Slashdot Mirror
Dvorak: Linux too much like Windows
inode_buddha was among a handful of folks who submitted linkage to Dvorak's latest column where he talks about Linux being to much like Windows. It's not really a slam, just a challange to be more innovative and look beyond feature creep and UI concepts that are old and tired. Hard to disagree with most of it.
... Believe me, buddy. I wouldn't be using linux right now if it wasn't quite as good as Windows. Windows came with this computer, and I'm not using it. That isn't because I'm some kind of linux religious freak. It's because I'm more productive on a linux box.
The same old command line? Somebody go tell this guy that linux (or any unix variant) doesn't have the same old command line as Windows. It's so obvious that they are different that I'm not going to type about it anymore
I'm getting the feeling that linux and windows are the same because they both run on computers. So they must be the same, right?
Sex - Find It
I like my WindowMaker. It's not a Win95/XP clone like KDE and Gnome tried to be. But they aren't fully Win95/XP clone that they tried for either, they all moved on. Gnome has multiple panels, as does KDE(ok, they keep up with each other instead of diverting, to me that is kind of pointless), as does Windows. But with Gnome and KDE is makes more sense to use the multiple panels, with Windows there really isn't a reason except to make it look better.
I do agree with Dvorak that WIMPs is a bad idea, but I do think that it is one of the best concepts out there. Although I don't have icons except when I minimize a window. What I would like is a scrolling desktop(and a CPU that could even support it if I coded it). I want to watch my MPlayer Window _over_ the Mozilla Window, but if I move the mouse towards the scrollbar(where MPlayer is covering), the Moz window would move over or the Mplayer window would dynamically shrink, to transparency would occur allowing me to use the scrollbar without having to move the mplayer window.
Everyone thinks that 3-D Window Managers are next. I say 3-d accelerated Window Managers, but having a box with windows on each side _really_ doesn't cut it in my book. It's neat. It's neat to program. It's neat to play with. Gotta get back to work now, good-bye. Just because 3-d is a big gaming thing and not used for regular Windows does not make it "The Next Big Thing(tm)" in my book.
What I would like to see, and this is off-topic, is XML menu specification. So you can download, install a program, and then install a menu item for it with whatever Window Manager you are using. It just needs a few fields. If someone wants to go with this idea and wants me to help(put my money where my mouth is) just e-mail me and I've got no problem.
What I also want to see is the death of X-Windows. It's served it's term, but it isn't getting any better. I want to see DirectFB succeed, but it needs to be multi-platform. I'm on FreeBSD so I can only run it under SDL ontop of X-Windows. But FreeBSD has something similar in the works set for probably 6.0 or whenever the person finishes it.
Communication and features between other type of hardware, specialized, would be great. And the framework to support it. Example, FingerWorks has some great products and great concepts. Once I get the money I'm going for their keyboard. I'd like to see a framework to make it work with any GTK, Gnome, KDE, GNUStep, and a generic library to add support for it to any program. That way have a custom gesture(when it is created) that will allow you to launch a program. I want to be able to hit numlock twice(Example) and type in 0805040206 and launch a program of my choice. For me, memorize 5 numbers, adding a '0' before it, and typing that in is much faster than moving the mouse, opening the menu, finding it, and clicking it. The generic framework, standardized would be best, would add the ability for, say, Mozilla to receive the two numlocks, to realize that it is a registered event handler, and to pass it off to the framework and do what is asked. Say, even passing it off to the 'server' so to speak to figure out what to do, although I think if it was implemented on a window manager level it would be best. That way you have a generic framework to work with as far as developers go, possibly a generic XML exporter of all your functions that you've specific(scanning the bar code, with your CueCat, of your favorite foot powder say, brings up userfriendly), and a generic XML importer to bring into the Window Manager. But having it Window Manager based, so that it fits in with Accessibility theory(I believe?). It _is_ a part of KDE Control Panel, it _is_ a part of Gnome Control Panel, it _is_ a part of that little WindowMaker configuration program. Easy for developers, easy for users, easy to switch between.
