Entelligence: Stains on the sleeve of my operating system
Entelligence is a column by technology strategist and author Michael Gartenberg, a man whose desire for a delicious cup of coffee and a quality New York bagel is dwarfed only by his passion for tech. In these articles, he'll explore where our industry is and where it's going -- on both micro and macro levels -- with the unique wit and insight only he can provide.
I originally started this column on my take on what an Apple tablet might be (I literally dreamed about it and started to write it down when I woke up). I was really into it, which explains why I didn't save it as I wrote. I think you can see where this is going.
Like a cartoon character who notices that he's no longer standing on solid ground and suddenly begins to fall, I reached over to save, but was too late. My trusty XP install suddenly blue screened. Muttering just a few choice words, I rebooted, only to blue screen again. No problem, there's always "safe mode." Too bad safe mode blue screened as well. With little hope of getting anything recovered, I gave up, fired up my Mac and started from scratch. It's not the first time this has happened to me, where for some reason or another I've lost work on my computer. I suspect it's happened to a few of you out there too.
But this latest bad experience changed my thought process from Apple tablets to what's wrong with the whole PC landscape and today's operating systems.
On the eve of the Windows 7 and Snow Leopard launches (which respectively appear to be the best efforts yet from Microsoft and Apple), the basic platforms we all use remain products that are largely unchanged from concepts introduced decades ago, regardless of all their speed enhancements and evolutionary changes,. It doesn't matter if you're running some flavor of Windows, UNIX, Mac OS X or Linux -- this isn't about OS wars. Every modern PC and operating system retains the flaws of initial product designs that go way back to a time when compromises were needed because system resources were scarce. It's a little like the urban legend about a garment vendor who sent a designer coat overseas to be copied, only to get back 10,000 coats with an identical stain on the sleeve. Yes, PCs initially had to be told to save and it would have been ludicrous to try and track revisions on an 88KB floppy disk. But all that should be history by now.
There have actually been some pretty interesting attempts in the past, at least conceptually, to change things. PenPoint from Go used a unique set of system wide gestures to control the device and had an architecture to allow for documents nested within documents that could be zoomed in and out. Apple's Newton OS was one of the first operating systems to save everything created by default. The SwyftCard created by the late Jeff Raskin could take an Apple IIe and go from cold boot to where you left off in six seconds. (Hit the image for the full-size ad.)
Isn't it past time for hierarchal file folders that attempt to re-recreate my filing cabinet? Even with universal search, there has to be a better way to store and retrieve information. Why, for example, does almost every program force users to save their work? Going from CTRL-KD as the "save to" command in WordStar to Ctrl-S isn't much of an achievement. Why can't we build on a real-world model that keeps everything I create by default, and only throws away changes? You can even extend the metaphor and keep every revision, since modern hard drives are hundreds of gigabytes in size. Yet with release after release, vendors add features that no one uses or cares about -- a problem that reached a nadir with the Microsoft Office Assistant, which was so universally reviled that Microsoft actually sent out press releases announcing Clippy's death when it was removed.
Yes, I know there are settings to allow for things like auto-save, but why are the default settings for applications often designed to punish users who don't follow the programmers' rules or understand the interface? Don't get me wrong -- we're making progress, but most of that progress has been restricted to mobile platforms. webOS, iPhone OS and Android all have driven some degree of innovation beyond the PC desktop. Even apps are tracking that way: Tweetie, one of the most popular Twitter apps, debuted for the iPhone before it was available for Mac OS. Now we need to bring some OS-level innovations back to the PC desktop and take personal computing to the next level as well.
One reason no one has been able to unseat Microsoft from desktop operating system dominance is that no one has offered a fundamentally different product that actually changes some of these longstanding issues. Although there are arguably better products out there, the stains on the sleeve are still there, and it's time for some revolutionary innovation that will truly change the computing experience. What do you think -- are today's platforms really revolutionary or are there still too many throwbacks to ideas that are thirty years old?
