Monday, March 22, 2010

Things I'd Like To See In Mac OS XI 11.0.

In a recent TUAW blog, I stumbled on the opportunity to tell them (and hopefully the world) on what I'd like to see in Mac OS X 10.7.

So I upped the anté... instead, I thought I'd tell the world what I'd like to see in Mac OS XI 11.0 (that's Mac OS Eleven to those who keep pronouncing the "X" as Ex rather than Ten!).

Just in case it never made it to TUAW, I thought I'd reproduce it (with slight edits and elaborations) here. So, without further ado, here's my list...

  • No app should have a "Save" command. This does not mean that apps shouldn't save documents! Rather, apps should be intelligent enough to know when to save their documents on their own accord, or with help from the system via notifications. The system software could provide automatic document saving triggers, so that applications can save users' work on a regular-enough schedule, meaning that users will always be able to return to their documents in the same way they left it, without having to micro-manage their computers to make it happen.

    The Apple Lisa had exactly this capability back in 1983 with the Lisa Office System.

  • No app should have a "Quit" command. This redundant command was a relic of the original Macintosh System Software (before MultiFinder), where users had to tell the computer when they wanted it to return to the Finder. Now that the Finder runs concurrently with any and every app on the system, and because the Darwin foundation provides an abundant amount of virtual memory, the Quit command is no longer a necessity. Users should only have to close windows of an application, and the system should decide whether or when to quit the application. If the system detects that the application is small enough, it could quit the application on the closure of the last open window; if it is rather large and is known to take up significant time to start, or the app runs services that other apps can use, then the system can opt to delay the shutdown of the command unless memory conditions indicate otherwise (usually never). If users really need applications to quit because of memory or performance issues, then the Finder or the Dock can tell the applications to quit via some user interface (such as the user interface controls that exist already).

    System 7 introduced a feature which allowed the system to be smarter with regards to memory management, where if a user launches an application for which the system has no memory left for it, the Finder would ask whether to automatically quit applications with no open windows, or suggest that the user quit the largest application currently running. While Mac OS X doesn't have any of these memory management issues, it could still be made to provide intelligent software management to the point where the Quit command, and the very concept of quitting applications, could easily be dispensed of.

    Update: iOS 4 has actually introduced the concept of automatically managed apps using its multitasking APIs and system services. Apps quit only when iOS thinks they should quit, otherwise, for iPhone 3G S and better devices, apps stay in memory and when they are a background process, the process is paused unless prior arrangements are made to obtain execution time from the kernel. This does away with any explicit "Quit" command, and better manages the system for efficiency.

  • Component software and document-centric apps should come back to Mac OS. Document-centric applications are (I'm guessing) probably easier to develop using the Cocoa frameworks than with any other software technology, thanks to Objective-C and its dynamic messaging features. Document-centric applications allow users to create content based on "parts", which can be controlled by separate "editors" in a system-supplied document shell, which in itself is expandable using a common plug-in architecture. Users can then choose the parts required to compose the documents, rather than rely on applications to provide every conceivable feature or interface to the system or other applications, and the part editors can be scriptable with AppleScript. Of course, not all applications may fit in an OpenDoc-like model, but I think many desktop applications can benefit from the architecture if developers work with it carefully. Pages, Keynote and Numbers could benefit from a component-oriented design where these three features could be interchanged amongst a common document shell, and provide the live-editing features that System 7 used to have with Publish and Subscribe in System 7.0, and OpenDoc 1.0 in System 7.5.3.

  • As an option, the App Store could provide software for the Mac. Here, I ought to emphasise that I am interested in the on-line software delivery and auto-installation mechanisms that applications could benefit from rather than a digital rights management service for Macintosh software. Applications for the Mac are mostly easy to install (about as difficult to install as Dashboard Widgets) so an automatic installation mechanism for a new generation of Mac software could facilitate this as an on-demand service if the user so chooses. Meanwhile, software that requires Installer intervention, or applications that are just drag-and-drop installed by the user to the Applications folder, or are built and installed by UNIX tools and facilities, should still be allowed for various, but hopefully obvious, reasons.

    Update: the Mac App Store is actually coming!

I have also included some things that could come earlier than Mac OS XI, because they are within reach of a 10.x release:

  • Apple should stop playing with user interface prototyping in their apps and have them all conform to their own published user interface guidelines. And I should highlight that I am not talking about bold user interface changes like iMovie 8, but rather, the implementation and re-implementation of Aqua standard controls! Many things, including Inkwell, are broken because of this, and it is high time that Apple should be setting an example for other developers rather than go their own way because they can! This goes for their Pro apps, too! Apple should clean up the Aqua user interface some more, and then stick to it, and make applications use standard system behaviours that allow all apps behave the same way regardless of the input technologies used (touch, electro-ink, mouse, whatever)! Cocoa Touch should be the foundation technology towards this effort (ie., any features solely based on mouse-move events should be given a make-over)--the iPhone OS already has a head start on this.

