A few days before the MacBook Air announcement, references to a MacBook Air appeared in Adium's server logs.

http://arstechnica.com/apple/news/2008/01/macworld-ars-macbo...

Given how Apple is so notably obsessed with security around new products, it certainly seems plausible that if they wanted to hide it, they would figure out how and make it happen. So I wonder what the probability is of a controlled leak.

And this info would be the perfect thing to leak, because it doesn't tip their hand to show anything we don't already know. It's already known that a new iPad will be announced on Wednesday. 2x resolution LCD panels have surfaced in China, so we already know that. And of course Apple is working on iOS 6, because they make iOS.

I think you might be ascribing too much to the marketing types. Regardless of the company, a "controlled leak" would probably be someone from PR (credible) talking to a single (or few) outlets. They almost certainly have no idea of (a) what a user-agent string is, (b) that it is logged to servers, and (c) that high-credibility news outlets are monitoring user-agent strings.

I'm putting my money on gaffe.

Considering how obsessed with image, marketing, etc, Apple is, I would be extremely surprised if the Marketing department weren't hugely competent in this matter, including knowing about user-agent strings, logs, etc.

Ask yourself: would Steve Jobs have known about this? Yes, of course. So, why wouldn't the person he hired? This is not a clueless marketing department we're talking about, here... it's probably the most competent marketing departments of any technology company in the world.

Or at this point, they simply don't care, given that:

1. an announcement has been made with a photo hinting at a Retina display

2. it is obvious that Apple is working on iOS 6.0

No one said it was a controlled leak from a marketing department. It could have been managers (at any department, or multiple departments) issuing a statement that the new development software/hardware can be used on the public Internet without worry. I'd imagine the executives or directors would know (or be told) what could happen, and they were comfortable with it.

At that point, it's not a release of information and it's not a leak, but rather a controlled leak.

Occam's razor

Basically what's happened here and elsewhere is Apple's profile is so big now and there is so much money to be made writing and breaking scoops about them that everything is under much more intense scrutiny then before.

Apple is like the country drifter who's been able to avoid the Barney Fifes at the local sheriffs offices but get's caught once the FBI forensic team is brought in.

The expectation I guess is more secrecy. When they're internally testing iOS 7 they sure as hell won't call it iOS 7 before the announcement. They may approach a number of these sites and offer payments for NDA, (this information is presumably more valuable to Apple then it is to Ars so they can just pay Ars). More intentional misdirection.

Meh. It happened before. Apple obviously doesn’t care at all.

in both cases, the products didn't start showing up in logs until after the announcement of the announcement. the cat is out of the bag. there's no doubt at this point that there's going to be a new high-res iPad announced on march 7. it's not so much a controlled leak as it is apple just doesn't care anymore.

I just sent ARS a ping with iOS 6.1 at 8192x6144 resolution. Lets see if they write another article.

from inside the apple corporate IP block?

is that impossible to work for Apple in California and bring a device with webbrowser that imitates this:

Mozilla/5.0 (iPad; CPU OS 7_0 like Max os X) AppleWebKit/546.1 (KHTML, like Gecko); 4096x3072

Also, I can specify whatever the hell source IP address I want on my packets.

Unless Ars have started serving HTTP via UDP, establishing a TCP/IP connection over the Internet with a fake IP address will be be near impossible.

Depends if they care about X-Forwarded-For, or not.

Hm. Don't Retina iPhones actually report their resolution as the same as non-Retina ones? When you're designing web sites etc. you can use the exact same pixel values in CSS, and the iPhone does the rest.

Maybe Google Analytics handles this by multiplying screen dimensions by window.devicePixelRatio. Or maybe someone is playing with Ars.

Yes, retina iPhones report their resolution as 320x480. And it's so ridiculously easy to forge a user agent. I don't understand why this is even news.

But they specifically reported in the article that those user agents were coming from Apple's IP block.

Pixel measurements in CSS are not actual physical pixels. I think that's rather silly, but on the other hand what can you except from web tech....

They are pixels at 100%, but a lot of mobile devices have a default viewport width making this meaningless.

Nope. CSS 2.1 specification says:

* pt: points — the points used by CSS are equal to 1/72nd of 1in.

* px: pixel units — 1px is equal to 0.75pt.

But in practice, doesn't every major non-mobile browser use 1 px = 1 pixel?

That being the only sane option, I sincerely hope so.

Years of trying to explain to designers from a print background that it's much easier to think in pixels than doing a mental transformation from mm to dpi to screen and now the whole world goes crazy.

Most desktop browser default to 96 PPI (which is also the default in Windows), which results to 1px being 1 pixel. 0.75 pt/px * 1/72 in/pt * 96 pixels/in = 1 pixel/px.

I was curious how publishing articles naming companies and user agents was allowed within their privacy policy. Relevant section:

We reserve the right to use the information we collect about your computer, mobile or other device (including its geographic location), which may at times be able to identify you, for any lawful business purpose

To be honest, I'm comfortable with websites publishing User Agent data. If I changed my UA to IE 5.5 for Mac and some sysadmin did a double-take and published it on her site, I wouldn't care as long as she didn't give away any personally-identifiable information like an IP address.