Michael Gartenberg is vice president of strategy and analysis at Interpret, LLC. His weblog can be found at gartenblog.net, and he can be emailed at gartenberg AT gmail DOT com. Views expressed here are his own.
I originally started this column on my take on what an Apple tablet might be (I literally dreamed about it and started to write it down when I woke up). I was really into it, which explains why I didn't save it as I wrote. I think you can see where this is going.
Like a cartoon character who notices that he's no longer standing on solid ground and suddenly begins to fall, I reached over to save, but was too late. My trusty XP install suddenly blue screened. Muttering just a few choice words, I rebooted, only to blue screen again. No problem, there's always "safe mode." Too bad safe mode blue screened as well. With little hope of getting anything recovered, I gave up, fired up my Mac and started from scratch. It's not the first time this has happened to me, where for some reason or another I've lost work on my computer. I suspect it's happened to a few of you out there too.
But this latest bad experience changed my thought process from Apple tablets to what's wrong with the whole PC landscape and today's operating systems.
On the eve of the Windows 7 and Snow Leopard launches (which respectively appear to be the best efforts yet from Microsoft and Apple), the basic platforms we all use remain products that are largely unchanged from concepts introduced decades ago, regardless of all their speed enhancements and evolutionary changes,. It doesn't matter if you're running some flavor of Windows, UNIX, Mac OS X or Linux -- this isn't about OS wars. Every modern PC and operating system retains the flaws of initial product designs that go way back to a time when compromises were needed because system resources were scarce. It's a little like the urban legend about a garment vendor who sent a designer coat overseas to be copied, only to get back 10,000 coats with an identical stain on the sleeve. Yes, PCs initially had to be told to save and it would have been ludicrous to try and track revisions on an 88KB floppy disk. But all that should be history by now.
There have actually been some pretty interesting attempts in the past, at least conceptually, to change things. PenPoint from Go used a unique set of system wide gestures to control the device and had an architecture to allow for documents nested within documents that could be zoomed in and out. Apple's Newton OS was one of the first operating systems to save everything created by default. The SwyftCard created by the late Jeff Raskin could take an Apple IIe and go from cold boot to where you left off in six seconds. (Hit the image for the full-size ad.)Isn't it past time for hierarchal file folders that attempt to re-recreate my filing cabinet? Even with universal search, there has to be a better way to store and retrieve information. Why, for example, does almost every program force users to save their work? Going from CTRL-KD as the "save to" command in WordStar to Ctrl-S isn't much of an achievement. Why can't we build on a real-world model that keeps everything I create by default, and only throws away changes? You can even extend the metaphor and keep every revision, since modern hard drives are hundreds of gigabytes in size. Yet with release after release, vendors add features that no one uses or cares about -- a problem that reached a nadir with the Microsoft Office Assistant, which was so universally reviled that Microsoft actually sent out press releases announcing Clippy's death when it was removed.
Yes, I know there are settings to allow for things like auto-save, but why are the default settings for applications often designed to punish users who don't follow the programmers' rules or understand the interface? Don't get me wrong -- we're making progress, but most of that progress has been restricted to mobile platforms. webOS, iPhone OS and Android all have driven some degree of innovation beyond the PC desktop. Even apps are tracking that way: Tweetie, one of the most popular Twitter apps, debuted for the iPhone before it was available for Mac OS. Now we need to bring some OS-level innovations back to the PC desktop and take personal computing to the next level as well.
One reason no one has been able to unseat Microsoft from desktop operating system dominance is that no one has offered a fundamentally different product that actually changes some of these longstanding issues. Although there are arguably better products out there, the stains on the sleeve are still there, and it's time for some revolutionary innovation that will truly change the computing experience. What do you think -- are today's platforms really revolutionary or are there still too many throwbacks to ideas that are thirty years old?
Michael Gartenberg is vice president of strategy and analysis at Interpret, LLC. His weblog can be found at gartenblog.net, and he can be emailed at gartenberg AT gmail DOT com. Views expressed here are his own.


















