What I'm interested in is the effect of SSDs on compile time, which is the biggest bottleneck to my productivity. In my case, the build time for Java projects in Eclipse. I've read conflicting stories of how SSD affects this. Over here:
Google also has a vested interest in keeping as many apps free as possible. Most free apps makes use of Admob (Google) as their primary source of revenue, this means that Google gets a cut of this. But for paid apps, the 30% that gets taken away from your sales goes to the carrier instead and Google does not see a cent of this. Therefore by limiting the number of countries from which developers can publish apps, as well as the countries where paid apps are available, Google is actually boosting their own profits.
I have almost given up on the Android market, Looking forward to MeeGo and Win7mo.
"One difference the mobile phone industry might well find appealing: Apple keeps the remaining revenue, but Google gives it to wireless service carriers, minus billing settlement fees."
With Google not getting revenue from the Android store, they have no incentive to improve it in any way. They won't be nearly as concerned about their developers making money, either. It sounds like a terrible plan all around.
They do have an incentive... but an incentive to make developers use ads-subsiadised apps. Maybe that was their main strategy all along, and they told the carriers they could keep any money made on the store to get some traction (and to avoid the carriers creating their own stores).
Huh, I stand corrected. That's really weird; can they tell what carrier you're on when you download an app, even if it's over wifi? Definitely evidence toward Google favoring free apps and not caring as much about the Android Market as much as they should.
First off, congratulations on your success on both the iPhone store and the Android market. As an aspiring app developer for Android, i'm wondering if you are willing to share some specific marketing/sales strategies that lead to your success.
I noticed that for your android app, there is a trial as well as paid version, does the trial drive a lot of traffic to the paid? How are the return rates like on your android app? I heard that a lot of people does the "buy-app, backup with astro file manager, return app , restore app" trick to get freebies.
I also noticed that you have a website for your app, does this help sales at all?
On Android it is fairly easy to some word out, only because it is a much smaller ecosystem. So hitting up all the standard blogs once your app is well reviewed and polished up will generally get you a mention.
From there it is pretty much organic, we did a bit of tweaking of our application name based on what we think common searches are 'news', 'rss'.
People like trials, though it is probably mostly relevant for an app like NewsRoom, where the real utility comes from using it for days. For smaller games I'm not sure it makes much of a difference, though that it is surely true that the Android market is especially harsh to the amusing for 10 minutes kind of games that sell quite a bit in the App store for 99c. Those apps make nothing on Android, as people return them within the 24 hour window.
The trial was also useful for international customers, as we allow people to get an unlock key from our website. Paid apps aren't available in a lot of countries that the Android marketplace lives, so this lets those users who want the full version have a way to buy it. We don't get a ton of sales that way, but a few a day.
I don't think the website helps much at all, though the video might help give people an idea of how it works. We consider those to just be checkboxes for getting blog coverage, most purchasing decisions are made on the phone I think, where people aren't googling the app first.
This is good news. However, in order for Android to become a more viable and successful gaming platform, google better make some DRASTIC changes to underlying Dalvik VM in Android 3.0 or provide us developers with better C++ libraries to write games.
Currently, the Garbage Collector in Dalvik is a total piece of shit. When the GC runs, your game can freeze for up to 100-200ms, totally kills the gaming experience. What most developers who code in Java does is to use object pool and manage memory manually, its a huge pain in the butt.
If you are a C++ developer using the NDK, you are in for a world of pain as well, you can't control EGL context from C++, you can't receive events, can't play/mix sound, can't access assets (easily). To do all of these, you are stuck with using a JNI layer, which just decrease performance.
On top of that, current devices with similar hardware specs (1ghz snapdragon, wvga screen) are all fill-rate limited to 30fps.
I haven't been following this very closely, but Guy Romain has mentioned a language called "RenderScript" in a presentation or two. RenderScript, to my understanding, is a sort of C-like language that the Android team intends to use to fill the gap between Java development and NDK development, enabling less painful game dev.
I'm not too familiar with it, so here's a few links:
Fortunately, there has been plenty of research on GCs that are fast enough for games. So the solution is to use one of those for Android, now that people are actually interested in writing games.
Your link mention two C++ facilities for supporting manual memory management that have no equivalents in Java, and thus (ignoring the smug sarcasm) supports my original point.
A better GC was already mentioned by the OP as another option.
1, The marketplace sucks compared to the app store in many ways. There are few ways to feature your apps on the marketplace vs apple which has the "what hot", "new and noteworthy", "top paid", "top grossing" categories. In comparison, the android marketplace doesn't.
2, On the apple app store, you are given more promotion room on your app page itself for more descriptive text and screenshots.
3, The android marketplace is FILLED with crap and spam apps, just go to the "whats new" section of your android marketplace and its filled with 95% crap-ware and those ringtone apps. There is no clear way for consumers to find good apps vs these garbage apps. This destroys the user's experience in the marketplace.
