Google Android

[westmeadboy]

The complaints are mostly about providing additional functionality with the extra space and taking advantage of the extra processor speed and graphics processing, and in some cases, storage. If you think Android tablet users won't be complaining to devs about not releasing tablet-optimized software of their phone software on day one, that is just silly.

When did I say that?

I suspect that the majority of apps (non-game apps) do not need to be changed just because a target platform has better processor/graphics capabilities. Most users will be just happier that things like, dictionary lookups are faster.

What we are talking about here is screen hardware configuration and my point is that existing Android apps will generally look good on tablets because most Android devs have designed the UI to run on multiple screen sizes/densities. Many apps will already have a layout dedicated to "large" screens because Android tablet devices have been available since last year.

 

[numble]

The complaints are mostly about providing additional functionality with the extra space and taking advantage of the extra processor speed and graphics processing, and in some cases, storage. If you think Android tablet users won't be complaining to devs about not releasing tablet-optimized software of their phone software on day one, that is just silly.

When did I say that?

I suspect that the majority of apps (non-game apps) do not need to be changed just because a target platform has better processor/graphics capabilities. Most users will be just happier that things like, dictionary lookups are faster.

What we are talking about here is screen hardware configuration and my point is that existing Android apps will generally look good on tablets because most Android devs have designed the UI to run on multiple screen sizes/densities. Many apps will already have a layout dedicated to "large" screens because Android tablet devices have been available since last year.

I was addressing your point that all the iPad optimization requests are because iPhone development is focused on one screen size. I will disagree with your notion that most people are happier with things being faster and that the majority of apps do not need to be changed. Simply go on the iTunes store and take a look at the iPad versions of apps and compare them with their iPhone counterparts. Even something as simple as 1password, which simply stores username passwords, undergoes an overhaul. NYTimes, NPR, IM, RSS reader apps, even the QQ app, all have drastic overhauls as well.

 

[westmeadboy]

I was addressing your point that all the iPad optimization requests are because iPhone development is focused on one screen size. I will disagree with your notion that most people are happier with things being faster and that the majority of apps do not need to be changed. Simply go on the iTunes store and take a look at the iPad versions of apps and compare them with their iPhone counterparts. Even something as simple as 1password, which simply stores username passwords, undergoes an overhaul. NYTimes, NPR, IM, RSS reader apps, even the QQ app, all have drastic overhauls as well.

I said

"This is because iPhone devs designed their apps to run on one screen size."

but I suppose I should have said

"This is partly because iPhone devs designed their apps to run on one screen size."

Note, I didn't say

"This is entirely because iPhone devs designed their apps to run on one screen size."



Again, you are using the iPhone experience to predict the Android one. In the iPhone world you have two devices which are very different from each other. In the Android world, you have a very even spread of devices across the various hardware configurations. So if the gPad was released tomorrow it wouldn't require swathes of developers to overhaul their apps because chances are they already have the code in place that takes care of that configuration.

To put it another way, your arguments are very device-centric whereas Android is not. Suppose some new graphics technology comes out on a bunch of new Android devices. Then an Android dev might consider optimizing for it. Suppose, a bunch of 9inch screen Android tablets flood the market. Then an Android dev might consider optimizing for that.

Let me ask you this:

One dev designs an app for one screen configuration.
Another dev designs an app for multiple screen configurations.
Out comes a new tablet and neither dev has a chance to change his app.

Which app would you guess would be better on the tablet?

 

[numble]

One dev designs an app for one screen configuration.
Another dev designs an app for multiple screen configurations.
Out comes a new tablet and neither dev has a chance to change his app.

To put it another way, your arguments are very device-centric whereas Android is not. Suppose some new graphics technology comes out on a bunch of new Android devices. Then an Android dev might consider optimizing for it. Suppose, a bunch of 9inch screen Android tablets flood the market. Then an Android dev might consider optimizing for that.

Which app would you guess would run better on the tablet?