I'll answer your question by asking another question : Why do we still use QWERTY (made for typewriters) when the Dvorak format is superior? Shit happens. And Microsoft and Apple will keep pumping OS's out.
Kearney: Hey Dolph, take a memo on your Newton: beat up Martin.
[Dolph writes "Beat up Martin" which the Newton translates as "Eat up Martha"]
Bah!
Nope. There's proof that Qwerty is better than Dvorak.
Because anybody who types knows qwerty
Only some people know dvorak
Switching would be like coming up with a complete new way of driving with laws, lanes, turn signals - everything. Do you really expect to re-teach everybody, including the elderly, and not have a major disaster on your hands?
Someone actually researched this: http://www.reason.com/news/show/29944.html
Switching from QWERTY to Dvorak is like America switching from Imperial units to Metric: It's not going to happen, at least not any time soon.
Why don't we all speak Esperanto?
(I do use dvorak btw)
Really we should all be using Caligraph (apparently):
[February 1889]: "In the Brooklyn demonstration he typed 142 words per minute in a five-minute test, 179 words per minute in a single minute, and 198 words per minute for 30 seconds. He was accompanied by George McBride, who typed 129 words per minute blindfolded. Both men used the non-QWERTY Caligraph machine" [from the Reason article]
Switching keyboard layouts isn't a good analogy for changing how an OS works. Keyboard layout is a personal choice, whereas changes in an OS would affect everyone. A better analogy would be switching from analog to digital TV.
@JWC
That article has been thoroughly debunked. Anyone who knows anything about how QWERTY and Dvorak compare can spot all sorts of errors just by reading it. The authors clearly used the Dvorak/Qwerty rivalry to highlight their economic theories, and didn't bother spending much time to make sure they got it right.
http://dvorak.mwbrooks.com/dissent.html
Qwerty was "good enough" for a day when converting a typewriter would have been extremely expensive. Now it's just crap.
I've meet dozens of one-time mac users switching back to Windows (heck, I'm one), I've never even heard of a Dvorak typist going back.
I have heard of plenty of dvorak typist going back.
about 5-6 of my friends have tried, and all switched their computers back. here are a couple of reasons which i observed:
1) they had to use qwerty anyways when they used somebody elses computer or a lab's computer
2) it was difficult to learn since they had to have an image on screen to remember where the keys were since they didnt want to buy little stickers. sure it can be overcome overtime, but it is a pretty large learning curve
3) when someone else used their computer curses insued and they had to rush over to change it back to qwerty.
4) their typing speed actually suffered since they hadnt fully memorized the layout. (although this problem would be overcome once they learned it, but again.. big learning curve)
5) everyone made fun of them.
another point...
the only real people that go to dvorak is people who actually want the extra speed. when 90% of the population cant break 30 wpm on qwerty, what makes you think that dvorak will improve their typing speed anyways?
"Why do we still use QWERTY (made for typewriters) when the Dvorak format is superior?"
I use qwerty because I have it available and I'm trained. I don't know why you use it, maybe you know. Clearly, qwerty is superior, since we use it.
@Clyde Berry (aka ignorant fool)
the US is pretty much officially metric. it's just that switching is incredibly difficult.
http://www.unc.edu/~rowlett/units/usmetric.html
Jason you just unwittingly proved my point.
Tried and gave up /= Mastered, judge inferior, and switched back.
For example in the Windows/Mac analogy, it takes awhile to get used to a new OS like a Mac. But very few people buy a mac and turn around and get ride of it in a few months. No, they generally use it for a year or two or three, and then purchase a PC as their next computer. You would never take someone's word for it who stuck with it for 3 weeks.
The very fact that people didn't finish, and switched back, proves exactly why QWERTY is still around despite its inferiority. As the standard, there are OUTSIDE COSTS associated with switching that have nothing to do with the inherent quality of the keyboard layout.
