isaiah’s web log

It's been about two days since my cursed Apple ID was changed. It's working right now and I'm feeling optimistic about this fix. This is the longest period of functinoality in at least two years. Even if this is not the One-True-Solution™, it's safe to say that this is a huge step in the right direction.

I've tweeted about it a bit, so people have started asking questions about what happened. I thought I'd blog about it a bit to let people know what was changed and why. I should caution, however, that my understanding of this whole system is limited to just this one experience from one perspective. I'm certainly ignorant of most of the details, reasons, and caveats. If anyone with more knowledge would like to fill in the detail, please don't be shy, I'll include anything else I find out here on the blog.

I'm not going to mention the customer service details specifically other than to say Apple is amazing and I can hardly imagine any other company doing anything remotely similar. My six-color faith has been refilled to the brim.

OK, let's do this…

The Problem

As I understand it, the problem is primarily that my account had a very simple login... basically my first name @a-common-domain. And that it's been around for so long. Other people attempted to sign in with that name, or connected devices to that name, or used it as their backup email address... etc. Without 2-factor auth enabled, there are only a few password attempts allowed before the account goes into lockdown mode and refuses to let you even try anymore. And even with 2-factor auth enabled the system still locks down -- but you can use your 2-factor backup code to re-activate things. I have tried both over the past couple years -- both end in an unusable account.

It seems that my account was connected to so many devices/people/whatever that it ensured it would get locked out almost immediately after I unlocked it. And that's been the status quo for quite a while: I unlock the account and within a few hours it's locked again -- even when I didn't interact with the account at all.

A Solution

There is a new 2-factor system from Apple for iOS 9 and El Capitan. The new system only allows approved devices to request 2-factor codes. If one of the other approved devices doesn't allow it, the attempts to acquire a 2-factor sign-in codes are blocked and do not lead to account-lockdown.

I'll be honest: I don't fully understand why this blocking happens in the new system, but does not happen in the old system. If anyone can explain that in laymen terms, I'd really appreciate it.

Irony, Served Ice Cold

The new 2-factor system is only available for devices running OS X 10.11 (El Capitan) or iOS 9. If you revisit my earlier post about my cursed Apple ID you'll notice that last summer I decided to make a clean break; to stop using the account; and just move on. When did I decide was the perfect time? You guessed it! It was at the release of iOS 9 and El Capitan!

I decided that when I got my new iPhone 6S and upgraded to El Capitan I wouldn't let those new things touch the cursed account. Since then I've only attempted changes to the account through a web browser… ensuring that I would be stuck with the old 2-factor system and the cursed lockouts.

I'm 100% positive this irony will seem hilarious just as soon as I've stopped crying about it.

How To:

Getting connected with the new 2-factor system is not difficult -- once you know how. Here's the basic details:

You can tell you are using the new 2-factor system if the 2-factor sign-in codes you receive are 6-digits. If they're only 4-digits, you're on the old system.

Turn off the old 2-factor

In order to get on the new system, you have to completely disable the old one first. If you're not using 2-factor at all, then you can skip this part.

Go to https://appleid.apple.com and sign in.

Under Security choose Edit

Click "Turn Off Two-Factor Authentication"

Turn on the new 2-factor

On a Mac running OS X 10.11 (El Capitan) open the System Prefs (I know you can do this via iOS too, but don't ask me how, I have no idea).

Choose iCloud and sign in if necessary.

Click Account Details

Click the Security tab. (this seems buggy -- i had to click this several times -- signing in each time -- to get it to display the Security sheet -- occasionally multiple "Sign In" sheets would drop down overlapping each other -- but be persistent, keep typing in that password, it seems to work eventually)

Click "Set Up 2-Factor Authentication"

After following the steps you'll eventually have to sign in again and will receive a 6-digit 2-factor code. This means you're good to go. From there you can add other devices, so long as they're running OS X 10.11 or iOS 9. I'm not sure what happens if you try to connect an older device. I'm guessing it's just a No-Go.

Brent Simmons wrote a nice little post about pop music. Nearly every song on the list was a regular on my sometime-in-the-90s walkman. It was a fun to listen, it felt just like getting a mix tape from a friend in college again.

Inspired I made a my own. Not the same, just a counterpoint of songs that or similar albums from the same era that are a bit more me. Here’s my list:

On Rdio: http://rd.io/x/QIkQL6tGzg/

Gigantic — Pixies

Really, the whole Surfer Rosa album.