The second dev, but that is not what I am arguing. Have you taken a look at the iPad apps on the iTunes store, particularly the ones most popular with users? They are apps that are an overhaul in design and functionality, not simply a readjusting of page layouts. It still doesn't get to the issue of different hardware configurations. The Notion Ink Adam sports a behind the screen trackpad, the Nook and Alex have mini-touch screens and a reading screen that users cannot touch, some tablets have styluses, etc. users of each will ask for optimization to their specific hardware profiles, even for something as simple as a dictionary app. And while the gaming argument is conceded, its still a big argument for marketshare. PC gaming has practically died because of the need of devs and users to keep track of multiple hardware and software parts--even though a more powerful version of Crysis 2 can be played more cheaply on the PC (the game will also probably be $10 cheaper), the PS3 version of the game will probably sell more.

 

[westmeadboy]

Have you taken a look at the iPad apps on the iTunes store, particularly the ones most popular with users? They are apps that are an overhaul in design and functionality, not simply a readjusting of page layouts.

We seem to be going round in circles here.

I suspect the iPad optimized apps are doing well because the ones that are not optimized work so badly (because they were never written with that hardware configuration in mind).

Here's something I wrote earlier:

Again, you are using the iPhone experience to predict the Android one. In the iPhone world you have two devices which are very different from each other. In the Android world, you have a very even spread of devices across the various hardware configurations. So if the gPad was released tomorrow it wouldn't require swathes of developers to overhaul their apps because chances are they already have the code in place that takes care of that configuration.

This doesn't just apply to screen configuration, it applies to other hardware configurations too.

Here's an analogy:

Suppose you have a shop that only sells apples and then you want to also sell maxi pads, then there are going to be lots of problems introducing such a drastically different product line.
Now suppose you have a supermarket that sells all types of fruit, veg and toiletries, then you can much more smoothly introduce some new line of sanitary towel

Maybe Mike could chime in here. What percentage of Pleco's iPad-specific code (code that would never be used in the iPhone version) is not screen-configuration-related?

 

[numble]

Here's an analogy:

Suppose you have a shop that only sells apples and then you want to also sell maxi pads, then there are going to be lots of problems introducing such a drastically different product line.
Now suppose you have a supermarket that sells all types of fruit, veg and toiletries, then you can much more smoothly introduce some new line of sanitary towel

Here's a comparison. Is it easier to go from designing for the iPod Touch (iPhone OS) to designing for the iPhone 3G/3GS and iPad (iPhone OS) or is it easier to go from designing for the G1 (Android) to designing for the Nook, Alex, and Adam (Android)?

 

[westmeadboy]

Here's a comparison. Is it easier to go from designing for the iPod Touch (iPhone OS) to designing for the iPhone 3G/3GS and iPad (iPhone OS) or is it easier to go from designing for the G1 (Android) to designing for the Nook, Alex, and Adam (Android)?

And that's the whole point, an Android dev would never have designed specifically for the G1 (or any other device for that matter).

 

[numble]

Here's a comparison. Is it easier to go from designing for the iPod Touch (iPhone OS) to designing for the iPhone 3G/3GS and iPad (iPhone OS) or is it easier to go from designing for the G1 (Android) to designing for the Nook, Alex, and Adam (Android)?

And that's the whole point, an Android dev would never have designed specifically for the G1.

Ignoring the point that Mike has said he'd only design/test a putative Android Pleco on just one or two phone models, I'll say that even if that's the case, Apps that run nicely on the G1 will not run nicely on the Nook without much more extensive optimization then is required to port from iPhone to iPad.

 

[westmeadboy]

Ignoring the point that Mike has said he'd only design/test a putative Android Pleco on just one or two phone models, I'll say that even if that's the case, Apps that run nicely on the G1 will not run nicely on the Nook without much more extensive optimization then is required to port from iPhone to iPad.

An app may look and behave completely different depending on the hardware configuration. It just depends how its configured. So just because an app is great/bad on a G1, says nothing about how it will be on a tablet.