Sure this counts as something that would be a lawful business purpose.

Given the fact that this has happened repeatedly with the launches of past Apple products, I find it difficult to believe that Apple would have made this "mistake" again unless:

a) They want news about the iPad 3 to leak out

or:

b) It's faked, something which, as several posters here have demonstrated, is trivial to accomplish.

I think b) is the most likely answer.

If I remember well, last year we stumbled upon a similar case. We'll know after the announcement. As far as I know there have not been any iOS 6.0 rumors, so this is a high bet.

If 6.0 is unvelied next week, then Apple is doing that on purpose, as they know Ars checks their logs for that kind of content. However, if they stick to 5.1, then we'll know it's some kind of prank, and Ars will probably think twice before publishing the same article next year

You think b) even though Apple printed and sent out hundreds of invitations featuring prominently an iPad3 with a clearly visible retina display?

or:

c) Of the few hundred employees with access to iPad 3s, at least one slipped up and visited public websites with it.

> Given the fact that this has happened repeatedly with the launches of past Apple products, I find it difficult to believe that Apple would have made this "mistake" again

This is the part that excludes (c). If they didn't want to leak, they would've configured the prototype units to report themselves as normal iPad 2s.

Why is this even news? Don't we already have multiple corroborated sources verifying the existence of the Retina display iPad3? And what's the other takeaway - that iOS6 will eventually be released?

Wouldn't the opposite of this information be what's newsworthy?

None of the sources is what I would consider reliable. Multiple unreliable sources doesn't improve the reliability that much unless they are reporting independently.

The iPad on Apple's invitation to March 7 event[1] clearly sports a retina display. Compare it to iPad tv ads[2] and you'll see the difference between the displays...

[1] http://www.loopinsight.com/2012/02/28/apple-announces-ipad-e...

[2] http://www.apple.com/ipad/videos/

This is news because no one really knows what Apple will release and not release.The rumors and chatter before the iPhone 4S indicated an iPhone 5 with a new design but that didn't pan out. That is the reason this is newsworthy and this is the reason for Apple's secrecy, to increase the speculation in the media.

curl -A "Mozilla/5.0 (iPad; CPI OS 7_0 like Mac OS X) AppleWebKit/535.8 (KHTML, like Gecko)" http://arstechnica.com/apple/news/2012/03/ipads-using-ios-6-...

Now Ars has iPad with iOS 7 user-agents in their logs too :)

from inside the apple corporate IP block?

thats iOS 7 running Safari 535.8 from iOS 6

Touche.

The iOS situation is certainly weird enough to make this seem plausible. 5.0.1 was hastily pushed out, but did not fix some folks' battery problems; then 5.1 was announced with Photo Stream deletions and an option to disable 3G on the iPhone 4S, yet nothing was released, not even the SDK has been updated lately. 5.0.1 is also by no means bug-free on my devices. What's going on there?

Doesn't this sound a bit like the situation just before Apple announced Mountain Lion out of nowhere?

But then, why would Apple let it leak via the user agent - because this is not considered big news? But why not merge it into the March 7th event?

I hope this is a hoax and we can all enjoy a super-mature iOS 5.1 next week.

They're probably adding a few new features to iOS 5.x to launch with the iPad 3. In years past they showed off the next major version in the March/April time-frame but this was when the iPhone was launching in June/July. With an iPhone release later in the year they don't really have to show off iOS6 until WWDC.

I don't think there's anything plausible about inferring anything from a UA string, but you make a good point here. 5.1 should come on Wednesday, hopefully for all current iOS devices and not just iPad 3.

I can't wait to hear from the PC laptop manufacturers after the iPad 3 is released. How can Apple release a 2048x1536 (10 inch!) iPad when laptop OEMs have been claiming for years that 4:3 screens are DEAD? And that 768 vertical pixels is about the max we'll ever get on smaller screens?

ThinkPad: "Since these manufacturers make more selling TV displays than laptop displays, the PC vendors have almost zero say in this change. We simply have to adapt. As much as I would like it to be so, 4:3 is not coming back." http://blog.lenovo.com/perspectives/display-ratio-change-aga...

They probably do have zero say if they don’t commit to buying ten million or so per quarter.

I don’t think you quite understand how big that difference is. Lenovo sells half as many laptops as Apple sells iPads per quarter. That’s not laptops with one screen size and resolution, that’s six million laptops with wildly varying screen sizes and resolutions.

Apple has one iPad with one model of screen†. They sell millions per quarter and consequently need millions of identical screens per quarter. They have billions of Dollars with which they can pay upfront. Apple can pay LG one billion Dollars per quarter (that’s about as much as ten million iPad screens cost) upfront and wouldn’t even notice it all that much. Lenovo has less revenue, a lot less profit, a lot less cash reserves than Apple.

Lenovo sells many different laptops with many different screen sizes and resolutions. Many people probably do want 16:10 or 16:9 screens. They could probably sell substantially less than one million laptops with a 4:3 screen to a few weird nerds – and they would probably need 4:3 screens in different sizes and resolutions. That is certainly harder than what Apple has to do.

