It’s recently been claimed that the iPad 3 A6 may in fact be dual core (not quad core). While 100% understandable (what apps would take advantage of it?) It’s still obviously disappointing. So, rather than speculate on that, I’m just going to ramble off what I think are realistic but ultimately hopes for the next iPad

• 1080p rear camera (with 1440×1080 stills)
• 720p front camera (960×720 stills)
• “Universal” 3G model (CDMA+GSM)
• Separate Verizon LTE model
• Quad core but behaves like Intel chips with their “boost” performance
• Finally 1024MB RAM BUT apps sandboxed to 256 or 512
• Same relative form, same smart covers (existing ones will work)
• 2048 x 1536 Resolution
• Appropriately beefed up GPU

Some explanations…

The cameras will remain kinda terrible because they’re for FaceTime, not anything else.

The only reason Apple wouldn’t make a universal 3G device is because it gives consumers too much power and Apple doesn’t have the arm twisting power over carriers it used to. LTE still uses to much power (relatively speaking) so don’t hold your breath.

The “boost” feature I’m talking about is napping unused cores and increasing the clock on the remaining one(s). If Apple does give it 1024MB of RAM that would put it at 4x the currently sold iPod Touch and iPhone 3Gs and Original iPad. Even with ARC, this might encourage developers to get lazy so I wouldn’t be surprised if the sandbox limited the amount of RAM apps could see.

2048 x 1536 = 2 x 1024 x 768. In other words, apps can run at 2X with no modification. This is the only way it will happen. I can’t believe we’ve spent over a year listening to morons claim it could be 2560×1600 or that Apple should “bump” it to 1600×1200. That is a lot of pixels to drive. Scaling 480×320 iPhone apps up to that will look even crapper than it does now. Either Apple will finally implement retina support if iPhone only apps (bringing their 960×640 to the same relative crappiness but not worse) or perhaps more drastically, give developers notice that this feature is gone in iPad 3 and will disappear on iPad 1 and 2 with iOS 6.

That’s right. 1%. How long has it been out?

This makes me wonder. Based on how popular iOS betas are among non-developers who shouldn’t be using them, is it possible that iOS 5 topped 1% before going GM or between GM and release? Hilariously enough, the only way to accurately get at this metric would be from Google’s logs.

What exactly do people mean when they say they should bring manufacturing to the US? Do they think the tech industry is like the car industry? Where all you need to do is cut and bend metal and screw/weld it together? As I’m sure has been pointed out many times before, Apple doesn’t make every piece of every products. Displays come from Sharp or LG’s prison/factories. NAND chips come from many different prison/factories. Intel processors come from prison/factories. Macs and even iPads are full of commodity pieces. This was even true when Apple allegedly proudly “made” products in the US (when they were 9 days from bankruptcy according to Steve Jobs).

Go open up any Beige PowerMac and you’ll find parts from Hitachi, Samsung, Sony, Matsushita, etc.

While I’m obviously not advocating Apple’s practices, I’m just saying even if they didn’t do them, they still would. Apple’s not ready to start manufacturing every last component (especially when Intel doesn’t let you) so no matter what some piece of every product will be stained with blood. Apple used to try to make all the parts themselves and it led to sky high prices that almost killed the company. This was not their fault, everyone else was doing it so it created the price disparity in the first place.

I’m going to blame Reagan. Perhaps if he didn’t usher in a culture of outsourcing Intel would be making their processors in the US and there would be hard drives and RAM for Apple to purchase in the US.

So, obvious disclaimer, as a developer, I buy lots of “unnecessary” new devices because it’s a legitimate business expense. But buying so many has given me lots of knowledge for when people ask me when and what to buy. Here’s my advice.

First up, Macs.

Do NOT run things into the ground. I don’t care if it’s backed up. When you wait until you need a $1100 motherboard replacement for a 5 year old iMac to buy a new one you’ve cornered yourself into not having a (decent/practical) machine until the new one arrives. This means you may forgo BTO options for faster shipping times and have to ignore rumors/reports of obvious impending updates. Tragic examples

• Buying any PPC machine after Intel transition was announced but before actuated.
• Buying anything with a CoreDuo (not 64bit) or with an Intel GMA card.