Every single thing you mentioned is one of them. I've never denied the costs. But I still proclaim the significant superiority of Dvorak when objectively compared to Qwerty.
There are several simple java applets floating around that straight up prove how much more efficient it is just by copy/pasting some text in.
BTW, I don't think Dvorak is the best layout out there. On a scale of 1-10, if Qwerty is a 5, Dvorak is a 9 (nearly double), and Colemak is a 10. So you could ask me, why don't I switch to Colemak? Well, the costs you mentioned are not worth it to me for the small gain, although I often recommend new typists learn Colemak instead of Dovark even though the costs are higher (Colemak isn't built into the standard OSs, so you'll always have to load it instead of just changing a setting.)
With my typing skills, I prefer an alphabetised keyboard.
From what I understand, the QWERTY keyboard was designed to slow down a typist to prevent the typewriter from jamming.
It's been known for a while that dvorak layout is better, but the qwerty keyboard is so ubiquitous that switching to dvorak would involve too many resources and too much wasted time (from people correcting their typos), and the returns will be minimal. Most people don't need to type 10 more words a minute faster (or whatever that number is). Sure it would benefit professionals, but even in their case it wouldn't be a good return on initial investment (and PITA).
Besides, there are tons of people who are using more and more mobile devices for most of their typing, and when you're typing with thumbs, I suspect dvorak or qwerty make little difference.
In any case, this first post with its entirety of comments have gone completely off on a tangent.
Metric System:
Few people in the US know how much a liter is. But everyone knows how much two liters is.
I tend to agree - QWERTY should no longer be around since its reason for being no longer exists. However, I confess that I'd rather have QWERTY rather than AZERTY. Whoever designed that keyboard layout needs to be shot since it requires you to hold down the SHIFT key for so many common punctuation marks.
There's no evidence Dvorak is better than QWERTY. There's only annecdotal evidence from people who have invested a lot of time in learning it (and confessing that you've just wasted a lot of time learning something useless is a lot harder than claiming that it's superior when nobody else can check), and military experiments which were overseen by Dvorak himself.
Did anyone actually read JWC's article all the way through?
It covers all sides of the "debate" yet you all still debated over it. In summary, neither provided adequate performance benefits from each other. It points out which studies are myths, or not of high scholarship, and which studies are.
@luke - your response to a summary of a scholarly article is a link to a domain with "dvorak" in the name and "dissent" in the filename? You could have looked a little harder to at least seem credible. The guy's declaration that the Caligraph layout doesn't count as competition because it doesn't meet his standards of being "different" enough is particularly charming.
This discussion = GAY
@ Luke
"But very few people buy a mac and turn around and get ride of it in a few months. No, they generally use it for a year or two or three, and then purchase a PC as their next computer."
Not wishing to turn this discussion into an OS-related one, but what on earth make's you think that?
Oh my god? Qwerty vs Dvorak?
Dvorak is no better than qwerty, the often cited navy study was run by Dvorak himself and the other studies done are either inconclusive, side in favor of qwerty, or were paid by Dvorak or a company marketing his keyboards.
It's sad that this urban myth is still ongoing. and the retraining takes time as well. so that makes Dvorak even less attractive. It may have a more efficient key placement but that doesn't necessarily speed up typists or reduce RSI.
I totally agree with the article that something revolutionary needs to happen in OS's. but Dvorak/qwerty isn't an applicable analogy. Dvorak/Qwerty is more like Windows/Mac slightly different paradigms that can each perform the same basic tasks.
And I really hope people aren't telling themselves that Linux is the answer. my favorite quote as of late is "Linux is free if you don't value your time". Something fundamental needs to change at the OS level. hardware is changing (core i7 gets rid of the FSB), OS should be following suit, innovating and doing away with things that are no longer needed (registries and dlls, maybe).
one of my other favorite quotes is "All computers suck, Macs just suck less" (if you want technical details why feel free to ask). I'm no Mac fanboy, but after living on Windows all my life, OS X is a real breath of fresh air.
http://www.reason.com/news/show/29944.html
read up. hard facts, with references that debunks the myth that Dvorak is superior and that we all use QWERTY because it was the first keyboard on the market.
