waffle

Aaron Holesgrove uses the time-tested technique of replying inline to shoot down Gruber’s “Why Windows 8 Is Fundamentally Flawed as a Response to the iPad“. I thought I’d do the same. I don’t agree exactly with either of these men, but the way Aaron responds makes me think he’s missing a few things.

Yeah John, Apple is the ONLY reason for all of this stuff going on in IT recently. I’m not trying to disagree for the sake of it but I would argue that Apple (and iOS) aren’t the biggest enemy of those three companies; for them, their biggest enemies have been themselves. It was predicted many years before the iPhone and its mobile operating system hit the market that all three of those companies would fall flat on their faces for their respective reasons (see here and here and here) – iOS is simply the straw that broke the camel’s back, I’m afraid. Despite its enormous and disruptive success, the Apple gets too much credit for how it affected its competitors – especially those ones.

Sure, Apple’s not the only reason the IT world is in upheavals the past few years. Windows Mobile hung out near the brink of destruction for several years and couldn’t be saved with the 6.5 coat of paint; I’m pretty sure it would have collapsed under its own weight sooner rather than later. Palm says they started designing the Palm Pré before the iPhone and I believe them. But you know what Android phones looked before the iPhone? More or less like a BlackBerry or a Motorola Q. That’s what smartphones were.

Actually John, iOS IS built on top of Mac OS X and its core principles. It is common knowledge that it is a modified version of OS X with a touch centric shell on top.

iOS is built on top of Darwin, and has a bunch of programming frameworks from Mac OS X. That’s it. There’s no way to “get back to Mac OS X”, there’s no Mac OS X window server, there’s no AppKit and no Carbon (the two frameworks available for running apps with the Mac OS X interface), there’s no TextEdit, Dock, Desktop or menu bar, there’s no System Preferences, there’s no visible user accounts, there’s no Dashboard, there’s no Exposé, there’s no Software Update, there’s no Finder.

Maybe it is “common knowledge” that it’s a “touch centric shell on top”. Maybe people who care enough to believe a little bit about this think that’s what’s going on. That’s like saying a rally car is like a fully featured Honda Accord with a racing layer on top. iOS is as much an extra layer on top of Mac OS X as Windows Phone 7 is an extra layer on top of Windows 7. It’s not an extra “layer on top” if you toss out nearly everything except some of the innards and replace the layers you threw out with new technology.

Microsoft’s demo video shows Excel — the full version of Excel for Windows — running alongside new touch-based apps. They can make buttons more “touch friendly” all they want, but they’ll never make Excel for Windows feel right on a touchscreen UI.

John, I don’t know what has you so convinced that Office has to look different in order to qualify that it belongs on a tablet. The ribbon interface was supposed to introduce bigger buttons along the top so that menu options were easier to find and I think a lot of users don’t want to re-learn how to navigate Office all over again – they just want it pretty much the way it is on the devices they want to use.

I’m a fan of the Ribbon and some of its features help in making the interface easier to use from a tablet. But it’s nothing like adapting a program for a tablet. If the Ribbon was truly supposed to introduce bigger buttons so that they’d be easier to tap, some of the core formatting buttons would be larger. The button for bold and the drop down list for font size are pretty much their old size from Office 2003 when in the Ribbon. The Ribbon includes many features that are dependent on a cursor, like live preview of styles in galleries as you hover and the disappearing mini-toolbar on selection, which disappears based on the cursor distance. Some Ribbon features incidentally make it easier for everyone to hit buttons, but it’s hard to mistake this for tablet-specificity.

The Office apps aren’t different from any other app; they don’t have to look different per se, they have to work differently. There’s basics: they have to accommodate a touch keyboard gracefully and move the content out of the way automatically. They can’t rely on right-clicking. They have to present the options available in a fraction of the screen space. They should embrace the ability for direct manipulation and multi-touch since people will come to expect it. Excel in particular will need work to differentiate between scrolling, moving cells and changing the selection. Without a shift key available all the time, how do you select a cell range spanning 800 rows without going nuts? Moving to the Ribbon has saved Microsoft numerous headaches in preparing for this work (imagine having all those Office 2003 toolbars with rows of 16×16 icons to hit now), but there’s work to be done.

