Which OS is more User Friendly and Intuitive?

The widely accepted majority opinion about operating systems is that for some unspeakable reason Windows is the pinnacle and a shining beacon of usability. Every single time someone brings up making Linux available for the masses (for example when discussing the Wallmart Linux laptops, or Dell’s Ubuntu line) some munchkin .NET developer and self proclaimed usability expert invariable pops up and starts counting the ways in which the discussed OS is different than windows. Not that I have anything against, .NET developers themselves – most are good people but you can hardly expect people who are professionally tied to windows platform to have an objective and well researched position on Linux. And trust me, it’s always some .NET developer, or someone who works exclusively on Windows and has only a second hand Linux experience. By second hand I of course mean that he reads Linux related threads on Digg and Slashdot but never actually used it himself.

Most of the time this attack is not done on objective analysis, usability experiments or statistical data but on “gut hunches” people have about what is accessible to the general public. And the bad part is that most of us share these hunches, and will at least in part agree with them. One such point that is usually touted around is that GUI is better than CLI. I can throw that one in the open, and most people will nod in agreement and say “yeah, for a novice user a GUI is way better”. How do you know though?

I guess part of the problem is our origins. A lot of Linux people my age and younger came from certain background – they are ex-Windows users who learned to do their first steps on the Microsoft platform. Perhaps they vaguely remember DOS, but most of their formative years they were kicking it MSWIN style. A lot of them probably remember feeling uneasy about CLI when they first started experimenting with Linux.

I can’t say this is my experience. The very first computer I used was a Commodore 64. The first computer I owned was Amiga 600 which had a full blown GUI OS, but also a functional shell that I used to hack various scripts and display ASCII art on bootup. The first “Computer Class” I took at school was done exclusively MS-DOS based. We didn’t even have Windows installed on these systems – and hell, no one was using it anyway. We considered Norton Commander to be a luxury. And you know what? We were getting along just fine. No one really complained it was to hard, or unreasonable – this is what we had to work with at the time, and we were fine with it.

I never feared getting my hands dirty writing scripts or dropping down to DOS when I got my first Windows machine. But even I for a while got trapped in that “GUI is Easier for Normal People” mentality. I forgot where I came from, and followed the wisdom widely accepted by the masses. Then few years ago I read a wonderful, heart warming story written by a guy who taught an introductory Unix/Linux class to adult learners. I wish I could link to it but I don’t have it anymore. At the time I might have bookmarked it, but that was 2 computers and several catastrophic hard drive failures ago. I have been searching for it ever since, but it vanished somewhere in the depths of iterwebs. Thanks to the magic and infinite wisdom of the internet, I have recovered the link! Thanks for finding it Newt!

I think the author made a really, really interesting point that at the time shook me out of the “Newbs must have GUI” trans. Many of his students admitted that they never really used a computer on their own. Others have been trying to “learn computers” many times and they always failed. This class was different than what they tried before – instead of working with the latest Windows platform, he taught them the very basics of command line unix using simple tools like bash, pico and pine.

And the astonishing part was that it worked. Few weeks into the course these people who had no technology background, and were throughly confused by Windows were already proficient bash users. Many were researching new commands by using appropos and reading man pages, and exchanged simple .bashrc tricks. Some of them actually remarked that this is the first computer class that actually made sense to them, and that the command line was much easier to understand than the multiple windows, panels and tabs of a GUI. It was like “talking to the computer” they said – you issue a command, and then it either does something or responds that it did not understood.

The common wisdom that CLI is hard and GUI is easy is obviously a fallacy, or at best a half truth. If you can’t use a computer at all, then a single task oriented, modal, command-response model of a CLI interface might actually be more intuitive to you than the multi-tasking, multi-window confusing interface. Let’s face it – staring at a dialog with 6 tabs, each of which has 2 or 3 buttons that open another dialog with 5-6 tabs can be intimidating for a novice user. This is one of the areas where GUI designers really can’t beat the plain old well documented configuration text file. They either emulate it by creating a web page like config screen where each option is accompanied by a paragraph of text, or do some bizarre, counter intuitive nested abomination with no explanations like the one I mentioned above.

