Google Android

[character]

Yes, but time-wise iOS 4 is roughly contemporaneous with Android 2.2, not 2.1, so it's not really a fair comparison.

IMO it's fair because it's more about fragmentation than who had what when. Can one drop 1.5/1.6 from consideration when developing an Android app? Yeah. Can one drop 3.1.3 when developing an iPhone app? Maybe. I know you have a strong support policy for older OS versions, so you might still provide updates for people running 3.1.3.

2.2's also at least a big a deal compatibility-wise as 2.1 given all of the performance improvements and the new apps they're supposed to enable.

Agree 2.2 is faster, but don't recall complaints about 2.1 being slow.

[...]we're still getting a significant number of update downloads per day, suggesting that a lot of Pleco iPhone users go at least that long between App Store logins; [...]

Not sure of the comparison here. Android updates are usually pushed.

[...]the most active users are naturally also going to be the most likely ones to upgrade to 2.2 / switch to a phone running 2.2.

But the article is about 2.x, not just 2.2.

ETA: From a developer perspective it's important, as one can drop a whole class of devices, the early ones such as the G1 with a ~500 MHz processor downclocked to ~325 MHz. Now the base can be 600 MHz class phones such as the Droid.

 

[mikelove]

IMO it's fair because it's more about fragmentation than who had what when. Can one drop 1.5/1.6 from consideration when developing an Android app? Yeah. Can one drop 3.1.3 when developing an iPhone app? Maybe. I know you have a strong support policy for older OS versions, so you might still provide updates for people running 3.1.3.

But then it's rendered meaningless by the fact that iOS 4 came out so much later than Android 2.1 - you're giving Android the edge simply because their release cycle is on a different schedule than Apple's. A better comparison would look at the relative adoption rates of iOS 4 and Android 2.x such-and-such number of months after their release.

(and we actually still support 3.0, haven't seen any particular reason to stop supporting it yet because backwards compatibility on iPhone is so easy)

Agree 2.2 is faster, but don't recall complaints about 2.1 being slow.

No, but my point was that 2.2's signature feature (its JIT compiler) isn't something developers can assume most users will have, so they need to continue optimizing around older releases. For backwards-compatibility purposes 2.2 is at least as big a deal as 2.1.

Not sure of the comparison here. Android updates are usually pushed.

I wasn't making a case about update frequency but rather about App Store login frequency - if someone's going 2 months without updating Pleco, then they're clearly not accessing App Store very much but they're still an "active user" for our purposes since they're still using their phone and they've still got Pleco on it. So a 2 week cutoff may be a little too strict.

In fact, given the number of Android phones that can't be updated to 2.1/2.2, if anything this statistic suggests that Android users with older phones aren't continuing to buy apps, which is a worrying sign. Either developers are dropping backwards compatibility so quickly that 1.5/1.6 users have given up hope of acquiring interesting new apps, which is a dangerous precedent fragmentation-wise, or Android users lose all enthusiasm for buying apps after the first few weeks of owning their phone.

You can't have it both ways - the non-upgradable 1.5/1.6 phones are still out there, so either the statistic is wrong or 1.5/1.6 users have stopped buying apps.

 

[mikelove]

I just realized that I ignored the possibility that the 1.5/1.6 phones are still out there but that the size of the overall Android installed base is growing very rapidly and most of that growth is coming from new 2.1/2.2 phones, so that even if the number of 1.5/1.6s remains relatively constant, they represent a shrinking percentage of the overall market because they're being overwhelmed by new users.

But in that case it's not really a fair comparison with iPhone because of iPhone's much larger installed base - if half of the iPhones in circulation were iPhone 4s, the iOS 4 adoption rate would be even higher than it is now. You're comparing OS upgraders on iPhone with new purchasers on Android. And of course as Android matures you can expect its growth rate to level out and new phones to represent a smaller portion of the installed base, so this doesn't say much of anything about Android's long-term fragmentation potential.