Now, all of that aside, I think the reason why we saw a full version of Excel in that video demo is because there is no way Microsoft would have demo’d Office 15 today as well – and from the screenshots leaked so far, it looks like Office is getting the Metro makeover too so your argument is moot.

This is actually a good point. I’ve been worrying about the rough break between Windows 7 and Windows 8 though, since not everyone’s going to upgrade to Office 15 and not every developer is going to want to or be able to remodel their interface for Windows 8. Having an OS so clearly with two halves with significant investment in both (Windows Explorer got a Ribbon and that’s what we’ve seen so far) is going to be worrisome. Compare and contrast to Mac OS X Lion which also brings full-screen in a big way, but which keeps all the familiar concepts going. You get to learn about switching fullscreen spaces and about disappearing toolbars, and that’s it.

Consider the differences between the iWork apps for the Mac and iPad. The iPad versions aren’t “touch friendly” versions of the Mac apps — they’re entirely new beasts designed and programmed from the ground up for the touchscreen and for the different rules and tradeoffs of the iOS interface (no explicit saving, no file system, ready to quit at a moment’s notice, no processing in the background, etc.).

Now we’re comparing apples to oranges. Firstly, iWork isn’t a ‘beast’, it’s a sexed up equivalent of Google Apps – a competent, entry level productivity application suite. What neither of those applications are, though, is Microsoft Office – say all you want about Microsoft products but Office has no peers, particularly in the enterprise, and has three times the amount of features of anything else. There are no comparisons to be made here.

“Entirely new beasts” are supposed to refer to wild animals and not to intimidating, capable animals. The point is that Pages on Mac and Pages on iPad look and work very different, and for good reason. We’ve been over this by now. It has to do with a million things, but being Word’s equal on features has nothing to do with it.

Microsoft, on the other hand, are looking to make tablets that are full screen computers which you can do anything/everything with – dock them as full computers, do full-screen computing using things other than touch – off screen gestures, voice control, etc. It’s a totally different kettle of fish.

No, Microsoft are primarily looking to make sure their existing OS continues to fly off the shelves because that environment is their crown jewel. (The Xbox and Windows Phone 7 departments are also of value, but they don’t contribute nearly enough to make a dent in Microsoft’s central focus right now.) Windows 8 now looks to be a pretty good start for tablets at the expense of shoving that same experience down every single Windows owner’s throat. No one knows how that’ll shake out.

The ability to run Mac OS X apps on the iPad, with full access to the file system, peripherals, etc., would make the iPad worse, not better.

Agreed – but just because that’s true of Mac OS, that doesn’t mean that the logic auto-applies to Windows as well. Mac OS has been through a giant state of flux – Mac apps used to be written in Carbon which was an afterthought in the migration towards 64bit Intel CPU’s and now Apple are becoming more partial towards their own CPU’s and designs. That situation is a mess currently.

Wow, yeah, Mac is the platform riddled with a gigantic lack of future vision. Apple’s dance from PowerPC to 64-bit PowerPC to x86 to x64 has hurt developers, but it hasn’t largely hurt real people. Aside from wondering why Photoshop is slow or where VBA went, they’ve been mostly unharmed. Meanwhile, Windows is the platform with a dozen different directions and technologies caked into layers over the past twenty years with no telling if they’re ready to make the hard decisions in order to continue staying relevant. And I don’t mean MinWin. Untangling your components, while arduous, is the easy part. Microsoft has just proven that they can make Windows run great on tablets, but not that they can sort out if the tablet experience belongs on the desktop/laptop and vice versa or what developers should be aiming for. No matter your opinion on iPads and their capabilities, there’s at least a clear line in the sand between what a Mac is and what an iOS device is.

iOS’s lack of backward compatibility with any existing software means that all apps for iOS are written specifically for iOS.