 

[mikelove]

character - OK, understood. Thanks for keeping it up for as long as you did. Though I hope you'll keep posting suggestions about iPhone at least

Everybody seems to be on the cusp of releasing an Android tablet. And they're going to be available in a vast and confusing array of sizes, screen types (e-ink versus LCD), with different input options (some may have keyboards and almost resemble laptops), pricing plans (subsidized on a carrier contract), etc. But the problem there is that consumers, developers, even tablet manufacturers don't exactly know yet what people are going to want to do with them - SNL had a joke on Saturday that the iPad's success was ushering in a new era of "people buying things to find out what they are." Google has shown with the Nexus One that devices aren't guaranteed to succeed simply by virtue of having the Google name on them - I have yet to see a single Nexus One "in the wild" and I've seen lots of other Android phones - so the idea that Google's tablet is going to introduce some unity to this mess is unrealistic.

Apple is now offering a clear vision of what tablet hardware is going to look like in their corner of the universe; they may still not know exactly what sorts of software is going to run on them, but at least the people developing that software have one specific form factor and set of capabilities they can start planning around. One device that's guaranteed to be successful enough that people can rely on its continued existence for at least the next few years. Schools all over the US are already evaluating iPads for classroom use, accessory designers are hard at work coming up with every conceivable kind of cradle / dock / keyboard / sweater cozy / etc for them, even the US military is supposedly investigating iPads for things like interactive maps / training software / keeping vast piles of documents ready at hand. Having the hardware question settled gives everyone something to focus on and plan around.

westmeadboy - nice to see you back in the thick of this again The screen-scaling thing has never been one of your better arguments, though, for the very reason that numble points out - there's a big difference between resizing an interface for a tablet (which happened automatically the first time we ran Pleco on an iPad after setting the "supports iPad" property flag - i.e., told it it shouldn't just run the thing in an iPhone simulator) and adding new UI controls etc to take advantage of the extra space.

As far as iPhone versus Android, I don't know if I'd characterize this as really being that kind of thread even now - it's more that it's turned from a "nice things about an Android version" to a "pros and cons of an Android version" thread. Many of those pros and cons are being presented specifically in relation to iPhone - Android does this thing which iPhone supposedly doesn't do - but that's logical, since the alternative to Pleco doing an Android version is Pleco not doing an Android version and remaining (at least on mobiles) an iPhone-centric shop for now. And the cons are very necessary - this sort of discussion isn't useful at all if it's just you and a couple of other people talking about how darn great Android is and me politely putting you off for half a decade.

I think, though, that the technical arguments either way are unconvincing enough that this is going to have to boil down to customers - the financial benefits of an Android port and the satisfaction of seeing more people using Pleco have to be weighed against their equivalents with a desktop version, and both of them weighed against continuing to throw everything at iPhone / iPad. Further development of our iPhone / iPad software could very well get us into some brand new types of markets - school site licenses, e.g. - so it's in no way certain that time and money spent on Android or on desktops wouldn't actually generate more new customers if they were spent on iPhone, even if Android eventually overtakes iPhone in market share. (which I've never found a compelling statistic anyway - Symbian says hello)

 

[westmeadboy]

westmeadboy - nice to see you back in the thick of this again The screen-scaling thing has never been one of your better arguments, though, for the very reason that numble points out - there's a big difference between resizing an interface for a tablet (which happened automatically the first time we ran Pleco on an iPad after setting the "supports iPad" property flag - i.e., told it it shouldn't just run the thing in an iPhone simulator) and adding new UI controls etc to take advantage of the extra space.

Hmmm, it seems I've done a bad job of explaining what you call "screen-scaling" because that was never my argument. I've never said that auto-scaling is good enough so I don't know why people think I did. I can only assume it comes from false assumptions about how Android supports multiple hardware configurations.

Android devs are already designing their apps with large-screens in mind (because the possibility has been there for more than one year), so if a gPad comes along, then their app may already have an optimization for that general screen size (AND I'M NOT TALKING ABOUT SCREEN SCALING!).

Here's another attempt to explain it (to show its not my argument thats bad, just your interpretation of it )...

I can think of three types of screen-configuration logic:

1. Scale (as if you were in photoshop and resized)
2. Intelligent scale - where you define the same layout for all configurations such that, for example, content panes will expand accordingly, but controls will retain their physical (dpi) size. This is a bit like when you resize your browser window. Side note - woah, how do all those websites survive all those on-the-fly screen configuration changes!?
3. Screen-configuration specific layouts - the app should look completely different on a large-screen TV to a watch phone.

