3.

Twitter is more a part of a distribution strategy than a platform to build an entire business off of http://t.co/r6wdLpMc $ZNGA $FB

Pardon my ignorance, but what is RFS short for?

Request for Startups. See http://ycombinator.com/rfs.html

RFS = Request for Startups, which is an allusion to RFC = Request for Comments (https://en.wikipedia.org/wiki/Request_for_Comments)

What are your thoughts as to OStatus?