Again, apples and oranges. This mentality works well for Apple products because basically no one could care less about them up until about five years ago.

No, this mentality works well for Apple products because it produces an alternative that is measurably different from what PCs have traditionally been. And I don’t mean in terms of file systems, I mean in terms of encouraging direct manipulation. Go look at the success the iPad’s having on learning, as an aid for autistic children or as a first computer for old people who has shied away from the traditional PCs, including Macs.

Most of these people don’t care about Apple the brand but about a different kind of user interface and even the tablet as a form of more approachable computer.

With Windows it’s different – people would expect Windows tablets to have backwards compatibility with old Windows apps because if it didn’t, they could have just settled for an iPad instead and been one of the trend-setters. Sure some apps in Windows 8 tablets will look ugly but at the end of the day, backwards compatibility with legacy Windows apps isn’t a drawback – it’s a feature, because that’s what the market will demand.

Exactly. That’s what I think will happen. But that also means that the Ribbon acting as a touch accommodation is a happy accident, like we said a few paragraphs prior.

And hey – if Windows 8 tablets are supposed to be operated only like iPads, people will have that option too – there will be a Windows 8 app store which will serve touch centric apps written in HTML5 and JavaScript.

I’m hoping they will, except that I hope Microsoft won’t completely toss their technology portfolio. (I know they won’t and I know there’s an awful lot Microsoft’s left for BUILD and won’t talk about today.)

There’s a cost for this elimination of complexity and compatibility, of course, which is that the iPad is also less capable than a Mac. That’s why Apple is developing iOS alongside Mac OS X.

Yeah and that cost is called “credibility”. Microsoft wants your tablet to be your total solution and just because Apple can’t do it, doesn’t mean that someone else can’t either – your roses colored glasses deceive you John.

Credibility? If Windows 8 for tablets is all Metro all the time, it could be a great thing, but once you are forced to swap to a second user experience, you’ve lost them, for the same reason that you’ve lost many comfortable Windows users today the moment you expect them to fall in love with running Metro side by side with classical Windows. Things just don’t change that dramatically when you’re in the middle of doing something.

Remember the Ribbon boo-hiss? Going on the currently released information about Windows 8, imagine the sound of Windows owners worldwide realizing that they have to learn about and live in two separate environments, one of which is by definition not fully intended for their computer’s shape, because of factors they don’t control. The Ribbon was mere foreplay.

It’s time for some hard decisions.

For far too long I’ve been too silent on my four apps; there’s a small bug fix here and there, but nothing really new. What it comes down to is as always lack of time. This time, it’s in combination with the state of each application in relation to interest and need. Suffice it to say that I either think something is perfect already, or that it deserves a bigger upgrade that I can only nudge towards slowly. Reading and thinking about the acquisition of NetNewsWire (an app I love by an awesome developer who helped me get started) has tipped the scales: I can’t keep doing this, so I’m going to do something about it.

• Hex Color Picker is by a wide margin the most used waffle software offering. Mostly, it does what it does with aplomb. There are two recurring issues.

  Various people have different opinions on which is the right way calibrated colors should be handled, and some are asking for more sophistication than the color picker and the color model in Cocoa will easily accommodate without in-depth knowledge. I’m out of my league in this field. I know about color profiles and gamma and calibration and I know what it’s all good for, I just don’t know it.

  Moreover, people keep asking for RGB sliders. I’ve been sort of working of a mockup of this that does much more but tries to stay as nimble and straightforward as Hex Color Picker has been since its first release, and we’ll see how this goes.

  Hex Color Picker is already open source under the BSD license and is actually shipping with some apps for internal use to my great delight. I might stick it on github yet, but it’s pretty good. I’m not planning any big upgrades, and it doesn’t need any. It’s the least of my worries, but if anyone wants to revamp its handling of calibrated colors (currently more or less deferred to “toggle this checkbox if you’re feeling upset and see if it starts working”), you will have my full backing.