Apple can because Apple is unique. Their position allows them to do stuff like that.

—

† After next Wednesday they might have two.

>"Lenovo sells many different laptops with many different screen sizes and resolutions - and they would probably need 4:3 screens in different sizes and resolutions.

My original comment was more criticizing OEMs for what you stated in that quote above, rather than actual outrage that Lenovo wasn't bankrupting themselves by pursuing 4:3 screens regardless of cost.

Why do they need so many models and screen sizes? Why are all the models impossible to differentiate and difficult to buy?

If Lenovo went "Apple-like" with ThinkPads, cut down their range to a 15" T series and a 12" X series, maybe we'd still have 4:3 screens.

http://arstechnica.com/hardware/news/2011/09/ultrabook-intel...

I totally agree that it's awful losing 4:3 - and flat panel monitors have actually gone down in resolution since 1998.

Well, but there's the rub: you definitely can get high resolution flat panel monitors today. You just have to buy a huge monitor to get that resolution.

So it seems obvious to me that Apple's not going to increase the size of the iPad. ( http://www.tomsguide.com/us/iPad-iMat-iBoard-Joke,news-5968.... )

The rest of us can fight over the scraps. Apple is big enough to dictate the design specs on their displays.

Because Apple has a billion or so dollars to throw away for getting a factory started that will build the said screens for them. And that all PC Manufacturers are fighting a dog fight for the razor thin margins at end of sale spectrum.

Could be a great prank, responding to the JS required to detect the resolution and sending a fake header are all easy enough, right? :D

They did say that the iOS 6 requests were coming from Apple-owned IP addresses.

The screen resolution was not tied to the IP addresses because it showed up in Google Analytics.

Apple employees are capable of pranks.

Hypothetically, say I'm an Apple employee who plays a prank on a tech website whereby I raise expectations for the product we launch in a week above reality.

How long should I expect to keep my job for?

Depends who told you to do it and who set up your device.

If this is a prank it's not a very good one:

1. If there is a retina-display iPad 3 then they're faking something that actually exists.

2. If there isn't then they're setting up the unreleased product to not meet expectations.

This is a great idea for a free web service.

Select/Insert user-agent and use cloud farms to hit a website randomly with it.

It is very easy, yes. As is faking the user agent.

Yep, it is advised to never trust client-side input for a reason.

It could be Apple is testing the new shipment of iPad 3 with iOS 6. iOS 6 is scheduled at WWDC in June, my guess is iOS 6 will be a refinement of iOS 5 just like Mountain Lion is a refinement of Lion. And there will be a few new things in iOS 6/Mountain Lion not reveal to public yet. Question is will Apple surprise all by giving us a preview of iOS 6 next week?

My conspiracy theory: Apple is pushing yearly iOS update to put further pressure on Android. Apple will be glad that Android 5 will launch later this year, at a time when Android partners are just starting to roll out Android 4 hardware.

Perhaps the beta iPad3 users are using it to track iPad3 related news on Ars?

I just checked my Google Analytics. I have (in my personal blog) 1 visit with this screen resolution, browser Internet Explorer. Then I checked one of my job's biggest sites (Apple related), 6 visits, Opera and IE. Checked a smaller site... Added Chrome to the mix.

TL;DR: The screen resolution argument is quite crapish.

There are a few 2048x1536 monitors. But that's not the only line of evidence the article's conclusions hinge on. There's also the OS, the version, and the source of the connections from apple's corporate IP block. Given all that, and given what we already know about the iPad 3 (it will have a higher resolution display) Occam's razor leads to the conclusion that these are pre-release iPad 3 devices. The 2nd most likely conclusion is that this is an intentional hoax on the part of someone within Apple, which is hardly a more credible idea.

As I said below, all together makes sense for the iPad3 testing. But they say that resolution is very odd, and I don't agree (like you said). I'm not denying all the article, just their resolution argument.

Same here - 12 visits with Chrome for Windows last month with this resolution

Btw, checked and found some "Safaries" - 535.8+ with a screen resolution of 1024x600 (operating system version not set, some with Flash support...)

My beta-channel Chrome is 535.19 today, so i don't think that's particularly unusual.

(Assuming Webkit versioning is Major.Minor and not decimal)

My wife's computer LCD monitor is a Samsung that uses this resolution natively (I think), so maybe it isn't that rare.

If the visits are from Apple's block of IPs, the argument is not really crapish. It can be that Apple employees are trolling Ars, but there's a better chance they are testing pre-release devices.

They affirm there are very few devices with this resolution. My GA data (and others) disagrees with this. As for the IP range, I think this makes it more legit. But I'd love Apple employees to troll Ars :)

>346 visits from a device with a screen resolution of 2048x1536

That's pretty weak, was that same device one that was providing a iOS 6 user agent? They don't appear to link those two together, so eh. I can see one Ars user who has an old CRT (like I had 10 years ago) running 2048x1536.

If this is next week's iPad, wouldn't we expect both landscape and portrait resolutions? If I were in the rumor-making business, in combination with that iOS 6 version, I would call this a new apple TV running firmware claiming to be an iPad.