I'm sick of hearing this qwerty vs. dvorak BS. people get as uppity about it as they get about the Mac vs. Windows debate (both are PCs, Personal Computers)
How many of you need typing speed above all else? Unless you write in English for a living, QWERTY probably meets your needs. I'm a programmer, I type for a living and I've never felt the need to switch (Dvorak offers no benefit for the operators, brackets, braces and semi-colon keys).
I don't believe you
Yeah I find it hard to believe that someone who dreams about apple tablets and judging by other things written by him is a huge apple fanbio would be using xp especially when he writes he has a macbook with him as well.
sounds kinda like BS, but i guess it's not impossible. his problem is that he's using XP. here's a tip: switch to Vista, or I guess at this point, install 7 when it comes out. even Vista is much more stable than XP. i think i've had Vista blue-screen on me once, and it was before service pack 1, because of a bad driver. Vista SP1 has never blue-screened no me, on two machines.
I've actually never blue-screened on my laptop's XP install, and the only times I've seen it on my desktop XP installation were all related to some form of incompatible or bad drivers. If you're getting a blue screen, you should look into it. Chances are, it's quite reparable. I just find it disappointing that someone who's supposedly tech-savvy thinks that the occasional blue screen is just part of the Windows experience, rather than actually finding the root of the problem to correct it.
You guys have all completely missed the point.
Yeah.... Pro versions of Windows have had file versioning for a while, and Apple's Time Machine does the versioning stuff pretty well too. There was also GoBack and related products. As for auto-saving, there's a few reasons why programs don't do this, but mostly it has to do with your intent. I often open up Notepad and Word just for jotting down quick things and getting rid of them later. Word is very smart in knowing that if I've kept the document open for a certain number of minutes to start auto-saving. This seems reasonable enough to me.
The rest of the article is just B.S. It's not the saving and information retrieval problems that are 30 years old--it's the interface paradigms. Most of the problems he's complaining about are issues with Computer Science fundamentals. Yes, innovations like WinFS will help, but NTFS serves admirably for most tasks as-is. There's only so much a computer can discern on its own (through algroithms, etc.) before you've got to step in and do the legwork yourself.
@ James
Proof that anyone who says anything negative about MS or postive about Apple is labeled a 'fanboi' within the Engadget commenting community. Why don't you do a little research on Michael Gartenberg and then tell me he is an Apple 'fanboi'
Its people like you that make Gizmodo look better every day.
Here, let me help you in case your copy of Windows bluescreened before you could Google his name;
http://www.techcrunch.com/2007/02/15/microsoft-hires-michael-gartenberg-as-new-evangelist/
Aah Skeezle the rare voice of reason in this cesspool we call the engadget comment system, where anything negative or insightful about MS's failings (as abundant as they might be) or Apple's accomplishments gets immediately down ranked, a place where an app on the iphone sucks, but the same app ported to Android is omfgwtfbbq best app ever in the world.
Its nice to see there is at least 1 other person who's mouth isnt full of ballmer. Now with your mouth full kids, say Developers, Developers, Developers!
Deberereors, Deberereors, Deberereors.
Were you using Word when you lost work? Whenever I crash it brings up a list of unsaved files that were open when it happened, and I simply select one and continue where I left off. This seems to be an answer to what you are talking about, and it already exists.
Even if he was, the situation he describes sounds tantamount to catastrophic hardware failure. In this case I am not sure what the OS is supposed to have done for him. Your reaction was identical to mine (I haven't had this problem in about 10 years since autosave has become ubiquitous). The flaw in this editorial is that somehow the hardware failure that is likely to be the true problem is somehow being blamed on the OS and applications that run on it.
It sucks when your machine dies, but blaming the OS for this is like blaming a builder for a building collapse when you yank out the dirt underneath the building. What was he supposed to do, exactly?