Sorry for the long post.
The fact is, any window environment must be similar to windows or users will get confused. New entrants must cater to the existing standard. Try building a new car with a different interface or maybe publish a book that reads up -> down. These items will fail. Look at the new BMW 7 series, all they did was add a dial that has extra functionality instead of a normal automatic shifter. Even though the traditional pedal acceleration and stop system remained. many buyers were completely put off by the idea.
Keeping Linux like windows is a good idea, getting rid of point and click makes no sense right now, but that doesn't mean in can't be done. With Linux people can write all types of crazy interfaces and environments, test them on a wide scale, and receive feedback. Apple and Microsoft can't afford to research 100 different window managers, but with Linux this is possible. The only problem with Linux is the developers, usually make decisions on the UI and look and feel. There needs to be a system in place where artists can make significant contributions to the DESIGN of open source software.
Dvorak brings up an interesting point, that interfaces are designed by coders and not artists, but oddly in the same article he says that linux shouldn't be going for a pretty look or features, merely to be different.
What is hands down most interesting about this is that for those of us who know his work, it seems to be a reversal of position. In the past Dvorak has ruthlessly bashed the macintosh operating system which stands for being different and had the original interface designed by artist.
There is some truth to the idea that an artist would make a better interface, but there are some guidelines which tend for better interfaces, and in general, a platform standard works well.
Apple provides the Aqua Human Interface Guidelines on their website for developers. This unification of interfaces on all application provices a unity over the system. In the MacOS a button in one application that is simmilar to another button should do basically the same thing. There are layout guidelines and notes for when to use different interface features, so a seasoned user will know what to expect when he or she does something.
The problem with impliemting something like this in the linux community is that there are many people working on any given thing, and too much varitation in X to do it well. Yes, it could be done but it isn't likely to happen.
Furthermore as far a putting features into the operating system, as someone above stated, that is what makes it easier to switch from Windows to Linux, and to that I say all the more power to us. Also Dvorak over looks the fact that any feature can be turned off, if the person dosen't feel like using it and wants more control over the system.
The point of Linux isn't to be difficult. It's to be open, free, and customizable. It is for those who don't want specific software shoved down their throats, and want to make their own software, edit someone else's or contribute to the greater good of their OS experiance.
Modular Redundancy--Because 4 out of 5 Nodes agree
Comment removed based on user account deletion
http://plan9.bell-labs.com/plan9dist/screenshot.ht ml
There are places where the networks are not touching,and there are places where they are-Boeing's Lori Gunter
Allright, that - in essence - what the article is all about. Yet, we know that the desktop metaphor is really the MINIMAL thing to implement before you can go on to other things. Because
1) users that were running Win/Max before don't want to change their way of working that profoundly
2) I for one think that the desktop metaphor will EVOLVE instead of just being killed and replaced completely
So, clearly with X/KDE/GNOME we are behind of MS/Apple by a more or less far shot. But I agree with the author, that - as some of us still are working on perfecting the desktop - we could work on possible "evolutions" and advancements.
One thing, for example, which will definitely be coming along in the not too far away future, is the "one-program" paradigm. The general idea behind is to
a) essentially have one "framework" interface for more or less all applications
b) really driving application-to-application interaction and data-transfer to a new level
c) employ new ways of browsing through data and software
d) making it possible to access the same data with multiple software modules while they are interacting with one another in a meaningful way
e) further degrade of the data-software boundary
So I guess we COULD put a lot of things together, if only OSS would focus more on the user side...
"The Innovator's Dilemma" by Claton Christensen suggests that Linux might win precisely by matching features. The main point of the book is that companies add features faster than customers demand them. Eventually, less expensive upstart technologies add features fast enough that the market switches en masse to the new technology, because the price/performace gets "good enough"
;-)
The quote "not quite as good but much cheaper" strongly suggests that Linux may be able to do this, even to Microsoft. Their weakness is that adding more features doesn't necessarilly help them against the Open Source community, who can add them as well. Unless a proprietary (patented) technology becomes indispensible to the user population as a whole, Microsoft seems vulnerable to attacks from low cost/no cost alternatives.