Ever Fallen in Love — Buzzcocks

I’ve swapped Clash for the Buzzcocks. More pop. And also because it was the first punk song I really loved.

Look On Down the Bridge — Mazzy Star

This is just a favorite Mazzy Star song. Not pop at all really. Just good.

How Soon Is Now — The Smiths

Swapped The Church for Smiths. Because Johnny Marr.

Fire in Cairo — The Cure

My wife is a big fan of later cure. I’m more preferential to their first album.

Walk on the Wild Side — Lou Reed

If I have to explain why this awesome then just leave now.

Lust for Life — Iggy Pop

I’m not sure what I swapped this for — but it seemed right.

Waiting for the Great Leap Forward - Billy Bragg

Pogues are awesome too. But Billy Bragg speaks to my personal political extremist bent.

Waltz #2 (XO) — Elliott Smith

Every mix tape must contain at least one Elliott Smith song. That’s the rule.

Birthday — The Sugarcubes

I love this song. That is all.

A Message to You Rudy — The Specials

I struggled to swap anything for coronershop. I don’t think there’s anything else quite like that. But this is also from multi-cultural working class England — work with me here.

Cold Tea Blues — Cowboy Junkies

This is not pop. But it is the junkies and is mine and my wife’s rather quirky and unusual wedding dance song. So I’m forcing you to listen to it too. ;-)

I like movies. Old ones, new ones, bad ones, and all the rest too. I don’t watch much TV, and for really crappy reasons, I don’t read as much as I’d like to, so I watch a lot of movies.

About five years ago I moved to a new place. The spare room came set up as a media room and since the previous owner couldn’t take his gear with him, he offered it to me at a steal. Long story short: I’ve got a little theater above my office. It’s kind of amazing.

iTunes

Being a longtime Apple nerd I originally played my movies over an Apple TV (gen-2 later replaced by gen-3). I would put movie files on my old Mac Pro and share them via iTunes sharing. This worked OK, but there were a few flaws:

Apple TV and iOS are very finicky about file types. This means re-encoding at lower bit-rates with a specific “container” (no MKV containers allowed).

iTunes isn’t a set-it-and-forget-it server. It requires interaction on the sharing end nearly every movie.

The iOS UX is crappy. Browsing a large library crashes the player, video stutters, and audio is sometimes out of sync.

In the normal Apple way, these problems simply exist. They are not solvable by tweaking or fiddling. A situation that’s great when things work — but maddening when they don't.

Plex

Recently, I’ve switched to Plex for movies. Plex is a cross platform media server & player. It serves pretty much the same function as iTunes sharing via an Apple TV, but it works on nearly any device. It’s pretty great. Really. You should try it.

I run the Plex server on my old Mac Pro, the same as I did for iTunes. I drop movie files into a folder. And then they can be played via numerous native clients for everything from iOS to Xbox (and even the web too!). It even transcodes video on the fly so no more handbrake settings and I can play movies on my old iPad 1. A feat that even Apple’s own native player can’t compete with. The server is very stable and requires no interaction. It’s all really amazing… until it breaks.

When it breaks you have to get out your propeller beanie cap and nerd-up. I’m split on this… on the one hand iTunes was broken. And nerd-cap or not, there was nothing to be done. On the other hand, digging down into log files and sorting through error messages is not for everyone. Which is better: fixable-for-few or unfixable? I'll be honest, I hate both of these options. I don't think I'm alone.

This all reminds me so much of the days of MP3s right before it got great. Right before the iPod.

This situation was laid bare yesterday. I picked up a new Fire TV -- a set top box a lot like an Apple TV. It’s a bit cheaper than the new Apple TV due out soon and comes with 4K video support (something my little theater aspires to someday) — but the Plex app for this platform was almost nonfunctional out of the box. It took me hours of digging through the net and carefully eliminating possible problems. At one point I even gave up and packed it back up into the box. I did find a solution in the end -- after hours of fiddling. Now I have very nice videos playing -- but only after a lot of tears first. Here’s my bug report post to the Plex folks.

Passion

TL;DR: This blog post is my last desperate plea to anyone with any contacts at Apple to please pass this info on to someone that might be able to help before I give up and abandon my online persona and thousands of dollars of apps and media.

For a while now my Apple ID, iCloud Account, and everything associated with them has been less than perfect.

It started during the beta of iCloud, years ago. Synced things would intermittently duplicate or disappear: notes, to-dos, and bookmarks, etc. Every few weeks I'd open up Safari to find my bookmarks duplicated, sometimes dozens of times. At the time I laughed about it, it was a funny bug. And I wrote it off a beta-weirdness. Things would surely improve as it matured, I thought.

