Stay on Top of Enterprise Technology Trends
Get updates impacting your industry from our GigaOm Research Community
It occurred to me tonight that OS X isn’t getting any new features in version 10.6 of this amazing operating system. It really sunk in. You know, geeks hear these announcements at WWDC or in the blogosphere and its only when there is enough time to reflect before it sinks in. That’s what happened to me anyway, and if you’re thinking the same thing then the next question applies. Does no new features bother anyone? I think I said a while back that 10.5 ‘could be the bastard child in a a long line of fine operating systems’ and thankfully 10.5 has been pretty good to the world. Granted it had a rough start, but Apple has delivered more significant updates than its peers in the same business. 10.5 is certainly far from perfect though, and it’s still good that the strategy moving forward is one about quality and design.
It finally sunk in for me what no new features meant. Being greedy and always wanting more, I pause in moments to remember a favorite quote.
“Perfection (in design) is achieved not when there is nothing more to add, but rather when there is nothing more to take away.” — Antoine de Saint-Exupery
Yet I still can’t stop thinking of ways to make 10.6 better. Sure we all suspect major changes in the way we interact with computers using a multitouch platform. All of the multitouch goodness on an Internet device, on my desk or in my pocket, and one that uses standards-based browsers has me very excited to be alive and geeky.
I thought of a few things tonight can apply to either OS X now and moving forward. For your enjoyment or resentment, feel free to read on. Bear in mind there is no ‘target’ for when these ideas are possible. Let it stay imaginative, creative, and fun. This is just my idea, free and clear, only because I hope that it makes sense and its attainable to do.
The way we install apps on OS X is stupid. There is a fairy simple set of changes that could dramatically simplify how installs and uninstalls function while making easier packages for deployment. Get rid of all thoughts on 4 current ways it is done now for a moment.
The Applications folder is where we want most things to go. By and large though an application can live anywhere in the file system and run, with exception to the Trash. The new installer method adds a check box to Get Info on a folder to mark a folder as an Application folder. This action requires authentication to change. There can be as many application-marked folders as desired. When an application installer is completely downloaded into ~/Downloads (i.e. Downloads folder as preferences provides), the Finder switches focus to this file (with an on/off option for setting focus in Finder preferences). When the user drags an drop the installer package onto an application folder, the package installs. Of course some Applications require some additional steps and there is no reason I can see why the installer bundle couldn’t run a configuration/installer when the user drops an app in this folder (or opens it for the first time). Microsoft Office 2004 is a good example of a drag-n-drop install app that adds files elsewhere in the filesystem during first launch. In other words, when ‘MyApp.app’ is dropped onto an Application folder the Finder looks at the package and locates a package installer, determining where additions files may need to be installed (adhering to file permissions of course) and any configuration UI may run with permission. The user completes any optional configuration steps and MyApp.app is ready.
Removal/Uninstall is done via a bit of XML containing support file locations. This list is read when the file is removed from the app folder and put into the Trash. A folder marked as an apps folder is able to identify legacy installer methods and proceed to process them in the background.
Benefits of this are enhanced security by identifying a folder location (outside the defaults) App bundles may execute. It unifies package management via drag and drop while still providing support for current methods. Lastly it provides a complete uninstall method with a developer defined external file removal, with (power) user editable XML files.
Yes yes, it’s unavoidable. The smiley face that makes you want to throw a monitor across the room sometimes. My personal favorite is “An error occurred -10060”. But in so many ways the Finder’s ideas aren’t even the fault of Apple. It goes further than that, back to when things went all wrong in the world of UI’s and computer sciences. Files are bad. Folders are bad. Look around the house and its pretty certain that the majority of us don’t have all the paperwork in a folder called “/” and all subsequent subfolders below that with symlinks and alias, references and pointers, or whatever you want to call them. Bottom line is, people don’t seem to think in this rigid sense the industry has forced things into. So I propose we get rid of files and folders.
Is it that radical of an idea? I don’t think it is, considering on my iPhone I never deal with files as a user. I have applications that store my data, but I certainly didn’t have to save them in some arbitrary location I need to remember later. I see many users say “Oh it’s in Word”, when there is a real separation between the file (a document) and the file creator (Microsoft Word). I still think there is, and needs to be separation, with these two things things. Attempting to simplify file management from a users perspective I realized its not important to create a hierarchical and logical folder list. Instead bundles contain the data an application creates. When an application runs, it reads the bundle and the user can search or choose from the last n# items opened. If something the user wants is in another bundle, then there is an application associated with it.
The 1-2 punch on these ideas
Moving forward though, if an operating system were organized around Application and Data it could be more secure. Application-containing areas and bundles that contain the users data are clearly marked. The OS X system in mind would only allow execution in these defined areas. Backups of specific data would mean copying 2 bundles. The simple finder realized and useful to even the power user. Think about it, most usage on a computer is about launching applications, either via the file or the file’s reader. Bringing the focus to these ideas and building from there should make a very easy to use and mature OS. In many ways, applications in iLife behave this way already. iPhoto and iTunes doesn’t require sifting through a myriad of folders to retrieve data. What if other applications adopted this concept? Seems we’re headed there anyway.
That’s my idea, love it or hate it it’s free of charge and open to discussion. What’s on your mind for OS X polish?