The interesting part about Christensen's book is his assertion that the disruptive technology needs to find a niche to take over, from which it can "attack upwards". The niche must be people who are not currently served by the market leader(s), who have lower requirements in key areas, and who resonate to the lower price point. Kind of sounds like the Open Source community:
1. Don't care if it's harder to use.
2. Admire the ability to tweak, instead of having everythig dumbed-down and hidden behind a GUI.
3. Did we mention it's free?
There's certainly a long way to go to displace Microsoft on the Desktop, but it's by no means obvious that Microsoft can defend itself by adding more features. Maybe Dvorak is right, but he hasn't addressed the disruptive technology issue at all.
What is it with people and their seemingly insatiable need to reinvent the wheel? The irony is that Microsoft's OS's caught on as well as they did not because Bill Gates is an Evil Genius *chuckle* but because Gates was dumb enough to write operating systems for the lousy x86 hardware paradigm when it began--the historical fact is nobody else wanted to. (The fact is that the guy IBM originally picked to do their OS decided to play golf instead of meet with IBM representatives as scheduled by appointment, and Gates was second on their list and he was in at the time.) Literally, no one else wanted the job.
Flash forward to the mid-late 80's. No one who was "anything" in the personal computer scene at the time would be caught dead using an x86 clone or DOS--they used Macs and Amigas which were brilliant concepts at the time, the Amiga especially literally being ten years ahead of Gates and Windows and x86.
Ironically, especially in light of the recent DOJ hearings, the reason the Amiga died and the Mac became a butt for jokes and received permanent niche status had absolutely nothing to do with Gates and Microsoft and IBM. The reason for those events was internal--for Apple it was a short-sighted and greedy Steve Jobs who did not want to license Mac clones; with Commodore it was a greedy and short-sighted Mehdi Ali who did not want to license Amiga clones (I recall at the time hearing from a source I trusted who informed me that Commodore had actually gotten a cloning agreement penned with Tandy and Radio Shack, where the company would have sold its machines in its thousands of retail stores under a clone name, but that Commodore pulled out at the last minute.) Both Apple and Commodore felt they could make more money by being the sole distributors of their hardware--neither company foresaw the incredible boom that would hit the personal computer industry in the 90's.
So it just so happened that Gates was the guy who grew up writing OS's for the one, single hardware standard which was open to tons of competition within--the IBM-PC clone hardware marketplace. In it you had dozens of companies all competing with each other to sell systems and peripherals--today there are hundreds of such companies all devoted to a single standard--the one that allowed clones--x86. Some people to this day do not understand that it was the hardware engine that drove x86 to vast supremacy--certainly not Gate's software--which back in the late 80's absolutely sucked compared to other OS's at the time. But because so many companies were selling x86 hardware so much cheaper than companies like Apple or Commodore, it was the x86 clones that were bought (most of the time Apple and Commodore could not meet demand for their hardware, which is exactly why they should've liscensed clones early on.)
And everywhere an IBM-PC clone went, a Microsoft OS was sure to follow. It's pretty simple to understand how Microsoft got to where it is today even though it was selling one of the worst OS's in existence for several years. Gates has never made a secret of it--there's the famous Gates-Jobs memos in which Jobs asks Gates what he needs to do to get the Mac into the mainstream and Gates writes back "License clones." It was advice which Jobs declined (which he now admits he should've taken.)
That's why I think Dvorak's bored...he wants something "new"...yet the only thing *he* can think of is some *old* crap nobody ever really pursued years ago *chuckle*...;) There's some inkling in his opinion that an OS should not be "functional" but "something else"--whatever the "else" is, Dvorak doesn't say....