 

[character]

But then it's rendered meaningless by the fact that iOS 4 came out so much later than Android 2.1 - you're giving Android the edge simply because their release cycle is on a different schedule than Apple's. A better comparison would look at the relative adoption rates of iOS 4 and Android 2.x such-and-such number of months after their release.

I'd agree except for the excruciatingly drawn out upgrade process which varied for each Android handset. It's not like Google held a press conference and "boom" :wink: 2.1 was available for some big subset of users. It's a milestone for Android even if you don't agree with the iOS 4 comparison.

No, but my point was that 2.2's signature feature (its JIT compiler) isn't something developers can assume most users will have, so they need to continue optimizing around older releases.

Yes, but that 'older release' is likely Android 2.1 running on a 600+ MHz handset, not as before Android 1.5/1.6 running on a G1-class device with a ~500 MHz CPU downclocked to ~325 MHz.

I wasn't making a case about update frequency but rather about App Store login frequency - if someone's going 2 months without updating Pleco, then they're clearly not accessing App Store very much but they're still an "active user" for our purposes since they're still using their phone and they've still got Pleco on it.

Given the base Pleco is free, I think it would be more accurate to call them "installers" than users. I (and I suspect a lot of people) will install free apps and not feel any urgency to use them. They looked interesting, and when I have time, I'll try them out.

In fact, given the number of Android phones that can't be updated to 2.1/2.2, if anything this statistic suggests that Android users with older phones aren't continuing to buy apps, which is a worrying sign. Either developers are dropping backwards compatibility so quickly that 1.5/1.6 users have given up hope of acquiring interesting new apps, which is a dangerous precedent fragmentation-wise, or Android users lose all enthusiasm for buying apps after the first few weeks of owning their phone.

Two things are going on: early adopters have upgraded, and there are likely more 2.x devices out now than the earlier non-upgradable 1.x devices. Could there still be a sizable group of people using old phones but not buying apps? Sure. But if they're not buying apps, I don't need to worry about them if I put out something for Android. The sorry state of the Android market is well known, and I'm not trying to argue that issue.

 

[mikelove]

I'd agree except for the excruciatingly drawn out upgrade process which varied for each Android handset. It's not like Google held a press conference and "boom" 2.1 was available for some big subset of users. It's a milestone for Android even if you don't agree with the iOS 4 comparison.

But the increased-sales-volume theory makes more and more sense to me - if most Android users bought their phone within the last 6 months it's only logical that there'd be a lot of 2.1 adoption now.

Yes, but that 'older release' is likely Android 2.1 running on a 600+ MHz handset, not as before Android 1.5/1.6 running on a G1-class device with a ~500 MHz CPU downclocked to ~325 MHz.

The actual speed isn't the issue, the issue is taking advantage of new features, which is why fragmentation is a concern - the sort of new apps that might only be possible with the extra performance afforded by 2.2 can't be developed until there are a critical mass of 2.2 users.

Given the base Pleco is free, I think it would be more accurate to call them "installers" than users. I (and I suspect a lot of people) will install free apps and not feel any urgency to use them. They looked interesting, and when I have time, I'll try them out.

Still potential customers, though.

Two things are going on: early adopters have upgraded, and there are likely more 2.x devices out now than the earlier non-upgradable 1.x devices. Could there still be a sizable group of people using old phones but not buying apps? Sure. But if they're not buying apps, I don't need to worry about them if I put out something for Android. The sorry state of the Android market is well known, and I'm not trying to argue that issue.

But the fact that they're not buying apps is a worrying trend if it carries over to newer OSes as well - more users stop buying apps, fewer users are left to sell to, and Android's impressive market share numbers get less and less interesting from a developer perspective.