Now, I'm not sure which the original iPhone apps fits into. I asked once before and the answer seemed to be (2). But maybe its (1)?
(3) is for when the iPad came along.

Here's how those scenarios play out in the Android world:

1. Very old apps (pre 1.5) or very badly written apps (less than 10% of Market I guess) will result in this type of scaling
2. Most simple apps will use this approach
3. More complex apps will have completely different layouts for larger/hdpi screens. This might mean extra panels of buttons etc that you would not see on devices like the HTC Tattoo (small screen).

I've been talking about (3) for the whole of today's debate.

My point is this. Many Android devs have already taken tablet-type hardware configurations into account and so if a gPad was released, it would not be much work (if any) to have the app working to a good standard on that device. Granted, you would need to put more work in at the beginning to make it look good on a variety of configurations, but as you are seeing with the iPhone (with the promise of one configuration for all) that scenario is inevitable.

If you have built your app from the ground up with that in mind, then you will win in the long run and you won't piss off so many of your users when the next latest and greatest device comes out.

Mike, what percentage of Pleco's iPad-specific code (code that would never be used in the iPhone version) is not screen-configuration-related?

 

[mikelove]

iPhone apps fit into category 2 - you tell the system how to scale or resize your controls when the screen size changes (and in which direction / axis / etc) and it dutifully does so. Developers have to support that even in portrait-orientation-only apps because the iPhone's in-call status bar lops off a good 50 pixels or so on the top of the screen. I wouldn't be surprised if a few Android developers started planning for tablets in advance, but I doubt there are many apps in Android Market that already have largescreen-tablet-specific UI features built in / ready to go the second someone sticks them on a tablet.

So it seems there isn't really much of a debate here - neither platform can magically conjure up large-screen-optimized apps, developers have to redesign around large screens to make their apps work well in both cases. You can make a reasonable argument for Apple's approach or Google's approach to non-optimized-app screen scaling on tablets; Apple wants to segregate iPad-optimized from non-iPad-optimized apps, Google wants everybody using the same Android software. Apple's approach on this is actually philosophically consistent with their ban on recompiled Flash apps - they've got plenty of developers already, they want developers who are willing to take the time to really make their apps sing on Apple's platforms. (the product of decades of looking at ugly, 18-months-delayed Mac OS ports of popular PC apps)

In terms of lines of code, maybe 3/4 of the iPad-specific stuff would never be used on iPhone, but development-time-wise, roughly half of the (few days of) iPad-specific development time went to code that could be and in fact will be used on iPhone too, much of it to adding the ability to keep a definition on the screen during a search. (something that a few people had requested even on iPhone but that there hadn't been reason to prioritize / finish ASAP before iPad)

 

[westmeadboy]

In terms of lines of code, maybe 3/4 of the iPad-specific stuff would never be used on iPhone, but development-time-wise, roughly half of the (few days of) iPad-specific development time went to code that could be and in fact will be used on iPhone too, much of it to adding the ability to keep a definition on the screen during a search. (something that a few people had requested even on iPhone but that there hadn't been reason to prioritize / finish ASAP before iPad)

I was thinking specifically of screen-configuration dependent code. numble seemed to be implying that you might have special code to take advantage of the faster processor etc...

 

[numble]

In terms of lines of code, maybe 3/4 of the iPad-specific stuff would never be used on iPhone, but development-time-wise, roughly half of the (few days of) iPad-specific development time went to code that could be and in fact will be used on iPhone too, much of it to adding the ability to keep a definition on the screen during a search. (something that a few people had requested even on iPhone but that there hadn't been reason to prioritize / finish ASAP before iPad)

I was thinking specifically of screen-configuration dependent code. numble seemed to be implying that you might have special code to take advantage of the faster processor etc...

