Are Today’s Operating Systems Behind the Times?
Are the latest operating systems really delivering the goods or just a lot of nothing much coated in eye candy?
Vista is continuing to get bad press—and not just from traditional skeptics, but from sources you wouldn’t normally expect. And when Apple fully revealed Leopard in June, there was a general feeling of letdown. People are still hoping Apple will still deliver some wild new enhancement in Leopard.
Have Vista and Leopard fallen short of what they should have delivered? Or is it that our expectations are too high?
Leopard
Each iteration of OS X has seen a raft of improvements—usually numbering over 150. However, among those, not many have been truly significant, if any. None have significantly changed the way we interact with computers.
Some like to argue it’s the changes under the hood that are more significant; but does that justify the dearth of true revolution in OS X over the last seven years? In the last seven years, in relative terms, OS X has progressed no more than Windows. Consider that both operating systems still work off file systems at least 25 years old—with a bit of spaghetti and eye candy patched on. In fact, many long time Apple users still consider Mac OS’s Finder better in many ways than OS X’s.
It doesn’t help Leopard’s cause that Steve insinuated the features not yet revealed were so significant they were top secret, lest Microsoft would stop the presses to include them in Vista. Bill must be really kicking himself that he didn’t get to put a 3D tab bar in Vista…
Vista
Word on the street has not been good for Vista. This is seriously echoing the Windows ME debacle. In the latest news, we read Gianfranco Lanci, head of Acer, saying, “The entire industry is disappointed by Windows Vista.”
For all its good looks, it doesn’t take long with Vista to see its Windows 9x roots. And even on its looks, it comes as a bit of a shock to see the facelift doesn’t extend much beyond the desktop. You quickly find yourself feeling like you’re just using an upgraded Windows XP.
Both Leopard and Vista are nothing more than incremental upgrades. Apple admits as much, with Leopard taking OS X from 10.4 to 10.5. MS is trying to convince us Vista is more than XP SP 3. However, all the really significant improvements were in Long-gone, um, Longhorn.
We’ve all written and chortled about what Microsoft left out of Vista, but Apple can’t make much claim to including anything significant. Time Machine, for example, is nothing new.
As Chris Seibold pointed out yesterday in his article Five Obvious Mistakes and How Apple Made them Work, what Apple does well is take something and make it easier to use. Time Machine borrows from a feature that Windows has had for several years, but makes it significantly easier to use.
Admittedly, usability is a key component of operating systems, but sometimes the user friendliness and eye candy glosses over the weaknesses.
Today
The last truly canyon-leaping change in personal computer operating systems occurred 24 years ago when Apple unveiled Lisa and its GUI based OS. Everything since then has been incremental.
Who knows, maybe as Apple users we come to expect more. Look at the iPhone; there’s a massive leap in phone operating systems, and for that matter, any operating system.
It’s probable, though, that the next 10 years—with multi-touch technology, voice interaction, knowledge management-like file systems, and so on—will be when we really see operating systems evolve into something unrecognizable from today’s incarnation.
But today, are OS X, Windows, and for that matter Linux, truly where they could and should be?
del.icio.us
Digg
Facebook
Slashdot
StumbleUpon
Comments
Well, what do you want them to be? A lot of the industry is self-propelling anyway. It only takes so many MHz to surf the web and work some office software. Yet what we see (apart from the money extortion system that is PC gaming) is operating systems eating up more and more power, requiring hardware upgrades for, as you say, very little benefit.
Apple seems to have put the priority on stability (OS X over 9), then accessibility (Bonjour, Spotlight), and now security (Time Machine). I think there is still a lot to improve before they can go ahead and reinvent the wheel, and I’d take a well working, stable OS over a newfangled thing every day.
Applescript + Automator = perhaps the best of the best in Tiger & Leopard that XP/Vista will never have. Think of the time and $$$ saved by professionals just by this awesome feature. The Finder desktop and 99% of all applications’ workflows can all virtually automated.
Take that Active/VBScript! I wish you were as flexible and as easy to manipulate.
So, Chris, are you saying the move from OS9 to OSX was insignificant? How about OSX’s agility that allowed Apple to move the entire platform to Intel x86? And now, ARM on the iPhone?
I beg to differ from this view that the OS *MUST* be new every few years. You are reminiscing the relic of the 1980s when Atari, Commodore, Tandy (remember them?), Timex, Sinclair, and even Sears (OMG!) were releasing new platforms every generation that obsoleted early adopters. Do we really want to go back to that era? I think not.
