iPhone 5 + new iPod Touch

[Vzzzbx]

The iPhone 5 seems to have a dual-core 1 GHz ARMv7 processor and 1 GB RAM. If that's a genuine report, this thing should fly.

 

[mikelove]

The iPhone 5 seems to have a dual-core 1 GHz ARMv7 processor and 1 GB RAM. If that's a genuine report, this thing should fly.

Outbenches the quad-core Galaxy S3, apparently, though I'm not sure whether Geekbench is entirely accurate for cross-platform comparisons. (and in real-world usage the more generalized differences between iOS and Android UI lag are likely to matter much more)

 

[numble]

character: The 3rd Gen iPod Touch is still being sold. Just look at the Apple Store online.

That photo you are using is incorrect, the iPhone 5 lowers the screen and decreases the bezel.

Just look at the top selling games on the App Store, see what they support, and you get an idea that companies believe those users are worth supporting (Angry Birds, Where's My Water, Temple Run, etc.)

 

[character]

character: The 3rd Gen iPod Touch is still being sold. Just look at the Apple Store online.

http://store.apple.com/us/browse/home/shop_ipod/compare Don't see it. Do you mean on some other country's store, or in the refurb section?

That photo you are using is incorrect, the iPhone 5 lowers the screen and decreases the bezel.

That photo was a joke.

Just look at the top selling games on the App Store, see what they support, and you get an idea that companies believe those users are worth supporting (Angry Birds, Where's My Water, Temple Run, etc.)

Games (esp. low-cost/freemium ones) are a volume business. While I laud Mike for supporting older devices/OS versions, people have shown they are willing to purchase specific devices to run Pleco and then spend ~$50+ on Pleco itself. In six months to a year, it would be interesting to see the use of Pleco on non-retina devices versus newer devices.

Mike: If you want to boost use of Pleco on the iPhone 5 + new iPod Touch, give users the option of an additional, customizable toolbar instead of increasing the size of the results tableview.

 

[cclaerhout]

The bigger news for us is probably the new iPod, since it looks like they finally added an autofocus lens - after 2 years we finally have a non-phone, non-tablet that can be used for OCR.

Exactly, and that's why I've ordered this new iPod. Oh and by the way a little feed back : two years ago I was the only student to use Pleco in class (Taiwan). Now almost all students i meet are using it. Congratulations !

 

[mikelove]

Games (esp. low-cost/freemium ones) are a volume business. While I laud Mike for supporting older devices/OS versions, people have shown they are willing to purchase specific devices to run Pleco and then spend ~$50+ on Pleco itself. In six months to a year, it would be interesting to see the use of Pleco on non-retina devices versus newer devices.

We're not anymore - the new minor update we just submitted has 4.3 as a minimum because that's the oldest version Apple's new developer tools (required for iPhone 5 support) will build for. And we're actually planning to require 5.1 in the big update, since switching to OS 5 with its systemwide interface customizations is going to let us get rid of a TON of old code we used for user-configurable toolbar colors and such (and virtually every device that supports 4.3 also supports 5.1).

Mike: If you want to boost use of Pleco on the iPhone 5 + new iPod Touch, give users the option of an additional, customizable toolbar instead of increasing the size of the results tableview.

Stacked toolbars at the bottom of the screen don't work - we've tried; it's too hard to gauge your finger's vertical location, the keyboard can deal with that by averaging how far you're shifted from the center of the keys but a toolbar handling a single button press cannot. For the 2.2.12 update all we've really done for the iPhone 5 is widen the list a bit in landscape mode, but as we use it more I'm sure we'll find some other things to do with it - should work very nicely for the document reader, for example, we might be able to have a much better permanently-embedded definition bar.

But the main game-changing thing about the iPhone 5 is processor speed - this is really going to take OCR to a whole new level.

Exactly, and that's why I've ordered this new iPod. Oh and by the way a little feed back : two years ago I was the only student to use Pleco in class (Taiwan). Now almost all students i meet are using it. Congratulations !

That's great, thanks!


More general iPhone 5 note: while most of Pleco works beautifully on it - OCR has never felt more "live" - unfortunately, the handwriting recognizer crashes as soon as you try to draw a stroke. (but since it's only triggered by that, the rest of the app is still quite usable if you just avoid doing anything with handwriting) This actually turned out to be a rather interesting bug - basically the A6 processor (Apple's first time designing their own core rather than using a standard ARM design) doesn't handle old versions of the ARM instruction set quite as well as previous iPhone processors have done - but rather than going in and hacking around with the old recognizer library the easiest fix was simply to replace it with the new one that we were planning to launch with our big update anyway.

Building around that newer ARM architecture required some other changes to our app to deal with some data structure alignment issues, but after a frantic late-night coding session we finished that all up and submitted an update to Apple early this morning, so hopefully within a week or so we'll have a version that works perfectly on the iPhone 5. And, as a bonus, people get that new recognizer ahead of schedule - it's a very nice update, more accurate (larger template database = more ways you can incorrectly write a character and still have it come up - and more refined since they've had a few years of feedback on problem characters like 的), and as a bonus adds support for Unicode Extension A characters (like our good friend 啰) and Japanese "kokuji" (国字 - characters that only appear in Japanese, like 働). The latter aren't part of any specific plans, they were just available and we figured we might as well add them since a few Chinese dictionaries actually seem to cover them.

 

[character]

And we're actually planning to require 5.1 in the big update[...]

That's moving quite quick, assuming the big update comes out this year.