• Google+Growl is the next most popular utility. It makes Google Notifier send Growl notifications. It, too, has two issues.

  The first is the Google Apps issue, which I’ve gone into before. Sadly, apparently nothing has changed.

  The second is that as Gmail Notifier became Google Notifier, I adopted Calendar notifications. I don’t think I have them right because everyone has a different opinion on how reminders should recur, and I think I never will because I don’t use Google Calendar myself. Doing something that I won’t use myself is something very bad that I try to avoid.

  Google+Growl is not open source, and this bothers me. It will be released under the BSD license by the end of June, new version or not, and I think I want to be able to hand over its maintenance to someone who’s dedicated to seeing through that calendar notifications work well. Open source means available to everyone, but handing someone the keys and the responsibility is not only useful for me but of value to everyone who uses Google+Growl and wants to see it develop.

• ThisService is interesting. I was dragged into the world of services by a friendly request from John Gruber and it turns out that it was pretty neat and useful. I can only speculate, and I do not mean to hog any kind of spotlight from the true pioneers, but I would love for there to have been a connection between ThisService’s popularity and the resurgence of the Services architecture in Snow Leopard. No matter the outcome, it makes it even harder to admit that I’ve done very little to actually put a version in people’s hands that takes advantage of the new service features.

  I’ve tried several concepts for ThisService 3.0, including trying to build in JSTalk support for an extra edge (I realized I’d have to build in CommonJS module support for that to be useful, and I didn’t want to commit to getting those details right, and I had a hard time trying to get what I thought was basic functionality to work in the first place) and jumping to a document-based app where you’d have a service blueprint document which you could open, manage bundled resources for and click a button to produce a service from with the latest files. I’m not sure at all what the response would be to the documents angle.

  To be quite honest, I’m not sure if ThisService is even needed anymore. Automator basically does the job for producing services, supports most of the Snow Leopard features and also does the whole workflow with interacting components thing, giving it a definite edge.

  The feedback I get regarding this will determine ThisService’s ultimate fate. If I do keep it going, I’ll focus on getting out a product that can handle the Snow Leopard features. And in any case, I’ll make sure the source code will be made available under the BSD license before the end of July.

• Monocle is the remaining of the four apps currently available. Getting something like it going well enough has been my desire ever since I released my first Cocoa app. Every release I’ve said “this is the release that’s going to be good enough for me to use, too”, and that never ends up happening. I’m not sure why.

  The current version of Monocle isn’t bad for what it does, even if the tradeoffs I made to get it out there stand out like sore thumbs. I didn’t really want to skimp on mouse support for the suggestions list. For a fleeting moment after release, it was fine, but now I run Google Chrome every day at work and I expect something more from search. There’s something new cooking in this aspect, but it helps keep any short-term upgrade to Monocle at bay and never seems to materialize. I started this prototyping work literally just before Snow Leopard and haven’t gotten much closer since, mainly because I’m trying to create something genuinely new, and implementing the level of flexibility and robustness I’d like to have consumes me. It’s hard work when I work at it, I feel bad when I don’t do it and more often than not, I have to go back and change something, breaking everything that’s been constructed. Perfect has become the enemy of the good.

  Beyond that, Google giveth and Google taketh away. I like the instant results, but the AJAX Search API it uses is deprecated, which means its days are literally numbered. In its place is a solution that’s for pay and which doesn’t provide what I wanted.

  Monocle will be made available under the BSD license before the end of August, because it’s what I consider the ugliest of my shipping code. Nevertheless, if my reputation drops as my apps improve, I think that’s a worthy tradeoff.

• I announced Rouse last year and hoped I’d be able to get close to a working version for primary web browsing over the summer. This has turned out to be optimistic. I’ve ran into bugs and limitations, but I’ve also given up far too easily and spent time doing other things. The good news is that Rouse is now moving forward again… and the bad news is that, because of all the above, I’m constantly feeling bad for not working on what I keep imagining is a shorter while on an upgrade for things that people are actually using.