I read that story, and it struck me – all these pretentious, self proclaimed usability experts don’t know shit. If you sit down a former Windows user at a Linux machine, then naturally they will be confused. But if you decide to one day teach your grandmother about computers do you really think it will matter to her what OS she will be using? I can guarantee you that she will be equally confused by both. There is no OS out on the market that requires zero training or getting used to. Unless we develop a full fledged, voice activated AI that will converse with the user (like computers usually do in Scifi movies) there will always be a steep learning curve, and many idiosyncrasies to be dealt with.

Is it really that much easier to edit Windows registry than to edit textual config files? I bet your grandma will do neither. Is it really that much easier installing software on windows than it is on Linux? I mean, hell – if you are using Ubuntu you pretty much have a one-click install mechanic. Choose software from the synaptic list, click install and you are done. But guess what – your grandma probably won’t be installing software anyway.

The OS that is user friendly and intuitive is the one you have learned to use proficiently. If you have learned to use Windows first, then you compare everything to windows. Whenever things are done differently you usually decide it’s hard, or counter intuitive. But really it is not. We are all conditioned by our previous experiences, but someone with no prior bias may experience things very differently.

So next time someone starts yelling that Linux is not ready for the desktop, just consider whether or not his arguments are objective, or is he simply following of these gut feelings we all tend to have about usability without actually testing it in practice.

[tags]linux, linux ready for the desktop, gui design, cli, command line[/tags]

This entry was posted in sysadmin notes and tagged . Bookmark the permalink.

39 Responses to Which OS is more User Friendly and Intuitive?

  1. Matt` UNITED KINGDOM Mozilla Firefox Windows Terminalist says:

    I have irrational prejudices against the command line… these would most likely go away if I knew how to use it better though.

    Stupid lifetime of using Windows… it’s like a scar in my brain that runs straight through the “brain-scars are bad” centre, thus making me not care about the scar.

    That said, if I took the plunge and just switched to linux (I keep intending to do so at least partially but never get round to it..) I like to think I’d figure things out. Maybe when the support from the wonderful world of hardware manufacturers and software writers gets close to Windows-like….

    That’s the thing I like most about Windows, even when I can see it sucks, it’s got everyone backing it…

    Reply  |  Quote
  2. Kiyu UNITED STATES Safari Mac OS says:

    I understand that you are primarily talking here about Linux on the desktop. Having run Linux as my sole desktop for over a year and being thoroughly versed in every version of windows there is, I feel very confident in my evaluation when I say that Mac OS is superior to both Linux and Windows desktops from every meaningful metric. As far as CLI, it is basically the same as Linux (a few differences, but they are trivial), and has all of the rich scripting abilities of that environment (and then some). With some compatibility libraries, nearly every Linux App can run under OS X – PLUS all of the programs which are written specifically for the Mac.

    The GUI is uniquely scriptable with AppleScript and all of the programs which support this kind of scripting by publishing an API for this purpose. In OS 10.4 Apple introduced Automator, a GUI for easily accessing scripting capabilities.

    From a programmer’s perspective, the tools and languages available for natively programming OS X (Objective-C) and its development tools are superior to those available on a Mac – but that is subjective… They also provide clean access to some VERY powerful OS level tools: CoreData for native data structures, CoreAnimation for making very cool visual effects, as well as “Core” libraries to graphics and audio. And if you don’t like Objective-C, you can code to one of the best integrated Java VMs available. Still don’t like it? How about scripting the interface with Ruby or one of the other languages that can interface to their interface builder tool.

    From the user’s standpoint the OS is simply more polished and easier to use than any version of Windows. Some of the best features of Vista are clearly OS X rip-offs – but they took a minimum 1GB ram to make it work. What is an OS DOING that needs 1GB RAM???

    Last, but not least, for those applications which one absolutely MUST have that ONLY run under windows, run Parallels in Coherence mode and everything just blends together VERY gracefully. I can even drag and drop files from my Mac desktop into a Windows app or explorer window and it works.

    Of course, I’m not alone on this evaluation. Your post points out the uninformed opinions of Windows developers regarding Linux. I would encourage you to look into the opinions of any respectable techie or programmer who has given OS X half a shake since OS 10.4 (Tiger). I have found (and many agree) that OS X is indisputably the most powerful, advanced and easy to use OS for both power-users AND novices.