Then, maybe four years ago, my Address Book picked up the weirdness too. It would randomly mix info from various contacts. This lead to some very awkward phone calls and text messages -- thankfully nothing too embarrassing.

Last year, things got really crazy when my account started getting disabled. Sometime in 2014 my Apple ID was "disabled for security reasons". Then it happened again after a month or so. Then it was weekly. Then daily. And it has remained this way for months.

A disabled account means that anything that requires authentication (buying apps, logging into the developer site, iTunes Connect, etc.) is inaccessible unless I first go through the many steps of reactivating the account using my 2-factor recovery key and a trusted device. It's very slow, very inconvenient, and more than tad frustrating.

Sometime during all this I started calling Apple support for help. The folks at Apple are, without fail, 100% the nicest support folks around. And most calls end in hopeful solutions. Some were things I tried (e.g. logging out and back in on all devices, disabling various things, password changes, 2-factor, etc.), some where things Apple tried (some "corrupted info" was found once and corrected, a filter was added for forged email bounces, etc.). However the end results are always the same, iCloud continues to corrupt my data, and my account is disabled often.

Recently a new issue has cropped up: trusted devices that are used to sign in to tho two-factor-auth have disappeared from the account. When this happens I get an email that says, "when you erase a device it will be removed from your list of trusted devices." And in case you're wondering: No, I have not erased this device -- or any device -- recently. Calls about this issue have not been very productive (Apple claims that there is nothing wrong with the account, the device was never a trusted device, and the email is not from them). So far it's just an inconvenience of re-adding the devices, although I think if they're all removed it will be a real problem. I guess I'll cross that bridge when I come to it.

I've had this account since Apple introduced accounts. It's my first name, an asset for someone doing business online so much. Well over a decade of music, movies, and apps were bought with this account. It's my developer account with several apps registered to it. But as hard as it would be to abandon my online persona and the vast amount software and media I've invested in, I'm considering it.

If I can't find the a solution before I get the new iPhone in a few weeks, then I'm just going to leave it for dead. I'd like the new phone to be completely free of any traces of this account or problems. On the day I get the new phone I'm also going to reformat all of my Macs and other devices and start fresh.

Nuke it from orbit, it's the only way.

I'm finishing up a 3 month beta test of Stacks 3. This post is a reminder to future-me when I begin my next one. It's tough to get my developer-brain-assumptions into the right frame of mind for collecting solid info from novice users.

Start talking

During a public beta I make it a regular habit to explain that the goal of beta testing is to root out problems so that I can fix them for everyone. For many novice users this is a new concept. Their experiences with tech support up to this point is usually quite different.

Normal Tech Support

Tech support is contacted when a user has trouble. The goal of tech support is to eliminate or work around the trouble for that one user. Usually this is done by showing the user how to use the product or work around a known problem.

TL;DR: I tell the user they're holding it wrong and teach them how to hold it right.

Bug reporting and Beta Testing

A bug report starts exactly the same: the user has trouble. But that's where the similarities stop. The goal of a bug report is to fix the bug, not just for one user, but for everyone. This is achieved by isolating and examining the bug, then fixing it. Any changes I ask the user to make will make the bug worse, not better, to make examining it easier.

TL;DR: I ask the user to show me how they're holding it wrong so that I can hold it wrong too.

Dear Future Me

In a public beta most users will contact you expecting help. Communicating your bug-fixing goals early and in a friendly way is key. But communicating your goals won't change theirs -- they'll still want to work around the problem and get back to productivity. When you've got your bug report, remember to return the favor and help your beta testers fix things and keep working.

Jared wrote a great post about pragmatic use of less than perfect code.

"There’s such a thing as judicious use of shitty code. "

People that know me and my perfectionist bent might be surprised to learn that I'm a strong proponent of shitty code. Not all my code is shitty, of course. But some is. And I think that's a good thing.

Moreover, I think you can be smart about how and where you use shitty code: the time invested in code quality should be relative to how much maintenance and/or reuse the code will see.

When you're building a one-off controller that's not likely to be reused, in a high-risk market, on a small app: code quality is a YAGNI. When you're building a reusable view that will be useful in several areas (high reuse) -- or is a key component of a large app (needs maintenance) -- or on a very large project (need refactoring) -- then by all means, take your time, do it right. The investment will pay off.

The important thing is to realize that code quality has a short-term cost and a long-term benefit. Invest more in the short term cost when the long-term benefit is likely to pay off.