 

[mikelove]

And here comes my previously-predicted app store fragmentation:

http://androidandme.com/2010/09/carriers/v-cast-app-store-to-compete-with-official-android-market/

Given Verizon's aforementioned default-search-is-Bing-and-you-can't-change-it deal, it seems quite possible that future low-end Verizon Android phones might only support their app store and nobody else's; would also let them ignore Google's compatibility / trademark requirements for Android Market devices and start to create something approaching their own unique Android-based platform. And it's a great deal for them since it creates carrier lock-in; if you've bought a whole bunch of V-Cast Android apps, maybe even some that are exclusive to V-Cast, you're going to be very hesitant to switch to Sprint / AT&T / T-Mobile for your next Android phone purchase.

 

[ipsi]

Wonderful. Have they said how this will impact international (outside the US) developers? That is, will they be able to submit apps, and if so, how much support will they have when dealing with bugs in the Verizon version of Android? I suspect the fact that I have to ask the second question means the answer to the first is "Hell No", but there we are.

I can't imagine Google being terribly happy with this, but there's bugger-all they can do.

Looking like Symbian, Mk 2…

 

[westmeadboy]

Hmmm, if Verizon continue like this they'll end up with a platform as closed as iOS. But wait, the closed nature of iOS is good, right?

 

[mikelove]

Hmmm, if Verizon continue like this they'll end up with a platform as closed as iOS. But wait, the closed nature of iOS is good, right?

Well most of the posts I've seen defending this Verizon move have suggested that better curation (a non-automated approval process, as Verizon seems to be using) is exactly what Android Market needs, though I think that would work much better / fairer if it came from Google and didn't entirely shut out rejected apps - this system has the potential to exclude small developers altogether, at least in big categories like games.

Honestly, though, I think what the last few months of carrier-screwing-up-Android developments - AT&Ts lockout of non-Market apps, Verizon's forced use of Bing and now their proprietary app store, several brilliant pieces of hardware screwed up by crappy UI shells - have demonstrated is that no mobile platform is "open" as long as the networks it operates on are closed; the "Android ideal" may only be achievable in a world where you buy your own hardware, connect it to the network of your choice and pay a simple flat per-MB rate for the bandwidth you use.

In the meantime, the increasingly-murky sales situation for small Android app developers can only help Windows Phone 7 - even the timing is perfect, Verizon is giving people reason to reconsider investing 3 months in writing their next Android app at exactly the same time that Microsoft's encouraging everybody to download / try out the WP7 developer tools. I wouldn't be surprised if Apple's recent reversal on 3rd-party developer tools / publication of their actually-very-reasonable App Store Review Guidelines was related, too... Google really needs to pull a rabbit out of their hats with Android 3.0 or they risk losing a lot of the momentum they've built up over the last year.

(I continue to feel like Android support is a good move for Pleco, though, since we're in a position to mostly stay above the app store fray assuming most Android phones continue to run non-Market apps)

 

[westmeadboy]

Not wanting to get too off-track but those App Store guidelines included the threat:

"If your app is rejected, we have a Review Board that you can appeal to. If you run to the press and trash us, it never helps."

Pretty chilling when you consider the AppStore is the only channel iOS app devs can distribute their apps through.

Of course, you could just read it literally, in which case it's harmless - but it's pretty clear to mean "if you run to the press and trash us, then the appeal is less likely to succeed". Otherwise, why mention running to the press at all? Are they just trying to save people wasted effort!?

 

[mikelove]

Not wanting to get too off-track but those App Store guidelines included the threat:

"If your app is rejected, we have a Review Board that you can appeal to. If you run to the press and trash us, it never helps."

What do you expect, that they'd encourage people to go to the press? I read that line as simply asking people to wait a bit, cool off, and try to get some clarification / review / a final answer from Apple before going off and whining to Gizmodo about their Clock With Interesting Backgrounds app being rejected. It's chilling if you assume Apple is evil, but if you give them the benefit of the doubt then it's really just encouraging people to act like professionals (or at least like grown-ups), which frankly I think the mobile app developer community in general ought to work a bit harder on.

 

[westmeadboy]

What do you expect, that they'd encourage people to go to the press?

Of course not. I'd expect a company not to mention it at all.