What we really need is *PROGRESSIVE* improvements to what already works. If there are kinks (and there may be hundreds or thousands - as in Vista’s case) then the vendor must work 16 hr/days to fix those and not just patching a band-aid on the problem(s).
Leopard and Vista offered such progressive improvements but the former outdoes the latter in that it will save adopters more time, hence more buck$$! Vista offers the adopter better looks but not much more. Vista may even be an impediment to productivity since you will be fighting mad about driver compatibilities. I warn thee from first-hand experience.
I disagree with your statements: “Both Leopard and Vista are nothing more than incremental upgrades. Apple admits as much, with Leopard taking OS X from 10.4 to 10.5.”
Tiger is hardly an “incremental” upgrade to its predecessor, and Leopard hardly seems “incremental” to Tiger. I think Apple has been using an incremental “dot” numbering scheme only to allow the retention of the OS X moniker. These upgrades, however, are substantial enough that Apple COULD have numbered them as 10, 11, 12… [X, XI, XII…] etc.
I think everyone in the computer world is waiting for the “next big thing” in a desktop OS.
But we’re not likely to see a paradigm shift like we did when we moved from the command line to GUI for probably quite a while. We’re pretty much entrenched in the current schema, and it’s hard for a large company like Apple or Microsoft to make that kind of seachange when they have to answer to shareholders.
To obliterate the desktop metaphor would probably require a wholly incompatible new OS.
Before anyone cries “Linux could do it!”, I also think that is wrong since Linux developers are also stuck in the desktop metaphor and what is required is a complete ground-up development of a new OS and file system.
Here’s what I would do if I had the venture capital:
Throw out the filesystem as it exists today. No more heirachies of folders. No more navigating to a file. The new “file system” would be fully database driven with keywords, meta-tags, etc. Perhaps with some broad categories at the top level “Applications, Documents, Music, Photos, Television, Movies”.
The system could store the mime-types (so you could search for files by kind, like PDF, word doc, etc) and when a user saves a new file, it would prompt for tags, or keywords in order for the user to search for the file faster. The file system might even have a built-in keyword auto-summary. You could add full revision control—every document save could create a new revision so you could roll-back to earlier versions. Auto-save could ensure you didn’t “forget” to save a document and create revisions every X minutes (assuming there were changes to the document).
When you want to lauch an application, you could select from the full list, or if that’s too large, type in the first few letters or a meta-tag, like “DVD” which would show me all applications with in that metatag (home movie editors, video players, rippers/burners, etc).
Opening documents would be similar. Show me PDFs with the word “Apple” in them or as a meta tag, or Word docs, or all docs… or just search by the name. Etc. All user actions are basically a search on the database for files or applications.
I think this is vaguely the direction Apple is looking with its Spotlight/iTunes/Coverflow implimentation of the Finder but…
The difficulty is that for it to be truly a new OS/Finder, the hard-disk-as-folders-heirarchy would have to be blown away completely which would break every existing application since the Open/Save dialog as we currently know it would need to be totally replaced.
However, if you look at iTunes today, you can see the seeds of this type of interface being planted in consumer minds.
Believe it or not, most technologies plateau. They reach a point where any additional improvements are either deathly expensive or are ergonomically unworkable.
Handheld calculators have been static for at lest 20 years. The keyboard layout much longer. Me, I’ve been pretty disappointed with the total lack of innovation in the wheel. (Though pre-Columbian civilizations built magnificent cities without them.)
I’ve designed the solution but I can’t take it anywhere without some startup investment.
http://www.ihoby.com/thebackdrop/
Yes, it seems that the focus seems to be on fine tuning rather than next steps. That can be good in many ways.
That said, I think the next evolution will happen in a distributed/portable computing manner, which may be outside of an individual operating system setup.
For instance, perhaps all our data (including mail, documents, photos, & music) will be stored on .Mac (including our time-machine backups), and we’ll synchronise a smaller portion to whatever computer we’re using (whether it’s a PC with all files, or an AppleTV or iPod which synchronises mainly multimedia).
Is that an OS level ability, or something else?
Take a look at the OS in parts, then you’ll see the significance of each. You’ll also see vast differences and similaritys to others between systems.