It's sad news to hear that InstaCast will be no more. Sad because it was my favorite podcast app and because it's a grim sign for indie development on iOS.

Why did this happen? I don't know. Other podcast apps are apparently doing well. So maybe they just got run over by the competition? That's probably it.

On the other hand, they had, at one time, a very popular app and a large user base. It's made me really wonder why it's so hard to keep a popular app going on iOS?

On other platforms revenue from upgrading a popular app is often larger than the initial launch. Sometimes much larger. Building a popular app is tough. Getting to 2.0 is tougher. But when it happens it's a glorious thing. Obviously for revenue, but also for users: it ensures that the best apps evolve and mature with the platform.

When iOS was getting underway I often heard, "What about upgrades?" The answer was often, "Just give them away. It will bring more users. There are so many users it won't matter." At the time is seemed odd, but almost believable. I'm not so sure anymore.

Apple's numbers are in. And they're even more amazing than predicted. Except for the iPad.

Some justify the news with convenient stories. Like hoping that it's temporary. Or that it's an early stage in the durability cycle.

But cheap tablets had a good quarter. LG and Lenovo sold a few more tablets last quarter than a year ago.

The less convenient story is the elephant in the room: there's no reason to buy an iPad. In 2013 I wrote that most tablet users are browsing the web or updating Facebook. Things a $99 tablet does great. Two years later, things have not changed enough.

Apple needs to give users a reason to pick up their iPad other than Facebook. The iPad needs to solve problems that cheap tablets can't.

And that means apps. Real apps. Apps that sell for more than 99¢. Sadly, the iPad is chained to an app store that appears bent on shovelware, brands, and free games (contains in-app purchases).

There are just three characters different between the two methods below.

- (BOOL)instantiateWithOwner:(id)owner topLevelObjects:(NSArray **)topLevelObjects

- (BOOL)instantiateNibWithOwner:(id)owner topLevelObjects:(NSArray **)topLevelObjects

One is OS X 10.7 compatible. One is not.

One will retain strong references. One will not.

One is deprecated in OS X 10.10. One is not.

Those are pretty important differences, but if you were building with the Mac OS X 10.9 SDK, as I was this past summer, there are no warnings generated if you mistype those three characters since both methods were completely valid.

I'm not really sure which is which, even right now. But I am certain I chose the wrong one last summer.

I have a toolbar right above an NSScrollView. To make things fit in with the new Yosemite look I've made the toolbar vibrant and moved it down to cover the top portion of the scrollview. It was easy and looks great.

But there's a little problem. The scroller is a bit underneath the toolbar. That's no good, sometimes you can't even grab the scroll thumb. The obvious solution is to modify the size of the scroller to go up to the edge of the toolbar instead of all the way to the edge of the content view.

Looking through the NSScrollView docs doesn't show any obvious way to do this. So after a bit of cursing at AppKit, I sat down and made a subclass that modifies the scroller frame. Not a big deal, but ugly, and probably not very future-proof.

Oh well. Code done and checked in. Moving on. That is, until I watched the 2014 WWDC session about vibrancy. It turns out that this situation was taken care of. And there are new properties on NSScrollView just for this use case. How did I miss this stuff?

It turns out the thing that's missing is documentation. There are no mentions of insets anywhere in the NSScrollView docs. But the properties are definitely in the NSScrollView headers and even documented too.

@property BOOL automaticallyAdjustsContentInsets;

@property NSEdgeInsets contentInsets;

@property NSEdgeInsets scrollerInsets;

If you're building something similar to me, then you'll need to disable the automatic adjustment, then set the scrollerInset top inset to the height of your toolbar. Two lines of code and done. Subclass deleted. Yeah.

I don't really need a watch. I'm not a watch person at all. I have one. It's OK, I guess. I hardly wear it.

When I first heard about the Macintosh I was sitting in front of a CPM machine (that's a very old kind of computer you've never heard of). My friend Alan told me how the Machintosh was a friendly computer. And graphical, with icons, and had a thing called a "mouse." I was skeptical. I had no interest in an easy computer. I could use CPM just fine, thankyouverymuch. Months later, while attending a tiny computer convention, I walked around a corner and there was a Mac running MacPaint at the tiny Apple booth. In the next 30 seconds I became convinced that this little beige box wasn't just a computer, it was the very best computer. The Mac was so different that you could hardly begin to compare it to other computers. And I wanted it. A lot.