  • And speaking of Inkwell, the Newton OS 2.0 still has better facilities for written input than Mac OS X ever did! Inkwell should be improved to support either touch or pen computing, with more facilities for editing electronic ink-written text. Some more things that also come to mind include gestures for correcting text outside the confines of Inkwell's notepad (where possible), in-document word suggestions and character-by-character text editing, which Inkwell does not support at all on the Mac.

  • QuickTime X should be made to play Flash content. Hang about... doesn't QuickTime 7 already do this?! QuickTime for both the Mac and iPhone could play Flash content if Apple really wanted to, and this is their best opportunity to support what people want, on their own terms. If something is too taxing to play back, QuickTime could say so, and give users the opportunity to select whether they want the Flash content to be played back or not by using an icon/button in the region containing the content.

    Update: it appears that Apple have discontinued Flash support in QuickTime as far back as 2006! QuickTime 7.1.2 was the last version of QuickTime to support Flash tracks without user intervention or at all, making this elegant idea unlikely to ever appear in reality.

These are my pet peeves of Mac OS X to this day, not counting bugs or poor implementations of existing concepts. The thing I miss the most is document-centric computing, where the computer could mix-and-match live data types onto a common document shell. This is something that I consider the most powerful paradigm for personal computing, which was abandoned rather abruptly in Mac OS 7.6. I have never seen a replacement technology for Publish and Subscribe, let alone OpenDoc, since the end of System 7.

Article last updated on 29th October: cleaned up the fluffy writing; correlated a few of my ideas with recent events.


—tonza

Labels: , , , ,

Sunday, March 14, 2010

But... It's So Grown Up!

I glanced upon this AppleInsider article which quickly touched on the fact that Apple TV has taken back seat to the array of iPod devices that are a part of Apple's iTunes ecosystem.

That made my enthusiasm for the device sink slightly, just because it was reported that Apple's attitude to the product is "it's only a hobby, for now."

I have tried quickly to find an Australian movie rental outlet that rents Blue-Ray discs, but alas, they don't seem to be popular.

I thought that Apple TV has indeed picked up in a lot of areas where it fell flat on its face beforehand. Yes, it's not as good as getting yourself a Blue-Ray player and/or games console and watch overpriced (either rented or sold) discs on your HD television set, but on my television set, it'll do. But I don't really use Apple TV for that.

I honestly use Apple TV as the only other device other than a Mac or a PC running iTunes as a music player for my home-wide audio distribution network using AirTunes. Listening to music around the house without dedicating a computer for the job is really the best thing next to having some other dedicated distributed homeentertainmentsystem do the job instead. But not only do I get the luxury of having music distributed throughout my home without the help of a dedicated computer (especially when it is expected or in danger of being restarted!), but I also get a device that can play material I have recorded with EyeTV or obtained through the iTunes Store on my television set. It's sheer liberation!

Actually, you need at least two devices for AirTunes: the Apple TV and at least one AirPort Express base station.

Where I wrote that the AirTunes functionality was worth the price of the Apple TV, I really meant it—it costs about the same as an iPod classic, and the Apple TV can do more than what the iPod classic and its Universal Dock can do. I use the Apple TV more as a music player for the home than as another avenue to watch videos on. But given that, I still do watch some videos, particularly the ones that would prove more difficult to watchany way else. And what other system allows you to play back Internet Radio throughout your house with the one device?

The Apple TV's Achilles' heel, however, is support from the iTunes Store. You need to feed the Apple TV somehow, and this is the easiest (but not the only way) to feed the device's hunger for data. At this time of writing, after using an Apple TV for a few months now, I can definitely say that unless you can get by with productions from America, and to some extent, Australia and the UK, you are probably going to get sick of the iTunes Store rather quickly. There is no provision for content from other countries, ie., "foreign film" or "world cinema" unless they have somehow made it to major American markets.

While I may have appreciated what Apple TV currently is and where it currently stands in the marketplace, I can also say that people expecting the very best in content delivery with Apple TV are going to be disappointed—currently, the iTunes Store doesn't stack up compared to going to your bricks-and-mortar video library or electronics store and selecting your favourite videos for your DVD player. So in my sincere opinion, this is no time for Apple to treat Apple TV has a hobby! Thankfully, as a product in its own right, I think the Apple TV is a very useful unit as far as the iTunes Store (or your resourcefulness) allows, and I hope that Apple are aware of this and have some surprising plans for us Apple TV users in the near future.

And if whomever are responsible for the iTunes Store in Australia could actually allow Australian Apple TV users access to:

  • more TV shows in HD quality rather than 576p, and
  • actually allow customers to purchase HD movies rather than just rent them,
that'd be great, too! It's all too often that Australia falls behind in the availability of commercial resources for products sold here, and Apple's iTunes services are no exception.


—tonza

Labels: , , ,

Tuesday, March 09, 2010

Linux Desktop Originality is Indeed Dead.