Mac OS: Has the ability to grow, and very fast while retaining stability as it can progress through controllers rather than directly interacting with other “drivers”. Leopard, from what I’ve seen playing with it, isn’t drastic (YET) but it brings improvements we need. It’s stable enough for Windows users in Development versions, so final cuts should be solid. I’ve only crashed an app twice, OS never went down. I agree a database type FS would be nice, but active indexing with a data scheme backlay would do the trick and would only take a few hundred lines to implement… Or is it already in leopard… Can’t say. Spotlight has taken care of indexing and from hand on use it doesn’t hesitate anymore in leopard…
Linux: See Mac OS. However these distributions need to standardize or they will fall FAST. Package management needs 1 system to make, find, install, optimize, configure on ALL distro’s. Otherwise your stuck compiling from source and that’s great for those that know but the other 99.9% just want Clink-N-Run.
Windows: This system is way too far gone to be salvageable. They had it right in 01’ when they started from the ground up, in 04’ they were still looking at a 09’-10’ release schedule, this is not good for earnings. So an updated XP with SOME longhorn code tossed in. They upped security by making a dialoged interface to put blame on the user rather than watching the proccess, Protected Memory with a monitor would have been the fix however no time left.
The main problems with all OS’s of late is hardware. As an OS developer they need to remember to leave some system for the applications we need to run. The OS handles the hardware and the software interactions, that’s it’s job. OS X handles the RAM like no other out there, but it’s not perfect. Linux can use more Filesystems than most but some applications can’t run on it then due to the lack of controllers to handle Software-Hardware interaction.
So it’s give and take. You give it some considerations, and take stability with ease of use in OS X. You give up some stability and get more flexibility with Linux. And you give up both and get popularity with Windows.
Nice article.
I could not agree more!
As a software architect and power user, I find that I’m often asking myself why aren’t Operating Systems, or the software they run, better. I think the majority of the time, the answer is due to a far-reaching legacy constraints. For instance, many systems were designed with now antiquated technologies or techniques. Business requirements like backwards compatibility only make the situation worse.
It’s too bad that initiatives like Taligent or BeOS failed. Software is complex and making economically successful software is difficult. However, someone really needs to design a new system from the ground up using modern-day technologies and techniques. You can use virtualization for backwards compatibility.
The question is, do we really need more functions in our systems? Can we use the current one in 80%? With every release of MS Office there is a new functionality but does it change the way we use it? Probably a bit. Why don’t we focus on maximising the great potential that is out there instead of waiting for new bells and whistles? We all like it - read tech news every days waiting for something cool and exciting but is there a point?
I find it to be insignificant. Perhaps in the Mac world it was significant at the time of the change, but compared to everywhere else, not so much. OS X is based on UNIX, which is 30+ years old.
It was significant in that Mac users got an OS with protected memory and true pre-emptive multi-tasking, but in terms of the user interface, not much has changed.
We got a dock, but there were a number of dock/app launchers during the OS 9 days.
Basically, we still point and click applications or documents and we’re still in a disk-and-folder centric file browser, instead of one that is user and data-centric.
From a user standpoint, not much has changed since we gained the advantage of running more than one application at a time (some of us remember the days before Multifinder!)
The Desktop metaphor is now almost 24 years old. The heirarchal file system is even older. In terms of the technology industry, these things are dinosaurs… The difference is we don’t have a metaphorical meteor coming along to allow the next idea (mammals) to take over.
It’s kind of amazing how things don’t really fundamentally change. We still drive cars the way we’ve done it for a hundred years - a steering wheel, some configuration of foot pedals, all on a chasis with four wheels. Designs come and go, bells and whistles are added, but it’s still a car as we know it.
Is that legacy? Or is it that practicality of form around the human physique gravitates to certain conventions?
So if the practical conventions of a computer encompass a keyboard and a screen, then we’re somewhat limited in our imaginations about what an OS can really offer. Minority Report is probably out of the question for every day users - you think carpal tunnels is bad? Try waving your arms around 8 hours a day flinging files several feet through the air. Such a paradigm seems confined to a tiny mobile device like the iPhone or a limited-use device like the MS Table.
Yes, there are certainly ergonomic considerations, but the keyboard and mouse aren’t really optimized for those either.
I don’t suppose we’re talking about replacing the keyboard and mouse completely, but more like enhancing them.
I think Google understands. It’s all about search… Whether on the desktop or the internet, you want to find information faster, easier and with fewer keystrokes and mouseclicks.
What we really want is the Star Trek-esque search:
“Computer, find me all songs that were in the Billboard Top 100 that include Donald Fagen as one of the artists”
or
“Find all files and email where I reference [this client, phrase, keyword] ... “
Natural language search, whether spoken or typed, is really the operating system Holy Grail (since normal non-computer people really don’t understand boolean logic), but again we’re talking about a file system that would be based on a meta-data database… not one organized heirarchally.