It seems to me that Dvorak is forgetting that most if not all of the "new" ideas as to what an OS should be and do have all been tried and the GUI is the best that anybody's been able to come up with. Maybe when the hardware gets here we can have 3D holograms on the desktop that will work in fundamentally different ways, but for right now and the foreseeable future we're stuck with a 2D display (even our "3D" is just simulated in a 2D display.) And the GUI seems to be everybody's consensus of "what's best" for an operating system interface (of course some people still prefer the command line, but that's not what Dvorak is talking about.)
Dvorak talks about "wintel roots" without realising that "Wintel roots" had roots of their own which came out of earlier computing projects--and accusing one company of "copying" another simply because it chose to adopt something as fundamental as a GUI is pretty ridiculous. It's like saying GM and Ford "copy each other" because they make cars with four wheels and rubber tires. Is it really that they "copied" each other, or more like the fact that these things are as fundamental to the design of a car (or computer OS) as doors are to houses? Of course, that I agree with the latter should come as no surprise.
The trend in Linux today toward workable GUIs that happen to "look like" Windows was not intentional, nor was it subconscious as Dvorak contends. Rather, Linux advocates and developers have always worked toward creating a better OS than Windows and a different OS than Windows. But the fact is there are only so many ways you can skin the GUI cat--only so many ways to make a GUI which is intelligible. Dvorak's "look and feel" arguments are pretty funny--I thought we'd gotten past that bit of nonsense years ago. It's like saying Goodyear should sue Firestone (or vice-versa) because the tires the other company makes "look and feel" the same *chuckle* The whole "look and feel" argument was atrocious from the beginning and it's gratifying to see it never got anywhere.
Here's the thing Dvorak forgets: so what if Linux versions "look and feel" somewhat like Windows? Who cares? The fact is it *isn't* Windows regardless of what it looks and feels like. If anything such superficial similarities might actually help spread the acceptance of Linux (if the community can ever get over the factional splintering of distributions--which is the one thing that could doom its ultimate success as a competitor to Windows--but that's another story.)
I guess Dvorak forgot the simple admonition that contains worlds of truth: don't judge a book by its cover.
On the contrary. We're so stuck in this windowed world of operating systems. Why do we have to rely on the model of "windows" and "start" or "application" bars? The computing world is no longer flat or one-dimensional. Until someone comes up with something COMPLETELY different from the windows/taskbar/buttons model, I'll stick with XP. I'm waiting for the next jump in OS development (akin to the jump from 3.1 - Win95). Now THAT was revolutionary.
"player 4 hit player 1 with 0 stroms"
I dunno... I hardly consider the switch from 3.1 to 95 to be revolutionary. None of the basic paradigms changed. The major difference is that programs are now launched by a menu instead of a "folder" view. Otherwise, it's really more of an evolution. A big leap in terms of OS power; but in terms of UI not such a big leap.
I'm sure I missed a few things, but most of those things are not revolutionary. In fact, in most ways, Windows 95 was merely catching up to Mac and some other environments.
Now, the switch from command-line to GUI... THAT was revolutionary.
Yeah, like there's no right-click context-sensitive menu, which IMHO is one of the best UI features since GUIs have been around.
(Getting OT here, but...)
I don't mean to troll against macs here (although I will ;-) ), but MacOS X is (IMHO) the most counter-intuitive UI ever designed.
We've had an iMac (one of the flatscreen with DVD-writer) in the office for a while - it's the only CD-writer there is, and it's hilarious watching people use it...
"Hmmm, pretty"
"Now, how do I burn a CD?"
"Let's try putting the blank CD in first."
"Where's the eject button???"
(someone walks past and points out the eject key on the keyboard)
"OK, name the CD....'untitled' will do"
"Hey, the icon appeared on the desktop! Neat!"
"Now all I do is right-click on the CD icon and..."
"How do I right-click with this thing????"
(much later)
"OK, there's 500MB left on the CD so I can just add another session..."