  I’ll continue working on Rouse and change my perspective from “this could kill OmniWeb for me” to “maybe this could turn useful”. Making a well-made web browser is a suicide mission. If I ever “get there” to where it’s good enough for my specific tastes, great.

  Rouse will also be moved to public development. I’ll make my progress in the open. I don’t expect anyone else to join me, but it’ll at least be useful.

• Shortcut Recorder is different in that it’s a control I started development for. Most of the actual development has been contributed by other people, and I’ve been the loudmouth in the corner making guesses as to what the current code base does and apparently earning the complimentary licenses (which I mostly turn down and always recommend go to the real heroes). While I designed how it works, it owes a bigger debt to everyone else who has kept it alive since. It warms my heart when I see it used in apps, but I’d love to completely turn over control to someone else as well.

If you are interested in helping out with any of these things or even just in poking around, please let me know over email — the address is at the top of the about page. Your thoughts can go here in the comments or to me in person.

To everyone who uses any of these products: thank you. It brings me great joy to see something I created help someone else. Words can’t describe how much I like hearing from you, how much I appreciate your feedback, your thoughts, your opinions and your critique. It has been an absolute pleasure, and I’d like to do what I can to make sure it stays that way.

It has recently come to my attention that the big Objective-C shocker on WWDC, ARC, is publicly released information.

ARC stands for Automatic Reference Counting and inserts release, retain and autorelease for you based on knowing about object lifetime conventions and deviations. If GC is “when you’re done, just toss them into that corner and I’ll clean it up for you”, ARC is some sort of robot frantically running around you, eerily knowing exactly when you’re done with something and snatching it. (Or something a lot less disturbing.)

Assuming it works and is well done, ARC is exactly the right solution for managing memory for Objective-C. Mind completely blown.

The world is getting several times more ridiculous every day. The Register reports on Apple “ripping off” a well designed app by Greg Hughes called Wi-Fi Sync that allowed for syncing iOS devices over Wi-Fi. It turns out that the comparison starts and ends with “something that syncs over Wi-Fi, is named the same thing and has a logo with the same ingredients”.

Let’s make something perfectly clear. When people ask Apple over a period of ten years (since the iPod was introduced) for “syncing” over “Wi-Fi” and Apple chooses to do this and name this feature “Wi-Fi Sync”, that’s not copying. Apple didn’t ride into town calling their dry wipes “Kleenex” when it was already taken; it was Greg Hughes that called his product dry wipes, and is now portrayed by the article as being astonished that other people can also name their dry wipes “dry wipes”. Literally the only thing the icons have in common is that they combine the two symbols for “sync” and “Wi-Fi”, oddly enough, that Apple has been using for years, where one of the symbols fits into the other in a visually obvious way. (Greg’s icon uses more elliptical sync arrows, whereas Apple is sticking to the standard circle shape they’ve been using since iSync and likely before that.)

There’s no evidence either way that Apple has or hasn’t copied Greg’s app’s technical approach beyond “yes, it does syncing over Wi-Fi“, because Wi-Fi Syncing isn’t even enabled in the first build of iTunes 10.5. Just following the trail of common sense, isn’t it quite likely that, whether you involve Greg’s app or not, Apple has been working on it for some time, possibly years, because people have been asking for it for years? That’s why Greg’s app found an ecstatic audience. It was also technically impressive because Greg effectively changed the transport layer without needing to alter iTunes, which is an achievement. The article mentions Apple asked Greg for a CV and I don’t blame them: of course they’re going to want talent like that.

I’m sure The Register means well, but I’m hoping that they just called Greg up to talk about Wi-Fi syncing and asked him how it’s funny that they chose the same functional descriptors both in name and icon, and he said “boy, yeah, I guess that is a bit funny” and not “they stole my life’s work”. When you invent something that should have been there for years and that something suddenly is there, there’s just not enough ground for outrage on the basis of originality. Especially when the only grounds so far for “copying” includes visually and in writing calling a spade a spade.