As the Apple Turns

To clarify on multitasking - the tablet multitasks in the exact same way that the iTouch does: Enabled on specific Apple apps like iTunes, disabled for third party apps. Note that Apple hasn't said how much RAM it has. RAM has been the limiting factor, the reason the iTouch and iPhone don't multitask, so if the iPad has more of it, that hints at future multitasking. It may very well be that Apple skimped on it to concentrate on low latency for a CPU that most likely doesn't have any big caches to keep it fed. Also remember the economics of CPUs: That A4 cost a bunch to develop and tape out, but is cheap now. You will see it in the iPhone for sure, and it was certainly made for it.

A stylus makes some sense, actually - the iTouch you can hold with two hands and type with your thumbs, but I don't think that will work here. That's one of the oversights, in my opinion, but I guess it comes down to the primary usage pattern: To watch video. I'm pretty sure that video is the main reason for this thing, with reading documents as a secondary purpose.

In addition to USB (via adapter), it supports 802.11n. That's plenty fast to transfer files. I agree that it needs an SD slot, though. That's the reason I'm thinking 2.0 on this thing: USB port directly on the tablet, SD slot, multitasking enabled, better battery life.

Regarding the A4 not that we've seen details as yet to it's exact make up we may not find out for sure until someone tears one apart. I do wonder the yield on the fabrication of these would be similar to any other chip like the Intel CPUs right?

So out of a batch that in theory can run at 1.3Ghz (Which i understand is top speed for Cortex 9 reference design) So we might see some thing like
?AppleTV 1.3Ghz (HD 1080p)
iPad 1Ghz
iPod Touch
iPhone

I wonder will we see the A4 in other products?

Like a dud one core version in the iPod nano.
or as co-processor in the mac family. I mean for the laptops it could be a really cheap clean way for Apple to add the Appleness into what would seem otherwise to be really stock products with the i5 or i7 mobile range.
But what about the Xserve, similar to the new cray set up in Each node an ARM processor to handle system messaging and an intel processor to handle the job processing.

I guess the more and varied situation they use the chip will mean more of each batch that gets used spreading the up front cost even further.

We had a discussion very similar to this in another forum the other day... I think that for the Apple TV, the important thing is the GPU. Today I think it still uses a GF7300 chip. Removing that should save quite a bit of money, but that requires the GPU in the A4 to be up to the task. It might be, but we don't really know.

Yields by clockspeed aren't what they used to be. I know that Intel process guys usually say that once their process is mature (a few months in), 95% of chips are top bin - ie, the top clockspeed Intel sells and with all features enabled. Given the 1 GHz figure floating around for the Cortex A9 in general, I guess that that is what it is aimed for. Sure you could have a higher clocked version if you overvolt it, but will Apple do that? Seems likely that they'll just use the same chip, add a coprocessor if they have to and then underclock/undervolt the chips for iPhones. Even better, of course, if they adjust clockspeed on the fly - the best way to save power is to let the device drop into standby as soon as possible, so maybe they'll keep it at 1 GHz as peak but only letting it stay at that speed for seconds at a time.

iPodLounge has a fairly balanced assessment, actually a letter by the editor, of the iPad's good, bad, and ugly:

http://www.ilounge.com/index.php/articl ... d-concerns

and within it he touches upon handwriting recognition, something Apple did well with the Newton and not at all with the iPad:

"* Handwriting Recognition. An OS-level tool to convert writing and scribbling into text and shapes. This was a gimme for the company responsible for Newton, and a big missing piece for users."

Ah... but in order to do handwriting on iPad one would need a writing instrument (stylus) or pen and that too was missing (as indicated and commented upon earlier in thread).

So, in retrospect, in a company as big as Apple, you believe that not a single person mentioned, even in passing, to Steve Jobs that "the old Newton did handwriting" and this hush hush Tablet project might need that?

More likely, Jobs hates pen input and there's a reason no Apple product that currently ships pays more than lip service to OCR.

Would that reason be Apple's 10% market share by chance?

Not completely outlandish here to expect a tablet computer device to handle basic OCR and offer stylus functionality... that is the point that I and others have echoed...don't have to take us to task for expecting it in Apple's newest offering!

I think handwriting recognition is missing precisely because of the Newton. It was mercilessly ridiculed at the launch, and I think Apple wanted to step away from it. Reports from people with SDK access indicate that there are rudiments of handwriting recognition in there. My guess is that it will return eventually in a shape similar to Palm's Graffiti but for complete words.

Note that a plain old plastic stylus won't work with the capacitive (multi-touch) surface - it will need something special. Such stylii exist though, no problem.

Um.....
It is expected to have the same Chinese and Japanese Keyboards as the iPhone.

Maybe the problem is handwritten english?

Handwriting recognition is not the way to go, and I am very very happy Apple didn't include it. It's sloppy because people write so different from each other, and it usually requires an extra bit of hardware which can be lost easily. Handwriting on a computer is so 20th century and just bad as a way to interact with a computer.