"What do you mean 'read-only'? What happened to multi-session?"
"What do you mean, I have to be administrator to erase a CD-RW???? It's meant to be rewritable, isn't it???"
"SO, there's a mere 100K of data on this CD-RW but there's no logical way I can write ANY more data on it?"
(user walks away in digust and finds someone with winblows on their laptop)
Seen it happen many times... I'm not joking
He describes the whole free software world as a windoze deriviative born on x86 by "boring coders" and other uncreative types that lacks "features" of the only true software, Microsoft.
From the article: Long ago, Microsoft recognized that features sell software--not code size, efficiency, or even a pretty interface.
Tell that to the "designers" of XP: all ugly interface fluff.
I don't know about everyone else here, but the number of features availably on my Linux machine are a whole lot more comprehensive than on my Windoze one. At home (using Linux) I'm running: an enterprise-level web server (with support for Java, PHP, Perl, CGI, SSL, you name it), an internal DNS server, a caching DNS server, a highly-configurable router / firewall, an SSH daemon, a mail server (one which serves as both a primary for some domains and a secondary forwarding server for others), two different database servers, a print server (usable by Linux, UNIX, Windoze, OS X), a networked file share (available via NFS and Samba), a networked scanner server, a modem pool, a fax server, a VPN server, a jabber instant messaging server, an add-filtering HTTP proxy, an OGG/MP3 networked jukebox, a tape backup system, an LDAP user directory (with integrated logins for my Windoze/OS X boxes and support for redundant mirrors on other machines), an internal DHCP server, and encrypted file systems.
Cost to date: hardware + my time.
Software cost to date: $0.
All the software I needed (with the exception of the jabber server and the jukebox) came with my distro. I even had a few choices for some of the stuff (sendmail vs. postfix, ipchains vs. iptables vs. whatever else, ssh vs. frees/wan, junkbuster vs. squid, mysql vs. postgres, dhcpcd vs. pump).
Show me a windows machine that can do that all that (on the same machine!) with that cost, and then I'll concede the features point.
moto411.com
People using Lycoris and Lindows most likely cannot tell the difference.
Let those two OS's use the Windows style GUI, but lets innovate now because we already have things as Windows like as possible.
I think its time to innovate. I've given many ideas to the mailing lists, maybe they will use a few.
If you use Linux, please help development of Autopac
I've been trying to figure out what the next big thing is... probably along with a million other people I guess :) I imagine it will have to do with AI and LESS user interaction. Just like the Command-line hasn't gone away just because the GUI arrived, the GUI/CLI won't go away when voice-control and intelligent agents arrive. The other big thing will probably be hyper-embedded devices that hide the fact that you're using a computer. Imagine if your computer was a tiny device that could interact with "smart surfaces" to project a UI onto it. You could point it at a table-top and see your email on the surface of the table. But the basic email metaphor probably won't change... it's centuries old and a fairly efficient form of communication (excluding spam, of course :) )
Not to be a jerk, but I doubt it looks half as good as Mac OS X. The Linux world has forgotten about the second half of the phrase 'look and feel.' IMHO, that's far more important than the color of the widgets. If it feels like a rickety bobsled, putting a big Mercedes Benz logo on the hood isn't going to make it feel any more reliable.
Don't get me wrong, I love Linux for serving (and I have two Linux servers) but almost five years after I started trying to convince myself that Linux could be a reasonable desktop, I'm further away from believing it than when I started.
I beg to differ. Mac OSX looks quite different from Windows; however, Apple's "switch" campaign is relatively significant and realistic impacts. That is, even though OSX does not look like Windows, Win users *are* moving over to the OSX platform.
So there.
Michael
So is "if I move this funny object on the table, the arrow on the screen moves as well".
Or "if I press down on this button, the letter on the button appears on the screen, next to where the blinking line is".
That's exactly the force Dvorak was arguing against :)
Sure, but: Baby. Bathwater. Don't throw out with. Change for the sake of change is a tricky thing, and often leads to disaster (or as linus would say, "that way lies crapness").