Do you think that if you do run to the press it will go against you or not?

 

[gato]

There have been very few reports of rejections lately. Steve's master plan must be working.

 

[mikelove]

Do you think that if you do run to the press it will go against you or not?

Given how many apps have subsequently been approved after the developer ran to the press, it certainly doesn't seem like it hurts much. And anyway, if the developer's app really was rejected unfairly Apple's going to have a very hard time defending that rejection against public scrutiny.

I suspect it bugs Apple greatly to have their name dragged through the mud just because one overworked App Store reviewer made a bad judgement call on a borderline app, and they'd like at least a second chance to get the app review right before you run off and complain about it publicly. It's not like they can stop you from running to the press if the Review Board rejects you too; in fact, the PR hit in that case would probably be much worse for them, since now the rejection would be as a result of an official, properly-applied Apple policy rather than one reviewer's opinion.

 

[numble]

What do you expect, that they'd encourage people to go to the press?

Of course not. I'd expect a company not to mention it at all.

Do you think that if you do run to the press it will go against you or not?

A lot of people have run to the press and explained how they don't understand why their App was rejected, and it's often resulted in Apple re-reviewing the app and often approving it. I don't know if there are many occasions where developers have gone and trashed Apple when their app was rejected though. It's not a good idea to "trash" a potential business partner anyway. If you trash Wal-Mart/Best Buy for not selling your item, is it likely to get them to treat you kindly? Remember that trashing is different from professionally bringing your concerns or questions to the press.

 

[westmeadboy]

Do you think that if you do run to the press it will go against you or not?

Given how many apps have subsequently been approved after the developer ran to the press, it certainly doesn't seem like it hurts much. And anyway, if the developer's app really was rejected unfairly Apple's going to have a very hard time defending that rejection against public scrutiny.

So, if trashing Apple in the press does not make a difference to the review process, then why mention it at all in the AppStore Review Guidelines?

It sounds like it better belongs to an iOS Developer Behaviour Guidelines document.

 

[mikelove]

So, if trashing Apple in the press does not make a difference to the review process, then why mention it at all in the AppStore Review Guidelines?

It sounds like it better belongs to an iOS Developer Behaviour Guidelines document.

It's a logical place to put that request - people are reading this document because they're concerned their app might be rejected, and Apple's reminding them that there's a process in place for appealing rejections and it isn't necessarily constructive to go running off to the press at the first sign of trouble.

Note that they say "never helps," not "might hurt your case" or anything more threatening like that - they're not warning of dire consequences if you do run to the press, they're just saying that the best way to appeal an initial app rejection is through them.

 

[numble]

Do you think that if you do run to the press it will go against you or not?

Given how many apps have subsequently been approved after the developer ran to the press, it certainly doesn't seem like it hurts much. And anyway, if the developer's app really was rejected unfairly Apple's going to have a very hard time defending that rejection against public scrutiny.

So, if trashing Apple in the press does not make a difference to the review process, then why mention it at all in the AppStore Review Guidelines?

It sounds like it better belongs to an iOS Developer Behaviour Guidelines document.

It also might be American slang, but saying "never helps" in that situation in America can also mean that that's not the best way to go about things.

"If the big kid is bullying you, you need to tell your parents or teachers, complaining to your siblings never helps."

 

[mikelove]

It also might be American slang, but saying "never helps" in that situation in America can also mean that that's not the best way to go about things.

"If the big kid is bullying you, you need to tell your parents or teachers, complaining to your siblings never helps."

This raises an interesting point - perhaps Apple should have worded their guidelines a bit more carefully; they feel somewhat un-Apple-like in general, like they were written by one or two people on the App Review team and weren't carefully run by marketing / legal / international corporate communications / etc.

Sorry you took offense at it, though - I do understand where you're coming from with your uneasiness about this sort of thing, wouldn't be doing an Android version of Pleco otherwise.

 

[yktan]

Any ETA on Pleco for Android?