The on screen keyboard is the best way to go - simple, exact, no hardware to loose, people already know how to type instead of learning some new palm-like way to make letters... It's simply better in every way.

The OCR and stylus recommendation was linked to the whole idea of handwriting on a tablet as opposed to a laptop; if I wanted to type I would use the utilitarian keyboard of the MacBook, while the open face tablet was for stylus note taking...

Newton had handwriting recognition, and it failed spectacularly. Palm had Graffiti, where you write certain characters that you learn. That worked better (it started as an app to install on the Newton) but it was still slow as you had to wait for each character to be finished, and it was crucial the exact order you drew your character. Something in between these, where you write printed characters but that are recognized as a complete word instead of a single char, is probably a better way to move forward - that or dictating. If you've ever tried Dragon Naturally Speaking, you know that voice recognition has come a long way. Maybe the A4 doesn't have the power, but combine that with something that runs on a DSP and a nice Bluetooth headset, and it should be enough.

Ink handwriting engine in Mac OS X works similar to that, in that it waits for you to complete the whole word before processing. I've tried it several times with a wacom tablet it works ok but seems to fall over on trade words and names. Also with a small tablet it didn't feel so natural writing one word per line. I know some one who uses a Wacom Screen Tablet they find Ink to be a lot more natural. Not only having the feed back but also being able to string a few words at a time.

Odd thing I find not just Ink but across the system. It's great having a central dictionary but why do I still have to train the dictionary with names that are in my address book?
That me would help improve all predictive text entry be it keyboard, pen or speach.
I don't mind writing block letters, after year on a drawing board block letters is my natural hand anyway.

Still Ink on the iPad might be an interesting thing maybe it's just not reliable to be a headline feature like last time they tried handwritting as a feature.

Eat up Martha

Nice!

I can see voice recognition on the iPad before silly handwriting is. I am a much faster typer than I am writer, even with pen and paper!

Noticed Panasonic's new Toughbook H1 Field tablet ("the world's most rugged handheld tablet computer.")... certainly expensive but far more utilitarian than iPad and built to Mil spec... and guess what...it features OCR handwriting analysis and a stylus...

I don't know. Everyone, myself included, thought the original iPod was the dumbest product ever to roll out of Apple way back when. Apple tends to solve people's problems before they know they have them. I'm hoping the iPad does that, and I'm seriously considering one as a replacement for my aging G4 laptop. I just wish they could have found a better name than iPad.

Apple sells Macs, iPads, and iPhones. Say that one a few times fast...

I would much prefer Ink on the iPad, even if I have to buy a 3rd-party stylus to use it. The form factor of the iPad does not lend itself well to speedy typing with the on screen keyboard, since the device is too big to type with your thumbs. The only way to type naturally is to sit on a couch with your legs crossed and the iPad propped against your calf. That's quite limiting. Sure, there's the dockable keyboard option, but that's just more junk to tote around, and that's really only usable on a desk or table.

If they could train ink to read cursive, that would be ideal. I was apparently one of the last generations of schoolchildren to have cursive writing be an important focus, when essays would still be turned in handwritten. I much prefer to write in script, as I can do it with quite a bit of speed. And also my script looks like the Declaration of Independence. It's been described as archaic, and that was meant as a compliment. I love it.

Agree on absurdity of "iPad" name wholeheartedly... further, not sure if anyone can make any kind of legitimate case for an "iPad" as a replacement for a laptop; the ridiculous "16GB", "32GB", or in its largest offering, "64GB" flash drive memory being offered for the "iPad" prevent its use for any serious computing today for anything beyond web browsing. Moreover, I understand traditional computing processing apps are incompatible with "iPad" installation, so as a "replacement" for an iPod the "iPad" may have legitimacy but certainly not as a computing platform replacement surely!

The Windows Tablet has this, too, but consumers don't buy it! Thanks for proving my point!

There are several "Windows Tablets in various stages of development, manufacture, and some are (or have already) hit retail; nonetheless, it is both premature and wrong to suggest that consumers aren't buying them, since - well - they are buying them or plan to.

There have been Windows tablets for quite some time - the entire point of the Atom processor was to make "Mobile Internet Devices" (basically tablets), and the reason to use it over ARM was that it could run Windows. That failed, and Atom was rescued only because Asus figured out that they could make cheap laptops out of it.

This doesn't mean that a tablet with OCR is destined to fail, but it does mean that OCR isn't enough. Is multitouch enough, or maybe multitouch + OCR for text input? Voice control? Don't know, but I do know that OCR alone is not enough.

In related news, Nuance (makers of NaturallySpeaking, the biggest voice recognition software on Windows) yesterday bought MacSpeech (who does the same thing on the Mac, with code licensed from Nuance). Think they might know something? I'm telling you, an iPad with a dictation interface and touch to correct mistakes could be the thing. NaturallySpeaking is impressive. OCR is not, and I've owned two Palms that relied heavily on it.