One thing Linux users don't understand is that Linux developers write applications to solve THEIR OWN problems, not to solve the problems of users of those applications. If we were paid BY users, to write software FOR users, the features and functionality might actually contain user requests and features, but WE ARE ALL VOLUNTEERS.
If you don't like it, submit a patch, pay us to help add the features you want, fork the code, hire someone else to add those features, or return it for a refund.
I think you made my point in reverse. If a wizard can simplify only simple tasks, why have a wizard? Why not just let the user do it?
I think, though, that wizards are a symptom of the problem. If it's that bloody hard to do something, then putting a big button on top to mash all of the little buttons at once is a problem.
Why bother switching? An innovative/paradigm shifting approach a la the original Mac over DOS would be a better strategy than always playing catch-up to Microsoft. I've been using Redhat since 1997 (I came over from Mac) and can totally predict the next release's GUI knickknacks by looking at the current version of Windows. These features tend to be buggier and more resource intensive than comparable features in Windows or OS X. I love Linux. It is an incredible server OS. I would never consider running my server applications on anything else. Yet it is a different story in the desktop arena: Back in 1997 Linux was the clear choice for me as far as having a robust OS in my desktop to do research and manage network applications, but now I am not so sure. At my current rate of frustration, I'll be using OS X in my desktop by the middle of next year.
I am a virus, put me in your
You realize bash works in Windows, right? So what's your point, exactly?
I wish I had some mod points to mod you up, but of course folks avoid modding in threads where they wish to reply, so I used them before I read this thread. Damn, because you make some very valid points.
I agree that KDE and Gnome are not pushing the envelope in interface design. I think most of the programmers working today have grown up in an world so saturated with Windows that they honestly haven't been able to imagine better ways of doing things.
Dvorak may be a whore, but like a very old whore he's seen everything. He started writing about PCs in the Altair era, and has witnessed nearly the entire evolution of the personal computer. He's written about Amigas, PenPoint, Deskview, NeXT, BeOS...whatever. And he gets demos of things before they come to market, including things that never came to market.
So, before people dismiss him as a buffoon, take a step back and consider what he is saying.
"How perfectly Goddamn delightful it all is, to be sure" Charles Crumb
Um. I think you are missing the point. Yes, of course KDE/GNOME are not linux. But the article was about the perception of linux by a first time user, not by Mr/Ms SuperGeek.
Personally I couldn't give a damn about whether linux achieves world domination or not. However, if you do wish to convert more people to using linux then we're talking about exactly those people who are going to judge books (or in this case OSes) by their covers. We're talking about people who will only use linux because it is better than Windows, and they're not going to spend months finding obscure, hidden software - it's going to be based on a first experiences thing. And having seen that KDE/GNOME is crud, they're going to avoid linux like the plague for the next couple of years, based on that one experience. I've seen it happen.
As I've said before, I couldn't care less - linux does what I want it to, and that's all I care about. But since a substantial population of slashdot seems to think that linux should achieve world domination, or at the very least give Windows a run for its money, then this article should be very, very relevant.
And shoving your head in the sand and refusing to listen to criticism, even when it is valid and constructive, is plain out stupid. (But hell, this is slashdot, isn't it - when did constructive criticism of linux ever get noticed? :)
The Windows NT Command Shell.
Full plate and packing steel! -Minsc
Dvorak knows something about computers, specifically desktops. Apparently he knows nothing about the rest of it, or he would have discussed the thousands of creative uses of Linux, in server clusters, network appliances, embedded devices of all shapes and sizes, incredible server clusters, renderfarms, the list goes on and on. He also ignores the numerous interface projects, both 2D WIMP enhancements, and the 3D interfaces that Windows does not have.
Dvorak is a fool, a pundit, he is the computer industry's Rush Limbaugh. Fortunately for the computer industry, Dvorak does not have millions of moron listeners who fail to look through his fallacies.