Hey, my comments have always been general about development in general, even speaking just of dictionary apps and not Pleco, excepting the fact that I said Mike said he'd only design/test a putative Android Pleco on one or two models. I make no assumptions about Pleco's code, requirements, or the sexes of its userbase.

I again point you to compare QQ, New York Times, NPR, IMplus, 1password etc. on the iPhone and the iPad--you can just look at the screenshots to see what I'm talking about. In general, users are not simply demanding better screen reconfiguration.

 

[westmeadboy]

I again point you to compare QQ, New York Times, NPR, IMplus, 1password etc. on the iPhone and the iPad--you can just look at the screenshots to see what I'm talking about. In general, users are not simply demanding better screen reconfiguration.

The fact that you are pointing me to screenshots indicates something graphic in nature

 

[Zeldor]

My main point was that it doesn't really matter what is the 2nd platform, but that Pleco should be available for iPhone/Pad + SOMETHING. Rest of my arguments was that Android makes the best 'something' right now.

I don't even have any idea about when iPad will be possible in many countries. And how much it will cost. It's mostly US thing to heavily subsidise phones. And having tablets/netbooks subsidised in Europe is very rare thing. So price difference between Android tablets and iPad may be quite important here. I don't think you should downplay foreign market, there is much smaller access to any Chinese learning tools than in US. But well, my vote would not count anyway, as I'm not paying customer. I'm not sure if I will ever become one if Pleco stays iPhone only [unless there comes really cool next-gen iPad with cloud-storage working in Europe + China [firewall?] or SD slot + usb port + camera + Linux/Windows emulation for some apps I'd really need and that are not on iPhone].

 

[numble]

I again point you to compare QQ, New York Times, NPR, IMplus, 1password etc. on the iPhone and the iPad--you can just look at the screenshots to see what I'm talking about. In general, users are not simply demanding better screen reconfiguration.

The fact that you are pointing me to screenshots indicates something graphic in nature

You can deduce elements of features and functionality from screenshots. Feel free to get 10 youtube videos to compare the functionality of each. Can you point me to some of the "many Android apps" that are ready for Android tablets, that you keep talking about? If the differences in features and functionality are as drastic as the differences between the 5 apps (iPhone vs. iPad) I listed above, I'll concede my point.

 

[radioman]

One of my reasons for wanting an iPad and will likely get the one with 3G is that it is NOT tied to a carrier and is NOT subsidized. I'll throw Skype on it and call it a day. (oh, and of course... Pleco...)

\
I don't even have any idea about when iPad will be possible in many countries. And how much it will cost. It's mostly US thing to heavily subsidise phones. And having tablets/netbooks subsidised in Europe is very rare thing. So price difference between Android tablets and iPad may be quite important here. .

 

[Zeldor]

Intel porting Android for Atom: http://www.slashgear.com/intel-port-android-to-atom-in-attempt-to-undercut-arm-1381500/

But also Flash hack for iPad: http://www.slashgear.com/ripcode-allows-ipad-to-play-flash-video-1381485/

 

[mikelove]

Android for Atom doesn't seem like a big deal to me; Intel's worried they're going to lose the very lucrative business they've built selling netbook CPUs and chipsets, and probably rightly so. (they used to be leaders in the ARM field too, but they aren't anymore) Generally I'd say it's only interesting for hardware and specifically embedded system developers; a dual-boot Android / Windows system might be nice for a high-end set-top box or other media-playback-heavy gadget, using Android as a low-footprint instant-booting OS and Windows to run regular PC software.

Aside from dual-booting Windows, there's not really any advantage I can think of to having an Atom-powered Android device over a device with a comparable ARM chip; performance will probably be better on ARM, the RISC versus CISC debate faded once Intel started recompiling X86 instructions dynamically on their chips (can't remember if they do that on Atom or not, but they definitely do in modern Core etc processors) but you've still got a certain amount of extra baggage on their chips versus ARM ones.

And of course an Atom-powered Android device wouldn't be able to run any Android software that uses the NDK, at least not unless Google decided to get behind the idea officially enough to start supporting multi-architecture "fat binaries" / cross-compilation / etc, which seems unlikely since it would be increasing platform fragmentation for no reason other than to help Intel.