These are cases when you know a much better machine is only one cycle away, but because you only have a brick as an alternative, you’re forced to buy something that’s clearly a “transition” machine that won’t have a long support life.

Another reason not to run a machine into the ground is that you’ll have some sort of Plan B in case your new machine needs warranty service or gets stolen. If your old machine isn’t just a brick, you don’t have to fall victim to buying at the end of product cycles. Older but functioning machines can also be reborn as file servers Home Theatre PCs.

So when should you buy a new machine you “don’t” need? That’s as easy as learning the “obvious” upgrades. Apple (and many others) usually ship some yesterday tech at low prices. Sometimes it’s just a few MHz, some RAM, etc. Sometimes it matters. Sometimes it doesn’t. Here’s today’s list of specs that are more future proof than others.

• 4GB RAM, 8GB supported (even if too expensive right now)
• NVidia and ATI/AMD graphics
• Multicore 64bit processor

Right now the MacBook Airs are the only things failing because of their Intel video card. The reason it matters is because they don’t support OpenCL. Intel says they will eventually, but that’s not good enough for me. There will be a time when apps require OpenCL. They won’t all obviously, but I see no reason why a future Final Cut won’t work on these Airs but WILL work on the previous generations with the nVidia 320m. Depending on the bench, the older ones can be faster.

Personally, I’m waiting for a new Mac Mini. I want a Quad i7 and nVidia graphics. Right now you can get a dual core i7 with ATI or a Quad Core i7 with Intel Graphics. I think I’ll wait rather than get something that might be useless before my current mini. (I’m also hoping for dual thunderbolts on the server model).

Regarding iPhones: That’s easy, always buy every other model at the subsidized price. It doesn’t matter if you’re using Odd or Even devices. You pay the subsidy monthly whether you upgrade or not so you might as well.

iPads? Not enough historical data.

My development cycle usually goes like this: (written during the reign of iOS 5)

1. Write and Finish and app that works well on my iPad 2 and iPhone 4 on iOS 5. At this point, no regard to what features might be iOS 5 only.
2. Check if the app still works on an iPhone 3Gs with iOS 5 (the “worst” possible device)
3. See if it even compiles for iOS 4
4. Test on iPod Touch 4 with iOS 4.3
5. Test on iPhone 3G with iOS 4.1
6. Add armv6 and try my original iPod Touch with iOS 3.1.3

• Universal apps that have @2X artwork are already ready, don’t even need to be recompiled
• UIImageViews scale using bicubic, not “nearest neighbor”, so images on websites won’t be blocky
• Unlike last time, the SDK you have now already supports Retina so things like UIToolBars should already be ready to go (unless you customize with your own graphics)
• If a website already makes @2X calls or uses larger images scaled down, it’s already ready.
• Auto Adjust includes @2X artwork. I bet I won’t have to do an “iPad 3 compatible” upgrade *for Retina purposes. There may be some new feature I want to implement.
• There is exactly one thing apps NEED to do: Provide a 144×144 (72@2X) icon (if Apple doesn’t just grab the 512×512 iTunes copy and scale).

The number of people who say “I want something like an iPad but isn’t made by Apple” isn’t nearly large enough to be served by a few dozen OEMs (now including Dell) to compete over. It’s the myth that demand exists for every piece of garbage, you just need to sell it. You cannot create demand for non-Apple tablets simply by making non-Apple tablets. Demand isn’t there. People want iPads the same way they want iPods.

Ultrabooks have demand, but supply doesn’t exist. Every single one that has come out does not beat Apple on price (and certainly not on anything else). The myth of the Apple Tax erodes quickly when you need to do Apple level engineering.

But Ultrabooks may not have as much demand as we think. When I use my 11″ MacBook Air I’m using a lot of full screen apps and swiping between them with the trackpad. I use the trackpad as a navigation device, not a clicking device. Right now Windows 8 doesn’t have that fluidity. It really is a unique and necessary experience. 1366 x 768 is not a lot of room to work and I’d be lost without spaces. Even using hotkeys (Cmd left/right) in Snow Leopard doesn’t feel good enough. Since I don’t think any ultrabooks have three-finger trackpads, that’s probably the best they can hope for, and it feels so 1990s.