    Reply  |  Quote
  3. Ian Clifton UNITED STATES Mozilla Firefox Mac OS says:

    I think the primary advantage of a GUI is the way that interaction can be done. For instance, it’s a lot easier to “play with a slider” to adjust gamma than to narrow it down with xgamma on the command line. Of course, in general, I hate the mouse. It’s slow and inefficient and damned annoying at times. I’ve maxed out the speed settings of the one at work on OSX and it is still frustrating. When I want to select a specific point, I slow down to do so, but then it slows down the mouse too so I’m not even inching toward the point I want; I’m pixeling toward it! Then I switch to Vim, hit a couple of keys, and everything is done.

    As far as overall which OS is more intuitive/user friendly… it’s honestly hard to say. I think gnome is intuitive, for the most part. The problem comes when you can’t do something you want to (for instance, setting a resolution to 1680×1050). The same is true of windows though. How many people know how to force a resolution that does not show up?

    OSX is somewhat friendly, but without QuickSilver, opening an app that isn’t on the dock is just annoying. Plus, try changing the desktop background to a regular flat black… you can’t. You have to select from the colors Apple has chosen rather than just dragging around in a color wheel or something similar. Or you can create an image file of a black pixel and use that….

    Usability is mostly not very good for a lot of computer programs (and operating systems in certain cases). The funny thing is, people expect computers to be so intuitive that they don’t have to put in any effort to learning how to use them effectively. Funny that the same people were so willing to put in time to learn how to drive a car, which is far from easy or intuitive.

    Reply  |  Quote
  4. Fred UNITED STATES Internet Explorer Windows says:

    I too first started computing on a Commodore 64 and the ever infamous TRS-80. I learned how to operate the computer by command-line programming. When Windows 3.1 came out, I couldn’t believe I could do by clicking what I did by typing. It wasn’t until they took DOS away that I had problems with Windows. I could tweak it and make it work the way I wanted it because I knew HOW/WHY IT WORKED. I feel the same way about linux now. I just installed my first Linux distro the other day and I have to say that it works extremely well. It took some time to understand some of the new screens, but it is most definitely ready for the general public. BRING ON THE LINUX.

    Reply  |  Quote
  5. Mackenzie UNITED KINGDOM Safari Mac OS says:

    Let’s face it, for sheer idiot usability, OSX is the best. It’s intuitive, it’s designed to be self explanatory, and it just WORKS. Everytime I use windows, I run into a pile of incomprehensible and unexplainable errors for the simplest of tasks, or have too bitch my way through a wizard. Let’s have an example:

    I wanted too print some pics. Ok, fair enough. Windows has a neat little idiot feature that shows various default configurations you can use for your photos, which is pretty cool and useful.

    Hit print, selected printer which was all plugged in and turn on and installed..

    Nothing happens.

    Do it again, nothing happens.

    Computer tells me that it’s finished printing. The printer is idling and bored.

    Check control panel, nothing out of the ordinary there. the printer is there, configured and installed, with plenty of ink and such.

    Turns out the printer only works on the User who installed in- Even though it appears to be accessible and workable for everyone. So when my mom logged in, suddenly it starts spewing paper everywhere. Like I said, incomprehensible and absurd.

    As for linux: Fuck that. I do NOT want too learn ANY code WHATSOEVER. Can’t abide by it. The idea of writing code too do a task which could be done on a GUI with a few clicks is just, Backward.

    I tried installing Ubuntu on a Windows x86 machine, didn’t work. Kept installing forever.

    Tried installing OpenSuse on the same machine, also didn’t work. Had too repartition everything rah rah rah rah rah after learning some command line.

    Will Joe Bloggs want too spend hours trying too wipe a HD too install linux because it gives him an obscure error during installation? I think not. I had too sit down and figure out what the problem actaully was because I don’t know tech-speak.

    4 hours later, Opensuse installs itself, then crashes.

    Repeat installation process x5 for Ubuntu, OpenSuse distros

    The only one that worked was DSL. Which is, ugly. Not exactly a home PC setup there. And incomprehensible. Had some obscure program for moving files around, and copy/paste and drag and drop didn’t exist.