When are computer software developers, whether open-source or commercial, and/or user interface designers, going to stop ripping off from existing designs and come up with their own insanely great alternatives that are not a derivative of Microsoft Windows or Macintosh?

Well, OK, IT history is full of repeat presentations of the same concepts and designs in computer-user interfaces:

and the countless Linux distributions that are hell-bent at applying either Windows or Mac OS X user interface design elements into GNOME or KDE themes.

Ubuntu apparently has a new theme, Lucid Lynx, as reported by Ars Technica... one that, again, has been ripped off from another platform. This whole "let's take the most popular desktop user interface and make ours look like it" thing is starting to get on my nerves, only because what I think ought to happen within the GNU/Linux community is that a survey be done on how people want to use their machines, and then have a think tank plan and design a user interface based on that feedback.

User interface design is never easy. It took Apple over 3 years to perfect their user interface designs since the Apple Lisa and early Macintosh, and over another 3 years for Xerox to come up with their own for the Star systems they created prior. Linux integrators and system builders who had anything to do with user interface design on Linux, such as the then Sun Microsystems, have simply taken other user interface designs from other systems (namely Windows) and incorporated them into their own in the form of GNOME or KDE.

Sun's feasibility study and user interface research, unlike Apple's back in the early 1980s, was reliant on a group of test subjects who were supposedly not familiar with computers, just like Apple's research candidates were over 10 years before, but somehow, Sun's designers were influenced on applying starkingly similar design traits and styles based on Microsoft Windows. The result is apparent: GNOME 2 is a Microsoft Windows rip-off, and it doesn't look like the trend is going to die any time soon as PC market share shifts more in Apple's favour every month.

I cannot fathom how a research group within Sun could have claimed independent research, yet come up with this awful result! I would have thought that for such a then talented company, they could have at least used what they already licensed: AT&T Open Look. But they didn't. Instead, they were obsessed at drawing from the pool of Microsoft Windows users and replacing GNOME as their primary desktop user interface. And despite all that so-called independent, clean-room research, all they had ultimately came up with is a Windows clone.

I think Sun, and the original development teams at GNOME, got either lazy or paranoid, and didn't want to believe that they could come up with an original, better user interface design that could be more useful and easier to use than Windows. Instead, they thought that they could obtain market share on the desktop by covering a wolf in sheep's clothing.

In contrast to what resources and preconceived ideas Apple had worked with for the Apple Lisa and Macintosh, they didn't have any competitors to worry about, and people didn't even know what a desktop user interface could ever be, so Apple indeed had a clean-room to work in, with pristine test subjects. The Lisa and Macintosh desktop user interfaces were not at all remotely similar to the then-established ISO Motif desktop user interface, which gave Apple the best opportunity to come up with ways to easily, flexibly, precisely and quickly control and monitor their personal computer systems.

I don't think that anyone in the IT world, unless they are strictly disciplined and highly artistic, can come up with a completely new user interface anymore, since there are so many popular computing gadgets that have biased so many software developers' minds in terms of both design and acceptance in the marketplace. However, Apple has proved that it is possible to come up with a unique and highly appropriate user interface unlike any other, and that made its appearance in their first commercially sold iPhones.

Yet despite this demonstration of possibility, others, such as contributors to GNU software (and the links provided here are not exhaustive!), Canonical, Symbian, Palm, Google, and even Microsoft, have also proven that it is immensely difficult to create an altogether "new" or otherwise a replacement user interface design without some degree of plagiarism. These companies have only managed to make "copycat" designs based on prior art, mostly from either designs created and employed by Apple's Mac OS X and iPhone OS, or from designs popularised by Microsoft Windows (which in turn, also has taken graphics and imaging bites out of the legendary Apple, as well as other places). The only thing common amongst all of these companies compared to Apple is that they all have Apple as the leader in industrial design, and they are trying to infiltrate, by stealth, the markets where Apple have the most artistic influence.

The question is, why is it that none of these major computing and electronics manufacturers are conjuring up new designs that don't take looks-and-feels from others? Is it a matter of guarantee, where the clueless in the industry at least have a fighting chance at being accepted in the marketplace because their products now bear some resemblance with the leading designs and methodologies they propose to compete with, or is it a matter of a distinct lack of courage, where they're just not willing to gamble such high stakes on their opinions of new designs and technology, in the hopes of proving they have something better in computer user interfaces?

Unlike Sun's interesting but unpopular attempt at a unique user interface design for Java prior to the Swing project—called "Hot Java Views"—Sun's research into creating GNOME for OpenSolaris, GNU/Linux and the Java Desktop, was coerced into maintaining familiarity with the status quo, including Sun's own Swing user interface in Java 2. This was the project's downfall, in my opinion. It prevented what GNOME could have been... a fresh idea, a new look, a better, easier way of doing things on a computer. Canonical's contribution towards any sort of goal towards the transformation of Linux from being a Windows clone has now made that focus shift towards Linux being a Macintosh clone rather than Linux being a system of its own.


—tonza

Labels: , ,