It has been apparent for a long time that JavaScript is not a foundation we can grow forever on. The current ubiquity and reliance on JS is an accident of history, like so many ‘standards’. Efforts have existed for awhile to extend beyond the limitations of JS, Coffeescript being a better known example. The number of projects that have been created to compile other languages to JS² strongly signals the need for something better.

That something better is emphatically not Dart. Nor is it NaCl. Both of those projects stem from noble intent but are tragically misguided and interesting evidence of fuzzy mission and management.

We are in a dangerous place right now. Google makes a lot of noise about being ‘open’ but they are pushing ahead with non-standard web technology right at the moment when both Microsoft and Adobe, arguably the biggest offenders of proprietary web technology over the last decade, are coming into the HTML5 fold.

Interestingly there is already a bytecode standard and runtime we could use. It has an open standardized specification and very solid cross-platform open-source implementation that runs on everything from mainframes to cellphones. There are dozens of languages that compile to it. More than a decade of performance tweaking. First-class tooling support on Mac, Linux and Windows. It is even able to support multiple cores and has first class security sandboxing. Can you name it?

It could be argued that the iPad and iPhone are also in a way inspired by PARC, (parctabs and pads) though it may just be that the guys at PARC saw an ecosystem that would make sense and we are just now catching up to it. That ecosystem was called Ubiquitous Computing and it has three major ingredients which are analogous to iPhones, iPads and iTV. All this TV talk is pretty boring though and misses the point; that ubiquitous computing (interfaces everywhere) is just as revolutionary a shift as general computing (one device that can become many things) was and goes way beyond media consumption.

That last device is one we have not seen yet, though speculation is now running rampant, especially since the SJ bio published this week has direct quotes from Jobs saying he “finally cracked it”. Whatever ‘it’ is, presumably a simple model for interacting with the display.

Joe Hewitt wrote an interesting piece¹ talking about how latency between the iOS device and the display would be a technical hurdle to overcome. I don’t think this is likely to be the case though; Airplay display mirroring exists today and works fine, but in the case of videos and music why would you relay a media stream through a mobile iOS device to the display? This may be what has been cracked; by moving everything to ‘the cloud’ here is no longer a relay issue, instead the mobile iOS device triggers the stream, but from that point the display is receiving the stream directly. The mobile device can stream directly, but in most cases won’t need to. This improves stream stability, means that the device doesn’t have to say involved full time, and saves on battery life and wireless traffic.

Others have been talking about how none of this can work from a logistical standpoint because of the mess of how broadcast and cable television works; that you can’t corral all these moving parts and get the various players to work with you to provide enough content. That you’ll have to provide at least a cable card connection. I would argue this fight is already past the middle thanks to the iPad and that the only connector an iTV will have is a power cord. Maybe an ethernet jack.

There are already multiple apps for the iPad that provide streaming video either from cable providers or directly from programming providers like HBO. You can already to pretty well with these on an iPad, certainly well enough for the early adopter crowd. There are already sports apps that provided nice real-time data, and live video is the obvious next step.

Apps that work across multiple devices and screens may become the norm. Your NFL app has stats and field positions, do-it-yourself replay and more in your lap while the larger display is tied into the video stream, but they are tied together so things can jump around as appropriate.

Your presentation software can be driven from your phone or tablet, but the display is seen by everyone. Your notes are on your screen, people are already doing this now, it just gets simpler when the wall display is standardized and well understood.

Stand in the kitchen and ask your display where your kids are. Hello the ‘family clock’² from Harry Potter.

There are many scenarios that will be possible, or more practical, and new ones will be invented. That is the point of general purpose computing, and now ubiquitous computing; that the devices are malleable, and new things will be created on top of them that their creators did not conceive of.

There is nothing necessarily brand new here, especially to hard core techies. “The future is already here, but unevenly distributed”. Just as I was time-shifting my video watching with a satellite dish and multiple VCRs five years before TiVo, and archiving video with a digital tuner to a server for years before iTunes and other services came along, so too do other ways of thinking about content and computing move from being esoteric to mainstream. People are growing accustomed to nice interfaces to their data and devices, witness the Nest thermostat that is getting buzz this week.

I suspect many techies will initially feel let down by whatever iTV turns out to be, but will gradually realize how revolutionary it is to have a “wall computer” finally be around when they start to show up in kitchens and board rooms.