    I have Never, ever, ever had a problem or glitch with my OSX that wasn’t solved by turning the modem Off and On again, or that wasn’t caused by third party underdeveloped freeware

    Reply  |  Quote
  6. James Heaver UNITED KINGDOM Mozilla Firefox Windows says:

    You might be interested in a (now defunct) project spearheaded by the (now defunct) Jef Raskin (of Macintosh fame) – Archy http://rchi.raskincenter.org/index.php?title=Home

    Although I don’t agree with all of Archy’s aims and premises I do think that it is fundamentally on the right track. Programs such as launchy and quicksilver show that the CLI is not defunct, it just needs to be updated and integrated with the GUI. Launchy is the incredibly powerful tool, especially when integrated with a service such as YubNub.

    There is an awful lot further to go with this new generation of CLI, the possibilities are only just beginning to be tapped. – beit the firefox address bar (especially in firefox 3), google, google’s appointment parser (dentist at 3pm wednesday), twitter, yubnub, launchy or even kuake – there is allot of innovative work going on, all we need now is to stop, think, and readdress the WIMP paradigm.

    (humanized.com are continuing elements of Jef Raskin’s work, and are comign up with some innovative ideas, but yet to have a really coherant program)

    Reply  |  Quote
  7. Luke Maciak UNITED STATES Mozilla Firefox Ubuntu Linux Terminalist says:

    @Kiyu – wow, that is a very good endorsement for OSX. I can’t really speak from experience because I never used it but I do believe that they have put a lot of research and resources into usability. You can’t really go wrong with a unix backbone, wrapped around in a top-notch GUI, no? I have already decided some time ago that my next laptop will be a MacBook. I just keep procrastinating buying one while my old lappy is still ticking. :P

    Please note that, these uninformed opinions I mentioned do not only apply to Linux but also to Mac. Yes, there are people out there who are hating on Apple and applying the same “not like windows” yardstick to measure Apple’s usability.

    @Mackenzie – Oh wow! Wow… Sorry to hear about your poor experience with Ubuntu. To tell you the truth, I have never experienced anything even remotely similar to that. In fact my experience was the polar opposite:

    Install windows:

    1. Wait an hour while the thing installs itself
    2. reboot
    3. boot into bare-bones windows
    4. following devices are not working:
    – Ethernet card
    – Sound card
    – Wifi Card
    – Video Card (I’m running 640×486 and 16 bit color)
    – modem
    – sd card reader
    5. now I need to somehow connect this machine to the internet and download drivers.. But how? Neither card is working?
    6. Use another computer to download Ethernet driver and install it
    7. reboot
    8. Download video card driver
    9. reboot
    10. install sound card driver
    11. reboot
    12 install wifi card driver
    13. reboot
    14. install modem driver
    15. reboot
    16. install Norton/Mcafee/Kaspersky
    17. oh… Wait, my machine got infected – I should have installed AV first
    18. format C:
    19. goto 1

    You get the pattern, right? Same machine, install Ubuntu:

    1. wait 20-30 minutes as it installs
    2. log into machine and everything sans wifi and modem works
    3. install ndiswrapper
    4. download and extract windows driver
    5. setup ndiswrapper
    6. modem is winmodem – can’t do anything about that
    7. ???
    8. profit

    I seriously never had “major” issues with Ubuntu installations. But these things happen.

    OSX does seem to have the “everything just works” thing covered. :mrgreen:

    @All – Interesting discussion. I think I have some more comments to what some other people said, but I need to get out of work and head home, so I will respond later. :)

    Reply  |  Quote
  8. ratchet UNITED STATES Mozilla Firefox Mac OS says:

    I agree with the author. It is as easy as you want it to be. I too learned MSWIN from DOS without any “Windows” installed. I remember making batch files to create a sudo-menu system. It was all fascinating at the time.

    A GUI isn’t easier, it’s just different. You still have to understand what the purpose for that GUI is… if not then it is infinitely more complicated than something you do understand the purpose behind.

    BTW: I’m a .Net developer, but I use Linux as my primary desktop (and my favorite desktop). I do .Net for the money. And for the record… all the MS talk about .Net being easier to upgrade and manage is all BULL. I can manage a PHP app about 100x easier than ANY .Net app. Have you tried to upgrade from .Net 1.1 to .Net 2.0? It works, but the question is “Is it easier than upgrading from PHP 4.x to 5.x? If you write you code correctly then your 4.x app will work in 5.x without a single code change (most of the time).