In 2001 a friend let me in on his secret Apple project, "It's an MP3 player," he said. "Lame. Already done!" I told him. Yes, I really talked like that then. When it finally shipped months later it was clear this wasn't just an MP3 player, it was the very best way to play music. It was so different that you could hardly begin to compare it to other MP3 players. I bought one of the very first available anywhere.

I was at MacWorld that morning in 2007 when Steve announced their new mobile phone. Everyone knew they were working on it, but I'm not really a phone person. I didn't need it. 30 minutes later I walked onto the show floor and saw it spinning in the glass case. I knew that this was not just another phone. As of that moment the best way to make calls was no longer a phone. The best device for phonecalls was a tiny pocket sized computer. The iPhone was so different that you could hardly beigin to compare it to other phones. I bought one on the first day.

After a day of attempts to fix a stuttering UI, I gave in and opened Instruments. results were, of course, amazing. When I understood the real source of the slowdown it was easy to fix.

I'd like to get in the habit of using Instruments first, before wasting a day… no, really, I would, honest… but Instruments isn't easy to use. It looks pretty enough, but the UI is strange bordering on arcane. I'm pretty sure I'm not alone in feeling this way.

Dear Twitter,

I hear you want to build a "relevance engine" to stay competitive. I can help you. I know an amazing simple and under appreciated relevance algorithm. Here it is:

Start with a list of the people that the user wants to hear from.

Display their content in chronological order.

There's no step 3.

I don't use an iPad. I own one, but my son uses it now, mostly for games. Between iPhone, Desktop, Laptop, and Kindle I could never find a place for it in my life. So I've started quizzing my tablet-using-friends about theirs. I ask friends, "When do you use your tablet and what for?" Almost unanimously the answer is: around the house for web, Facebook, email, and books. Basically a living room couch web browser and e-book reader.

What I noticed though, is that all of these tasks are largely platform agnostic: It's tough to tell Facebook running on an Android device from Facebook running on an iOS device. There are some great platform specific apps of course, but they seem to be more niche use cases, at least in my anecdotal polling.

This made me wonder if the iPad's platform agnostic nature will have a long term impact on the the platform. Without app lock-in will less expensive tablets become more popular?

The new iPads are now both very light and both have amazing screens. Just refinements, perhaps, but amazing ones. However both features seem to be nearing the point of diminishing return. Further weight reductions and screen improvements are unlikely to make for a better Facebook experience in the living room, we've reached "peak Facebook" already.

And here's the crux of this post: what else is there? How does Apple make the living room web browsing experience better than this? Or how do they change what users are doing with their iPads? When Samsung catches up in a few months what will keep people from choosing the least expensive device? Or will there be any reason at all?

Update: i initially posted something different here. but it was too abstract. no one understood. and confusion isn't what i intended.

So I'll just leave this here instead.

At WWDC this year I attended a fund raiser to support the Cartoon Art Museum thrown by Pixar tech god Dr. Wave. This year Andrew Stone gave a demo of some of his early NeXT graphics software collaborations with Pixar and told stories about the early days of the OS we now keep in our pockets. Andrew said it was a magical time because suddenly, with the help of a NeXT cube and its free development tools, a lone developer could build software that rivaled what large companies were selling for hundreds of dollars.

We all know that new tools bring about radical changes. When a single developer can build and sell software that competes with companies that are staffed by dozens or hundreds, the game is no longer on a level playing field. A single developer invests only their own time so there’s little risk and even a comparatively tiny revenue is a success. It’s just a matter of noticing the tools when they arrive.

In the software industry, new tools are invented all the time. Word processors used to take a large team to build, a publishing company to distribute, and a retail chain to deliver. Now we can write a basic Markdown text editor over a free weekend and send it to the App Store to reach millions of people.

The past ten years has seen an explosion of social services, but for an independent developer there are not many ways to participate. We can only build clients beholden to services and their advertisers which is no fun. The reason, of course, is that servers, storage, and bandwidth are expensive. Business models that require large capital investment are pretty much out of the question for independent developers. What we need is new tools.

App.net is a service that provides developers with the tools a social service needs: a social graph, file storage, and a robust API to group them together. As icing on the cake, it has a lively installed base of over 100K users that are hungry for more apps (which handily solves another social problem — the “there’s no there, there” problem — but that’s for another blog entry). But some services have become developer-hostile when they start to sell their product (your users) to their customers (advertisers). In constrast App.net's business model sells premium service to users directly (like Dropbox and Flickr), so your customers and their customers are one and the same. Aligned goals make for better business partners.