If the hard drive didn't fail (as in -- hopefully it was something else), then he should be able to find the autosaved file by slaving the drive to another PC -- should look something like ~xxxx.tmp in the folder he normally saves Word docs in ( like "DOCS" or "My Documents", etc.).
Hmmm. Well I agree a system that saved everything would be nice in the instance that your computer blue screened or beach balled to death, but then we're getting to the point where we take all responsibility away from the user. Say for some reason it didn't save, and a user is so used to just closing the program without saving, well now it's Microsoft or Apple's fault all over again. I also wouldn't want all my changes to be saved. Say I open a word document, is it going to automatically save what I type, even if I decide I no longer want those changes? Right now I could just close it, not save, and be back to where I was. But if it saved automatically...well that's a different story. Unless it was implemented in a way where it created new files, but then again I don't want to have to go through hundreds of files just for one document because it was automatically saved all the time.
It's an interesting idea and in the instance that your computer crashes on you it would be nice to just have your work back in your face, but I'm just not sure about it.
*Ideally* what would happen is this.
The program saves a copy after every x changes or x minutes. However, the previous versions are also saved. Preferably they would not be separate files, but rolled into the main document. so lets say you highlight the doc, then in a preview panel all your previous revisions would show up in a list, allowing you to select one.
Seems like the computer is tracking more and more of what I do. That's a good thing?
Google Docs anyone?
I think what he's suggesting is a system that uses journaling to keep constant backups of documents as you're typing them. They wouldn't appear as actual files but would exist in the background for the OS to handle. The oldest revisions of the documents would be over-written when it is deemed necessary(so they aren't really allocated for anything and will show up as empty space). That would make it so you can go to any past revision of the document without keeping millions of copies AND without needing to save.
Take a look at the Google Wave demonstration and its timeline. You want an older version? Just drag the slider backwards, and you see the document un-type itself.
Why doesn't every program that doesn't save a huge amount of data (anything that outputs text, xml, etc) have something like this?
Why do my digital camera's photos have to be called "DSC8805.JPG" and reside in /Users/Erik/Pictures/Camera/2009-05-03? Why doesn't the computer just *know* that I want to see pictures from my camera taken May 3, 2009, and they're just there, instead of having to box them away somewhere and then put a "layer" on top like Adobe Lightroom to do this tagging and sorting?
I really wish Microsoft hadn't neutered WinFS and then delayed it into the distant future. Even an absolute piece of crap version of this product would provoke others into improving on it.
They have had autosave in office since the late 90's in case your computer crashes. I know he is just trying to get his point across but lying about his computer blue screening to do so makes him a douche. He obviously doesnt know enough about technology to be writing these kind of columns.
evilspoons got it exactly right. Systems like what was demonstrated with google wave, or at the very least auto-saving with simple versioning, should be the norm. It may not be quite feasible for large image or audio files, but documents, emails, freaking text files (like source codes for html, js and css files for example) should all be continuously saved with a time line and periodic backups (just in case something gets totally messed up and even the previous versions in the time line aren't accessible).
And the whole file storing system needs a huge overhaul, and everybody knows that, even Microsoft. Unfortunately, seems that it turned out to be a bigger task than what everybody thought, and probably will take a few more years and maybe a windows version or two.
In any case, these kinds of fundamental changes take time, PC industry is too big of a ship to turn around quickly. Still, we know very little about Google OS - maybe it'll provide some of those things, or even all of them, and will serve as a kick in the butt to MS and Apple.
OpenOffice recovers even unsaved documents after a crash or power outage.
I think what the author is stating is interesting if it had some modifications. Autosaving revisions per application would be evolutionary instead of revolutionary, as backup software currently saves revisions as a whole. Saving would become 'tagging' a certain revision. I believe 'save as' would still have to be utilized, though, as sometimes I am working on a paper, and I alter it so far as to create a separate entity. I still want to expand upon the first revisions, but this later version has wholly developed into something on its own.