    Reply  |  Quote
  9. Josh UNITED STATES Mozilla Firefox Windows says:

    An end user confused by a couple of windows on the screen?! I hope those people can multi-task to keep hands on the steering wheel, a foot on the brake and eyes on the road, all at the same time when the traffic light turns red. Hopefully it does not start raining at the moment, because turning on the wipers would be waaaaaay toooo much.

    When do we stop using a half brain dead, 3rd removed cousin on grand-mother’s side as measuring yard stick? If you teach a monkey to click a couple of buttons in GUI, then GUI is 100% better then CLI? There are cases when GUI is better, then there are cases when CLI is better tool for the task. Just use the right tool and end of discussion.

    When your so called “user friendly OS” does not perform certain functionality at all? Is it still friendly?

    Reply  |  Quote
  10. Jess UNITED STATES Mozilla Firefox Linux says:

    I decided to put in my two cents even though I’m a new user to linux. I’ve been using different distros for the last year or so.I learned a lot. It’s been mostly fun (not all). I’ve jumped to different distros trying to find one to use that done everything I do. I want something pretty PNP. I must say it’s getting lots better and more fun and exciting. At 63 yrs old, it takes quite a bit to get me excited. I really like being free from Windows and I am very glad I took the steps I did.I have two systems that I built myself. Fairly up to date not cutting edge. As a matter of fact right now I am using one system and Taking windows off my “main” system and will end up with three distros on it able to multi boot to which ever one I choose on start up. I’ve had it with windows and the blue screens and the errors. I’ve only had that (error) a couple of times with linux and I probably induced it myself. I won’t keep rambling. I think the quickest and easiest for beginners would be PCLinuxOS. I know they say Ubuntu, but PCLinuxOS is better to start with.

    Reply  |  Quote
  11. Luke Maciak UNITED STATES Mozilla Firefox Windows Terminalist says:

    @ratchet – as I said, nothing wrong with using .NET in itself. I have nothing against the platform, an I did use C# on several projects I did in the past. I’m just poking fun at the few career .NET people that I who consider themselves to be experts on everything (especially linux) despite the fact that they have to ask me, a filthy linux enthusiast fix their Windows XP box after infecting it with 50,000 trojans . :mrgreen:

    I don’t use .NET to know the pain of 1.1 to 2.0 upgrade but I can imagine the fun it entails just from the way you describe it. ;)

    @Josh – you missed my point. I used the example of a “brain dead” (as you say) person to illustrate the point that CLI is not inherently more difficult.

    My point is that while most people think that novice users need big shiny buttons, it is not always the case. You can teach them CLI just as easily.

    I used the technologically-inept people as an example, because as opposed to us they do not have built in preference. They are a blank slate onto which we can imprint knowledge. They were not taught that GUI is easier, and so they are not afraid of the scary CLI but embrace it.

    Reply  |  Quote
  12. Nick UNITED STATES Mozilla Firefox Debian GNU/Linux says:

    @jess PCLinuxOS is definitely a more user friendly OS that Ubuntu based on the fact that is a lot more stable and won’t break as easily.

    But in my opinion Sabayon is probably one of the best OS’s to start a linux User on, as it comes with 10+ gigs of software and every time there is a new release (not sure exactly how often that is), they just need to run the update wizard which will update their entire OS and all of their software from binary packages. They don’t need to manage it in any other way as it already comes with compiz, media codecs, and proprietary drivers (if wanted) installed.

    Reply  |  Quote
  13. vacri AUSTRALIA Mozilla Firefox Windows says:

    Is it really that much easier to edit Windows registry than to edit textual config files?

    Yes, because at least the registry is all in one place. No need to play ‘guess where the .conf file is’. <3

    I’m slowly moving to linux from windows because I loathe the licensing, but it’s been a bumpy ride so far (especially when X dies). Moving to OSX isn’t on the table for me at the moment because of the licensing and the hardware lockdown.

    The biggest problem I find with the CLI interface is that, while more powerful than the gui, it’s difficult to ‘explore’. It gets even worse when the man page or –help was written by someone who can’t write help files.