4, Related to above, the lack of regulation has fostered a system where comment spamming other apps's page and creating apps that does shady things with user's data is actually more profitable than creating quality apps. see here: http://tinyurl.com/2f3r4g9
5, There is a culture of "free" and "opensource" in the entire Android ecosystem and most apps are released for free or are ad-supported. Google seem to love this since if you use their ad system, they get a cut and therefore they don't seem to have an incentive to push paid apps to have wider availability.
On the other hand, I'm looking forward to what Meego and Win7MO will bring for us developers. I hope they will follow Apple's model of having tighter control over their app stores rather than the mess that the Android marketplace is.
It's ironic that what developers hate about iPhone is precisely what makes it ideal for developers who want to make money. This being the walled garden, the inability to easily install free third party programs, and also the fact that there is a financial barrier to entry (a Mac, plus a dev fee) means that other developers are reluctant to make their apps free in the first place - meaning you don't have to compete with people giving their apps away!
It's a brilliant, (albeit arguably evil), strategy by Apple to create a viscous circle where developers bind themselves to iPhone, which means the good apps wind up there and thus users buy the product and pay for superior apps, which makes it more lucrative to develop for, and so on. And they sell some extra Macs and make money on dev fees, and take a cut out of every app store sale. It's really a simply ingenious grand strategy which has worked unbelievably well for them, however unethical or limiting some people might think it is.
> It's a brilliant, (albeit arguably evil), strategy by Apple to create a viscous circle where developers bind themselves to iPhone, which means the good apps wind up there and thus users buy the product
The only problem with your theory is the 'Android Sales Overtake iPhone in the U.S' part. Perhaps Android will need a bigger market share (double?) to really start stealing the higher quality developers from iOS but if the trend continues this will happen.
Not really. Firstly, if you're going to count all Android device sales against the iPhone, you're going to also have to count all iOS devices, such as the iPod Touch and the iPad. At this point, iOS is still well ahead of Android in the US.
Secondly, the US is not the whole world. Here in France I don't know anyone other than tech geeks that have Android phones, but a majority of my friends have iPhones or Blackberries.
Thirdly, you have to take into account the fact that Apple has created a culture of users paying for content. They have also made paying for content as painless as is humanly imaginable - select item, enter password, and it downloads straight to your phone. Android devs have a long wait ahead before they start to see the returns that iOS devs are seeing.
Please stop. This is what every single company on the planet does. Few things are inter compatible and you'd be naive to hope that they would. Apple has solutions (implementations) to a problem which are different to everyone else's.
One thing that may address your first two points is that Google demoed at I/O a web presence for the Android Market that would work in conjunction with the push service on the phone. Buy an app on the web and it's pushed to the phone.
I imagine that if done right this would allow them to match Apple in terms of providing more real estate for promotional content from app developers, and reduce the barrier to buying since you can preview an app, hit "buy," and immediately have it on the phone.
I also think the lack of a good web presence for Android Market is holding back device adoption. I know of several people who are dissatisfied with their iPhones but are hesitant to jump to Android without being able to see if there are third-party apps out there that are analogous to what they've already got on their iPhones.
I agree that one of the major problems with the Android Market at present is certainly this culture of "I won't pay for an app -- and if I have to pay, I'll copy it to my SD card and delete it within the 24 hour return window." A developer has little incentive to make a really, really good paid app if a good proportion of their users would pirate it. There's a tipping point where you can instead make more money by making a mediocre ad-supported app.
I think this is a good thing so long as developers are also generous enough with the grace periods that legitimate users aren't affected.
If you travel to a foreign country for a week and don't have international data roaming turned on, will you need to find a coffee shop and connect to wifi sometime in the middle of your trip to re-authorize your apps?
The nice thing is that it does leave these decisions in the developer's hands. If it's a network-centric app that accesses a web service that costs the developer money to operate (and thus piracy results in monetary loss), there's no reason not to have the license be strictly checked every time the app is opened. If it's just a game, maybe not so much, because someone probably will want to play it on an airplane.
I agree that one of the major problems with the Android Market at present is certainly this culture of "I won't pay for an app -- and if I have to pay, I'll copy it to my SD card and delete it within the 24 hour return window."
I haven't seen any evidence that piracy on Android is worse than on iOS. My paid app has a 5-10% return rate; even if that's all piracy, it's not worth worrying about.
The problem with piracy on android is how EASY it is to pull it off. On the iPhone, you have to jailbreak your phone, a step which many consumers are weary of doing in fear of bricking their phone, after this, they also have to find a cracked .ipa to download/install. On android, you simply need to buy an app, copy to SD, return it and viola! free app. Based on stats of my paid android app, there are about 3x as many pirated users as users who actually paid for it =(
>5, There is a culture of "free" and "opensource" in the entire Android ecosystem and most apps are released for free or are ad-supported. Google seem to love this since if you use their ad system, they get a cut and therefore they don't seem to have an incentive to push paid apps to have wider availability.