To reproduce:

1. Go to site, may be in a new tab/page, view current version of page.
2. Follow a link and view linked page.
3. Hit back to return to page.
4. Page loaded is out of date rather than the version you were just looking at.

I have clearly observed this bug in Safari on iOS 5 over a dozen times now, on both an iPad 2 and iPhone 4.

Timeline:

August 8 — Initial application through online form. I also faxed in documentation of my LLC the same day as I knew this would be required. I received thank you emails in response to both actions, the first surely automated, the second must be in response to someone processing the fax and tying it to my application record.

August 18 — Ten days later I received an email requesting a letter providing a physical mailing address (my LLC in this case using a PO box address) and affirmation that I have legal authority to enter into agreements. Fax sent back within the hour.

September 1 — Thinking a reasonable amount of time has passed I send an email asking for status. I get back an email “We are currently reviewing your inquiry and will get back to you as soon as more information become available.”

September 1 — Later that day I received an automated follow up email asking how my support experience was. I respond with appropriate displeasure at the slow and opaque experience I’ve had so far.

September 18 — No further movement to date. I lose my shit.

I can understand if they are completely buried in applications, really I do. If that is the case I hope they are hiring quickly to add capacity. In the meantime how much trouble is it really to give some indication of how long the queue is? Am I waiting another week, another month?

Two questions:

1) Is this typical? Are other people who have applied in this timeframe experiencing similar delays? This should be pretty straightforward to figure out that the LLC is real and approve or ask for more information. Which leads me to believe nobody has really looked at my information yet.

2) Is Apple so buried in applications of questionable origin (i.e. app scammers looking to make a quick buck with shovel ware or stolen apps) that it has effectively become a DOS attack on the application system itself?

For most of the last 20 years I have primarily developed for and with Microsoft products. It is well known that Microsofts main strength is that they know how to support developers. This experience so far really highlights the contrast.

As I’ve argued before Android will become meaningless to consumers. Watch a non-techie and how excited they are about their iPhone, that level of brand recognition will not occur with Android as it exists today. Worse (for Google) the only company that is going to really get anywhere with Android tablets in 2012 is Amazon. And it won’t be an ‘android’  in any way that Google would want it to be, in fact they seem to be shut out.

So here we are, back to a ‘Google phone’ as they should have pursued in the first place. The Android misadventure has been interesting, and useful in that it blew all kinds of value out of the market and forced OEMs to rethink their strategies. Perversely Google also probably just secured the #2 smartphone slot for Microsoft in a couple years.

Bottom line for me in platform targeting: Apps on iPhone, wait and see on Win Phone 7 but be ready for it. Books on iPad, but get ready for the Kindle tablet. Not sure how to get ready for that, it may be an Android development stack, but it is very likely there will be a publishing kit for interactive books that does away with the app layer. To be seen…

There are hundreds of people working for the various carriers and manufacturers who seem to think that their job is to create some sort of competitive advantage for their particular flavor of device. Instead of focusing on making the experience of the user as smooth and seamless as possible, which trust me is way more work than ‘differentiation’, they shovel cruft into the phone.

So we end up with huge compatibility problems. Apps that crash. Inconsistent user experience. Battery life that makes the device a toy.

The fundamental basis of a good user experience is TRUST. The user has to trust that when they tell the device to do something it will, that they don’t have to recoil in fear of doing the ‘wrong thing’ and destroying data, texting naked pics to the wrong person, having no phone because the battery died. The UI needs to be consistent so they can find their way around in a new screen quickly and intuitively.

Google had some inkling of this early on and tried to enforce standards, but they’ve since failed to hold the line. Now that horse is out of the barn and it is not coming back in.

The destiny of Android is huge market share as the layer under the cheap and inconsistent phones that will be replacing the ‘feature’ phone segment of the market. No more meaningful than things being Java based turned out to be. This is not a platform segment in remotely the same way that iOS is, or that future Win Phones have the potential to be.

I come to the Mac from a strange place. I’m not a ‘switcher’ in the sense that I am not a consumer user who when it came time to upgrade an aging Windows PC decided to try out a Mac. In fact I was very reluctant. Fortunately I have impeccable timing.