    Reply  |  Quote
  14. Luke Maciak UNITED STATES Mozilla Firefox Windows Terminalist says:

    @vacri – I actually find Linux config a bit more intuitive. There is actually just a few places where config files live.

    Your personal config files live in your home directory usually in the form of .appname or .appnamerc.

    The systemwide config files usually live in /etc/ or /etc/appname/. There might be few apps that put them in different places (like X server for example) but it’s rare and usually well documented.

    Windows may have it all in one place, but it doesn’t make things easy to find. Sure it’s easy to get into it, but once you run regedit you are faced with a huge tree with few dozen branches each of which conceals a hierarchy 6 or 7 levels deep. It is really hard to find anything in registry unless you know what you are looking for. The paths to correct registry keys are just about as convoluted as paths to config files.

    I agree with your second point though – GUI’s are easy to explore. This is especially true for exploring the file system. However, looking GUI panels is not always helpful. I will give you an example:


    Both are GUI applications, and you can explore at will but it won’t help you if you are new to windows and don’t know much about registry, or the way services are launched on startup.

    In fact we are slowly moving away from this model and more into the “search” paradigm – instead of browsing through the Start menu we now use Launchy. Instead of browsing the filesystem we search it.

    Reply  |  Quote
  15. Muhammad SINGAPORE Mozilla Firefox Windows Terminalist says:

    My personal take on OS usability is that it is all about user preference. No OS is truly superior over the other.

    I use Vista at work. My gaming rig at home is on XP. I took my fiancee’s old P4 rig (when she got a new PC) and turned it into a game server primarily running a MySQL database with Ubuntu Gutsy. The only OS I haven’t experienced at all is Mac OSX, mainly because it is not a publicly available OS.

    To me, all OSes have the same degree of complexity and usability. In all 3, XP seems the easiest as it is the OS I am most familiar with. But having used Vista for almost 6 months, and used Ubuntu Gutsy for almost 2 months, I can say for certain that if one took the time to learn and use an OS (and google a lot), everything becomes easy and intuitive.

    I really don’t understand the bashing some can say about other OSes, just because they run them once, don’t know how to do certain stuff, and instantly assume that that OS is poor. I have “gaming” friends who swear that XP is the best OS ever made, “linux-guru” friends who hate all Microsoft products, “haven’t-touched-Vista-before” friends who hate Vista, and “slight-air-of-superiority” friends who run Mac exclusively. And on this note, some of my colleagues also said that “MS Office 2007 sucks!” just because they are comfortable with the old Office 2003 toolbars and didn’t know how to use the new ribbon interface.

    I think it’s all just a user problem. Whether you can accomodate and learn to use a new OS, it’s all up to you. My advice to all of my friends is simple: if you’re comfortable with using your computer and getting things to work, just stick with it.

    Quoting from X-Men, people just fear that they don’t understand. :) :P

    Reply  |  Quote
  16. Muhammad SINGAPORE Mozilla Firefox Windows Terminalist says:

    [quote comment=”7737″]Quoting from X-Men, people just fear what they don’t understand.[/quote]
    “that” should have been “what”. Sorry about this! We need an edit button here! :P

    Reply  |  Quote
  17. Newt UNITED STATES Internet Explorer Windows says:



    “Then few years ago I read a wonderful, heart warming story written by a guy who taught an introductory Unix/Linux class to adult learners. I wish I could link to it but I don’t have it anymore.”

    rang a bell.

    Could this be it?:

    “http://www.osnews.com/story/6282/The-Command-Line–The-Best-Newbie-Int erface//advertise.php”

    Cheers. N.

    Reply  |  Quote
  18. Kiyu UNITED STATES Safari Mac OS says:

    @Vacri The “hardware lockdown” to which you refer is one of the strengths of using Apple computers. How hard would it be to write a piece of software that is supposed to interface gracefully (read: stable and predictable) with thousands of pieces of hardware tied directly into the depths of your program? That is the Operating System developer’s task.

    Both Windows and OS X have to interface with thousands of peripherals but Apple has to interface with a dramatically more limited number of motherboards, bus controllers, PCI controllers, video cards, etc. – that is, all the really core stuff. The people who work on those drivers for OS X can focus all their attention on many fewer variables and instead put their efforts into making their drivers REALLY stable, reliable and FAST.