One of the easiest sources for finding the elusive unmet demand is yourself. When app store searches come up dry, when you hate every last TV in best buy, every time you wonder “how could no one make _this_”? Steve Jobs told many stories about going after markets because something good simply didn’t exist. Yes we had “smartphones”, but we hated them. Apple tapped that unmet demand. Sure there were pocketable mp3 players but they only had like 64MB and awkward ways of getting things onto them. 20GB players existed too but they used 2.5″ drives and weren’t pocketable. Again Apple tapped unmet demand.

I’ve done it too. In the early days of the app store there were two types of photography apps: Ones that did it all for you (for better but usually worse) and “portable photoshop” type apps that were clever, but not for the general user. Auto Adjust tapped that unmet in-between category: an app that does it all for you (and well) by moving sliders that you can then move yourself. It also did it in realtime, which back then was unheard of.

That line probably cost Romney the general election because of the ease in which it’s taken out of context. But really he has a point. If you don’t like a service, Capitalism gives you the ability to pick someone else. The only problem is that capitalism in the US has produced a lot of duopolies, triopolies, a certainly cartels.

I’ll try to be more tech than politics here so I’ll pick telcos and smartphones as my example. I have an AT&T iPhone. If I don’t like them I can just fire Apple and AT&T right? Let’s say I switch to Verizon. I’ll get better coverage, but the same bill. If I don’t like the price I can fire Verizon and switch to – oh wait (don’t pretend Sprint or T-Mobile are real carriers). That’s it. I’m “not allowed” have an iPhone without a voice plan, or without a data plan, or without a texting plan.

My theory as to why handset makers (as well as auto makers) do not suffer from duopolies is because international players are a big part of the game. The biggest Android manufacturer is not a US company.

Until about a year ago when FiOS landed in my area, my central jersey apartment had one choice for cable (cablevision). This was not due to a contract with my renter. There simply wasn’t a choice. Firing them would mean living without TV (I choice I decided on eventually) but also without internet. In case you aren’t aware, New Jersey is the most densely populated state. You’d think we’d be crawling with competition for services – there are multiple auto service centers in every zip code after all. Diners too, we love diners.

Gas prices have so many externalities that the differences between oil companies and gas stations have become negligible.

Romney was talking about health insurance which is no different. Few providers are national. Most are regional. Really you have at most 3 choices if you buy it yourself. Your employer usually gives you one choice of provider and maybe two choices for plans (the married and unmarried version). Pre Obamacare, buying your own insurance too often meant being declined for having diabetes or high cholesterol, so really your one choice was to get hired and covered by your companies blanket policy.

“We all like to get rid of our insurance companies.” Yes. A right to healthcare would be great.

Romney is right that choice is good, but only if there’s enough choice to be had. If he wants to salvage this without looking like he’s doing damage control (which is often more damning than whatever went wrong) he should campaign on making capitalism more competitive by eliminating “socialized” corporate welfare *cough* GE *cough* and encouraging small businesses. Maybe appeal to disenfranchised environmentalists (yeah we’re pretty underwhelmed with Obama) by investing in solar companies. Sure the right will scream Solandra but who cares, it’s not like they’re gonna vote for Obama.

From HuffPo: At CES, Microsoft’s Steve Ballmer Strains for Relevance in Keynote

I couldn’t agree more. I’m tired of Ballmer. Read anything form the Building Windows 8 blog and you’ll see Steven Sinofsky is the guy you should be in charge. As such, I recommend you use that as your source for Windows 8 news and just ignore everything Ballmer does.

Also, who decided it was a good idea to have Seacrest there? He’s “cool”. Microsoft isn’t. He could only have made them look foolish, which begs the question, was that someone’s intention?

Newscasters that give “both sides” of climate change (a 99-to-1 consensus) equal coverage are cherry picking.

Newscasters that only interview the craziest OWS participants are cherry picking.

(I have tons more… maybe I’ll add them later)

John Gruber pointing out that Apple leads in profit share of both handsets and apps despite marketshare is not cherry picking. He is correct in his reasoning that this is why I, as a developer, have no interest in expanding to Android. If all of a sudden a report came out that said Windows Phone customers like paying $5.99 per app I’d be all over that (maybe…). Some metrics do in fact matter more than others.