This isn't entirely consumer-driven, IMO. I have no problem supporting developers, and I've bought the "donation" version of a number of apps to support developers, but the fact is that often as not when I look for an app to do something, the free app is the best choice available.
There are times I find myself wishing there were more quality apps I need out there that the developer would take money for, just to help end this notion that Android users don't want to pay for software.
1, The marketplace sucks compared to the app store in many ways. There are few ways to feature your apps on the marketplace vs apple which has the "what hot", "new and noteworthy", "top paid", "top grossing" categories. In comparison, the android marketplace doesn't.
Android Market does have "Top paid", "Top free", and "Just in" for the overall market and each app subcategory.
>I hope they will follow Apple's model of having tighter control over their app stores rather than the mess that the Android marketplace is.
It is really interesting that you hope for tight control, while I am waiting for a platform with as little carrier and/or platform vendor control as possible.
In the ideal world, I want everything to be "open" and "free" too. However, there are simply too many spammers/unethical people out there who abuse such a system. Therefore I rather have a high quality app store over one that is filled with spam apps but is "open".
I'm not referring to "free as in beer", but for freedom of choice. I want to be able to install & use apps (for which I paid) without carrier or platform vendor meddling. I don't want to root the phone, don't want to unlock the sim, don't want to be able to use only those apps that are approved by unseen/unknown authority. "Spam" apps as you call them are not a problem, as they are not in "pc land" unless you install everything you come across.
1. I really don't think this is the job of a marketplace with tens of thousands of entrants. In the Android world there has been a heavy push for QR codes to apps, so I seldom ever browse the market, and instead read reviews and blog entries on apps, following the QR codes to those that sound interesting. I'm not sure why developers think that someone else should do the promotion for them, and all classic promotion vehicles are still necessary.
3. So is the Apple market. There are 300+ fart apps (probably much more by now). It is absolutely true that the Android market is full of junk, but don't say that as if it's the exception. 99% of iPhone apps are make money fast garbage.
4. Spam is definitely a problem, and it should embarrass Google that they haven't dealt with that. On the "shady things with data", sorry but your iPhone is far more susceptible to this.
5. Bullshit. On the iPhone there was a novelty aspect that is quickly fading away. I remember having peers gushing every day about all of the apps they purchased. That has faded away and now they use a very small cross-section of top tier apps.
"On the other hand, I'm looking forward to what Meego and Win7MO will bring for us developers"
Indeed.
"I hope they will follow Apple's model of having tighter control"
This is such a bullshit myth. Apple's "control" is superficial and facile. That you promote it shows that you're a bit gullible I suspect.
Unfortunately, the Android Marketplace as well as the SDK still sucks for us developers. Buggy and inconsistent behavior across different devices; Bad performance (MotionEvent anyone?); The lack of paid apps in different countries and the prevalence of spammy crapware apps on the unregulated marketplace. I wonder if google will do anything to address these issues as Android gain a wider audience. Otherwise I can see this as a hindrance to the adoption of Android.
I agree. I would care so much more about this _iff_, as a software developer, I could capitalize on it. Looking at numbers from various posts and outlets, iPhone (or even iPad) app sales are orders of magnitude larger than equivalents on Android Market.
If I were a _device_ manufacturer, the "Free as in Everything" aspect of Android is definitely alluring, as it saves from Symbian or WinMo royalties. That's where it matters: to the Motorolas, HTCs, and (eventually) Nokias of the world.
For a hacker doing software, though, it just doesn't come close to the App Store as far as outlets for monetization.
It looks like Apple gains market share in web usage, and doubles Android in growth, according to the report. So, it looks like even though Android may overtake iPhone in sales, the users don't often use one of its primary features as a smartphone. Is it the browser experience or laggy hardware that stymies the growth?
Actually the iphone Cocos2D is much further along in development than any of the android engines. Also, if you are referring to this cocos2d port here: http://code.google.com/p/cocos2d-android/
it is a buggy and rather useless port due to the liberal use of temporary object allocations which totally chokes the garbage collector on dalvik
The problem with rokon (when i looked at it) is that it runs both the rendering and game logic on one thread, which is not optimal for performance. AndEngine takes this a step further by running it on 2 threads and only locking them into sync during the geometry submission phase of the rendering pipeline. By doing this, you save about 10-12ms per frame to do your game logic while an eglBufferSwap is happening (on my Mytouch3g), which in some cases can cause a big speedup for your game.
http://stackoverflow.com/questions/499889/ssd-and-programmin...
its mentioned that CPU is by and large the bottleneck but over here:
http://javamoods.blogspot.com/2010/04/hdd-ssd-battle.html
the author mentions that the SSD gives big speedups to compile time.
Anyone has any data to share in regards to this?