    So here I’m going to do a little Windows bashing AND a little Linux bashing: the people who develop drivers for hardware on these platforms or the interfaces to which those drivers connect are frequently cursed on one or more of several fronts: bad interfaces (OS side – mainly Windows), poor hardware documentation or not-so-cooperative hardware manufacturers (open-source driver developers), “little-brother syndrome” (every hardware manufacturer who puts any resources into Linux driver development – Windows is their primary market so most time and talent is focused on Windows).

    Linux is a superior server platform: ALL of my Internet servers are Linux. It is THE MOST FUN and MOST EDUCATIONAL operating system to play with as a tinkerer – anyone can get as deep into it as they feel comfortable and learn, change or add; Linux is (IMHO) THE BEST OS for embedded or highly customized systems (routers, TiVo & other DVRs, smart appliances, etc.). I’m never going to say that Linux as a Desktop should not exist – that would be like saying people should only play with the command line. The GUI offerings in Linux are varied and innovative. GNOME and KDE both have “traditional” (read: Windows & OS X) feels, but when I saw Enlightenment and Blackbox I thought “hey, there ARE different ways to do this stuff… I don’t like it, but it is good to know some people are thinking in different directions and I’M glad that I’ve been exposed to those thoughts.”

    HOWEVER, if you are interested in getting something DONE and you want as little impedance between you and that goal, OS X is the desktop environment for you. I feel that a common path for many power users is this: mainstream (Windows), non-mainstream on the hardware you have available (Linux), and then power-productivity (OS X).

    If you don’t believe me, think of the MOST innovative power-user you can think of, people who really shape the technology world or people who just PRODUCE – and ask them (or research them) how they feel about Apple and OS X. Check out the brainiacs at Google or Yahoo, what are they running? Why? You will see that among the “computer elite” (not that I am one of them, just that I’ve observed this) there is an inordinately high preference for Apple. The reasons are many but the “It just works ‘phenomenon'” is a big one, but the accessible power-user features available in OS X far exceed those available even in Linux. (I say “accessible” because if you have forever to tailor your OS to your preferences, yes, Linux could certainly have more) – even without QuickSilver.

    For those that read Ian’s post and don’t know, QuickSilver is a keyboard shortcut third-party app for OS X; Ian, I think that your statement is rather elitist-sounding and inaccurate. The Dock has always supported folders which work similarly to the Windows Start menu or the menu systems available in GNOME or KDE; they are better in fact, because you can have several of them, rather than just one. The new Stacks feature just makes these even prettier; but if you must have a keyboard shortcut, Command-Shift-A brings up the applications folder where most, if not all of your applications reside.

    I also take issue with your comment that pokes fun at people’s expectations of usability and unwillingness to invest themselves into their computer; first, MOST PEOPLE ARE NOT LIKE YOU AND ME; they are forced to use this thing they fundamentally do not understand to do their job to pay their bills. They just want to do their job. Second, poor usability is a programmer’s failure. The attitude that users should just figure it out or that the user is the problem is what I see reflected in the interfaces and workflows of plenty of software on Windows and surprisingly little of on Mac OS. I think there’s just a philosophical difference there. The bottom line is that the POINT of using a computer is to make a job EASIER, FASTER, MORE EFFICIENT, etc., therefore it is the JOB of a programmer to make software that makes the computer do just that – and “frustration” and “easier” don’t coexist.

    On the flip side, I was surprised that I could not, in fact, simply set the background to black. Good point. Of course, that would make all of the window shadows that make it more obvious which window is in front useless. So the designers at Apple made a decision on my behalf that prevented me from readily decreasing usability. THAT is an example of being a good programmer.


    Reply  |  Quote
  19. Ian Clifton UNITED STATES Mozilla Firefox Ubuntu Linux says:


    …opening an app that isn’t on the dock is just annoying.

    I’m not sure how that is either “elitist-sounding” or “inaccurate,” particularly since it is an opinion and I didn’t express superiority (as far as I know). My guess is that you are a heavy OSX user and don’t like people pointing out imperfections in that OS.

    The Dock has always supported folders which work similarly to the Windows Start menu or the menu systems available in GNOME or KDE; they are better in fact, because you can have several of them, rather than just one.