I can still recall very clearly the first time I saw a Mac (the original) and I an also recall seeing a Lisa very early on. In both cases they were in stores, the Lisa I don’t think was even powered on, sitting in what I think was the business center of Sears, back when Sears sold computers. The Mac was on, and I was able to play with it for awhile, I was enchanted with Paint. From that point on I lusted for a mouse, not necessarily a Mac as it was well out of any reasonable price range for my household. We did get a PC fairly early on, a Panasonic ‘luggable’ monster, and years before that we had a TI-99 4a that was the machine I cut my teeth on as a programmer.

PCs were where it was at for me, with a detour into Amiga land. But then there was NeXT and the cube. This was a damn exciting piece of hardware and software, and again vastly out of my reach. It was released while I was in college and I read avidly about it but it dropped out of mind as I moved on and started doing commercial development work on top of DOS and then Windows. I moved to Windows NT in the 3.1 era (1993–94 I believe) and from then forward Windows was unquestionably ahead on the engineering side of things.

More recently Vista got short shrift. I found it to actually be very solid when run on new hardware and configured well. I built an engineering department on it and had no issues that were attributable to Vista at any point. Windows 7 is the best OS Microsoft has ever shipped. It is Vista with only small changes.

All that time I ignored Apple. Yes, I got my iPhone the first week of availability, but I continued to scorn Apple engineering due to the miserable experience that is iTunes on Windows.  Proof enough of poor software practices and an often uneven and frustrating user experience.

At the beginning of this year I got my first Mac. Just a Mini, just to play with and see what was going on and to dabble with some iOS app ideas. It was a strange experience for me. At that moment in the evolution of OSX and Windows it was pretty clear that Windows was actually ahead in user experience. For all the praise that gets lavished on Apple for their great design it does sometimes come at a cost (case in point the sharp aluminum edges of my new Air that are digging into my wrists at the moment…) in actual comfort of the user. The experience of managing lower-case windows on the Mac desktop prior to Lion frankly sucks. Especially in comparison to the ease with which windows in Win 7 can be docked to screen sides and switched between via the task bar. Switching between windows on the Mac is kind of a joke.

But this changes dramatically with Lion. One very simple conceptual change; full screen apps are their own ‘spaces’. Spaces was never a concept your Mom was going to understand. Flipping between programs though, that makes sense. Grandma is doing it on her iPad, right?

When is a Mac not a Mac? The Mac ‘operating system’ was terribly primitive during the 90’s. Even the Win32 based versions of Windows were technically ahead, and there was no comparison with NT. The current Mac internals are something that fascinates me, because it is a NeXT station under there.  Obviously this has been the case for about a decade, but I missed the party. I’m glad I did. Now that I’m paying attention the Mac side of things and catching up with how things have evolved I’m pleased with my having missed much of the birthing pains that got us here.

I’m liking Lion a lot, and have a strong preference for apps that use the Lion way of going full screen. Fast switching between apps with a gesture is so natural. It is one of those things that makes you slap your head and wonder why things didn’t always work that way. It is so good that I will take it over multiple monitor setups. This is a big deal; I got my first 21″ monitor in 1994, it was $2000. I started using multi-monitor setups in 1996 when they were rare and exotic. I’ve had at a minimum two 21″ screens for most of the last decade, often three or four. Nice LCDs up on arms. I put my money where my eyes go. But now… I’m finding myself surprisingly happy with a single 13″ screen on this Air. Six months ago I would have told you it was impossible.

Two things will actually ‘kill’ it:

First the developer experience is awful. Technologically and ecosystem-wise it is a distant second to iOS. This has slowed development of quality apps for android, and injured the developers that have targeted the platform. Apple will continue to enjoy ever-increasing lock-in as people have a larger investment in apps that they want to bring forward onto new devices. You hear vehement arguments on both sides of whether or not Android suffers from fragmentation issues. From a dev standpoint, yes it does.

Second the platform itself will continue to fragment, and not just in different device form-factors. I think it is very likely when Amazon comes to the party they won’t even mention ‘Android’ anywhere. It’s just a color Kindle or whatever and that’s that. Want apps for it? Sure, we’ve got an app store, but it is going to be different enough from the get go and will start to drift rapidly from the Google version. Amazon is will see advantages in being their own beast, they don’t want or need compatible devices out there from other manufacturers. The original Kindle did just fine thank you and the next will do that much better thanks to expanded capabilities.