Stacked toolbars at the bottom of the screen don't work [...]

Sorry if 'toolbar' was a loaded term; would be happy to see more buttons as an option near the top of the display.

This actually turned out to be a rather interesting bug - basically the A6 processor (Apple's first time designing their own core rather than using a standard ARM design) doesn't handle old versions of the ARM instruction set quite as well as previous iPhone processors have done [...]

Were you actually writing assembly, or was this an Apple problem (compiler problem, etc.)?

 

[Vzzzbx]

it's too hard to gauge your finger's vertical location

Assuming Apple allows such discussion: is that a hardware limitation? i.e. is the screen not as accurate as it feels?

But the main game-changing thing about the iPhone 5 is processor speed - this is really going to take OCR to a whole new level.

I've got to say, it's spooky how fast everything is, especially combined with LTE. For those of you who've not yet seen it in action on an iPhone 5, the speed of the current Pleco OCR is imperceptible – text is picked up before you've even managed to hold the phone still. Skritter was laggy and jittery on my iPhone 4, but now it's like a hot knife through butter.

after a frantic late-night coding session we finished that all up and submitted an update to Apple early this morning

You are all amazing. Seriously.

And, as a bonus, people get that new recognizer ahead of schedule - it's a very nice update, more accurate (larger template database = more ways you can incorrectly write a character and still have it come up

Wow! It's hard enough to trick the current HWR. Brilliant.

 

[mikelove]

That's moving quite quick, assuming the big update comes out this year.

Skritter's been requiring OS5 since last summer with very few complaints - anyway I believe any device that can run 4.3 can also run 5, and we have to require 4.3 in order to be able to fully support the iPhone 5. We'd require OS5 already, but there isn't really any point since our "old" app doesn't require any of its new APIs.

Sorry if 'toolbar' was a loaded term; would be happy to see more buttons as an option near the top of the display.

Well we already do have two toolbars there - can we really get away with three? It'd get increasingly challenging to hit the search bar in that case.

Were you actually writing assembly, or was this an Apple problem (compiler problem, etc.)?

Not sure, this was actually in licensed code - we licensed a handwriting recognizer for iPhone just after the first beta SDK came out so they may have been using buggy tools or there may have been some old ARM assembly still in there.

Assuming Apple allows such discussion: is that a hardware limitation? i.e. is the screen not as accurate as it feels?

No, it's really just a user problem - people can't judge the vertical position of their fingers relative to the screen that well. Something that seems to be borne out by the design labs of larger companies, since you virtually never see two narrow toolbars at the bottom of the screen - the few widespread uses of stacked buttons we've seen at all have all had huge tap areas for them (e.g. the menu bar on pre-4.0 Android phones).

I've got to say, it's spooky how fast everything is, especially combined with LTE. For those of you who've not yet seen it in action on an iPhone 5, the speed of the current Pleco OCR is imperceptible – text is picked up before you've even managed to hold the phone still. Skritter was laggy and jittery on my iPhone 4, but now it's like a hot knife through butter.

Yeah - this part's fun but then funner still is what happens when developers start trying out things that were previously too slow to be practical

Wow! It's hard enough to trick the current HWR. Brilliant.

Very glad to have that upgrade, yes - actually with the "early access" Classical dictionary Extension-A support is going to have some practical benefits.

 

[Vzzzbx]

No, it's really just a user problem - people can't judge the vertical position of their fingers relative to the screen that well.

Ah right. I'd not considered it before, but I've always had more difficulty triggering vertically close elements than horizontally close ones.

Yeah - this part's fun but then funner still is what happens when developers start trying out things that were previously too slow to be practical

That I definitely can't wait for! I can see the possibilities now that I have the device in my hand.

Incidentally, I think the screen is too long now. With one hand, if I've just tried to hit something at the top of the screen (e.g. a new tweet button), it's a hassle to get back down the bottom in order to type. I need to keep shuffling the phone up and down my palm.

 

[Entropy]

That one I'm not seeing - is anybody really interested in a larger-than-tablet touch monitor? Past that point the most comfortable interaction method is to control what's going on on the screen remotely with a mouse.

That's a question we're going to have a chance to find out when Viewsonic releases its VSD220, a 22" Android 4.0 all-in-one that also serves as an HD monitor. Since this runs Android... we should be able to run Pleco on it! Hurrah!



http://www.viewsoniceurope.com/uk/products/vsd220.htm

Now that I think about it again, it would certainly be easier to enter data on a computer-sized device. This might be a way to get Pleco onto desktops without a huge investment.

 

[mikelove]

That's a question we're going to have a chance to find out when Viewsonic releases its VSD220, a 22" Android 4.0 all-in-one that also serves as an HD monitor. Since this runs Android... we should be able to run Pleco on it! Hurrah!

http://www.viewsoniceurope.com/uk/products/vsd220.htm

Now that I think about it again, it would certainly be easier to enter data on a computer-sized device. This might be a way to get Pleco onto desktops without a huge investment.

Except that to work well we'd have to optimize our UI around it, and devices like this rarely sell well enough to justify doing that. But you never know, if this proves to be a popular concept we'll certainly consider it.

 

[Entropy]

Except that to work well we'd have to optimize our UI around it, and devices like this rarely sell well enough to justify doing that. But you never know, if this proves to be a popular concept we'll certainly consider it.

It's big enough that you could probably just display every button at all times.

 

[cclaerhout]

I've received my new Ipod Touch today. Pleco is working great on it. And the screen extra space makes the handwriting more comfortable.
Well done & thank you !