Chinese carriers are already making their own forks of Android as well. This will only accelerate. Carriers and manufacturers that are pursing the ‘old’ model of differentiation through small feature set differences between phones will continue to duke it out over ever small slices.

I hope I’m wrong. I hope Google figures out a way to herd the cats more efficiently, because Apple needs the competition, but I will happily predict that the tablet market is already lost to them. iOS owns it, and Microsoft will come in very strong with Win 8 devices that can play in their existing Office hegemony. Even Google knew to hedge their bets, Chrome OS laptops anyone?

Computer-in-your-pocket phones for the moment are iOS and everyone else. Microsoft is doing great work here but nobody knows it yet. RIM is RIP. Interesting times.

Would your Mom know if she had an Android? Would she care?

The specific cause of the failure is not actually important for lessons learned to anyone but Amazon. We’ll have to wait and see if AWS does the right thing and gives a detailed post-mortem as part of restoring trust. The explanation so far revolves around a cascading failure of EBS infrastructure, but still missing is an explanation as to why the failure crossed multiple “availability zones” (i.e. more than one physically separate data center… maybe). My best guess is EBS, which was already proving to be fragile, was pushed into a cascade failure by tenant applications trying to fail over to another AZ when the first failed. AWS better have as one of their lessons learned adequately communicating status to their customers.

The broader lesson learned, by those who needed to learn it yet, is that “The Cloud” is not a magical place populated by unicorns and data faeries. It is in fact data centers as (TK) noted, based on servers and databases as Larry Ellison is happy to explain.(TK)

The lesson all CEOs and CTOs need to take away is make sure your shit is architected properly. Stuff happens. Hopefully your boards don’t take away the wrong lessons¹.

A number of popular services were completely hamstrung by the outage, including Quora and Reddit. (After a five year relationship with Reddit I have to say it being down was probably worth several man-years of additional productivity around the world this week, but it is the one shining exception.)

Try not to take away lessons learned that are Amazon specific, i.e. Don’t trust EBS, or AZs don’t mean what you think they mean. Instead focus on redundancy, avoiding single points of failure, etc. Resiliency costs money, so make your trade-off between a few hours/days of uptime and increasing your costs by 50–100%.

The bottom line in resilient design is assume stuff will fail.

For the edification of anyone trying to bullet-proof their systems and my own future reference here is a round-up of lessons learned etc.:

Twilio was not impacted due to their design principles.² And there is also an old presentation that explains how they organize their virtual infrastructure on AWS.³ That Twilio did not go down is great news as the cascade failure would have taken out even more services, and it is solid proof of the importance of good engineering.

Sulia stayed up due to their doubly redundant infrastructure planning⁴.

Some design thoughts from Agile Sysadmin⁵

Pounding home the assume things will fail lesson, Netflix is AWS based and did not go down. They wrote a tool for themselves months ago that deliberately attacks their infrastructure called Chaos Monkey⁶. Note that this was posted five months ago. There is also a current discussion of their infrastructure resiliency at Hacker News⁷. And slides from a presentation given in March⁸

George Reese wrote a nice summary of design considerations⁹

Some load balancing and IP routing thoughts related to this from James Cohen¹⁰

Cloud services can be designed so that they don’t have single points of failure, and this virtualization is directly underneath your app rather than under your virtual OS. Unfortunately Heroku, Engine Yard, etc. do not yet have an architecture that can truly avoid the virtual infrastructure failures as they themselves were harmed by the AWS outage.

Here’s to sunnier skies…

The behavior I experienced was suddenly most, but not all, apps on the iPad had no sound and I couldn’t get it to come back. The volume rocker switch did not function in most apps. Apps that did work correctly included YouTube and other native apps.

I tried restoring it multiple different ways, no dice. Even restoring to default, like a new iPad, did not correct this problem. Stunning.

The cause (and suddenly obvious solution) is when you mute using the hardware switch, then go to settings and change the behavior of the switch to rotation lock without undoing the mute. You are now permanently muted, though apparently not in Apple built apps, just 3rd party. The simple solution is to change the switch behavior back to mute and toggle it.

It really amazes me that a full restore did not correct this problem, it suggests that there is a toggle for audio that is handled outside of iOS?