    To me, that is far more elitist than my statement not to mention inaccurate. You can add more than one menu in Gnome and you can fully customize it. Either way, my point was about a blank system, not one that is customized to the user.

    …if you must have a keyboard shortcut, Command-Shift-A brings up the applications folder where most, if not all of your applications reside.

    And Alt+F2 brings up the run application dialog box in Gnome, but those aren’t intuitive controls, which is what this whole post was about. It should never be difficult to get to a simple program (like a text editor).

    Personally, I use all major operating systems (though limited experience on the latest versions from Apple and Microsoft), and often the systems are in labs that do not have my settings. If the dock has the app, a user can just click it. If it isn’t there, the user has to go into the apps folder (assuming the shortcut exists) and find the particular app among dozens (or hundreds on some systems) of others. Sure, you can use spotlight, but I don’t think that is intuitive (and can be slow when you have to go to a “fresh” system each time). On some of the systems, there isn’t a direct link to the app folder, so users have to go to “finder” and then hopefully notice the applications folder and go from there.

    Windows is not better, because the idea of categorization failed at some point (notice the accessories folder and sometimes a gaming folder and yet 99% of shortcuts don’t go in there). That’s speaking from a 3.1 to XP experience (have barely touched Vista). Gnome uses categories. Applications -> Internet -> Firefox; Applications -> Graphics -> GIMP. To me, categorization is much more user friendly than the typical “all in one folder” approach. This isn’t to say that you can’t set up any of these systems differently, as you obviously can, but a system has to be reasonably friendly and intuitive before you are able to customize it.

    User friendly is about being easy to access and understand from the layperson’s perspective. Obviously, each person sets up a system how s/he wants (maybe using Launchy, QuickSilver, Gnome-Do, Katapult, etc.) after understanding the basics. You say that being unable to easily customize a desktop BG to black is an example of a good programmer and I totally disagree. Pick another color arbitrarily and try it. Suddenly the “black wouldn’t make the shadows as useful” excuse dies. Give the user a color wheel! This isn’t the only example is OSX. Try taking a screenshot! Let the readers here know what they have to do to get a screenshot saved in OSX and I’ll do the same for Gnome.

    Press “Prt Scr” and click save. That’s it! That’s how it should be. Don’t get me wrong; there are plenty of things that are difficult in Gnome that are easier in OSX, but I don’t think OSX is the pinnacle of usability.

    the accessible power-user features available in OS X far exceed those available even in Linux

    I am curious to hear more about what power-user features are available for OSX, though, so please do enlighten me (seriously). I don’t get the chance to play around with OSX as much as Gnome, so I know far less about some of the features available for OSX. I’m not sure that it necessarily takes any large amount of time to get the tools for Linux, but perhaps some time to take full advantage of them, though I think that’s likely true for the tools in OSX too (but as I said I’m curious to learn more about them so that I can make a fair comparison).

    Reply  |  Quote
  20. Josh UNITED STATES Mozilla Firefox Windows says:

    “My point is that while most people think that novice users need big shiny buttons, it is not always the case. You can teach them CLI just as easily.”

    It is like saying that when you throw a coin up, is is not always the case that it would land heads up or tails up. Because once in a blue moon it can stuck edge down into the ground. How many consumer electronic devices are out there that are controlled by a terminal with a full keyboard VS big shiny buttons???

    Just because you can teach something easily does not mean it is better. Some bad examples of GUI do not translate that CLI is better for novice users. GUI was designed to make it easier and friendlier to interface with a PC. The large number of commands your PC can perform is always ahead of limited GUI options, this makes the case for CLI. But do novice users get to that point right away? Most likely they get there because the provided GUI failed.

    Reply  |  Quote
  21. James Heaver UNITED KINGDOM Mozilla Firefox Windows says:

    A great example of users of CLI / GUI are professional secretaries. People who use computers an awful lot of the time, know some deep intracacies of their office applications, but have no interest or understanding of the underlaying system or OS.

    You can generally tell how old a secretary is (or atleast how long they have been a secretary) based on their opinion of the GUI vs using the keyboard. Many of these people grew up using wordperfect and other dos based systems, and those that did swear by them.

    On the other hand, the younger ones who have only used win/office much prefer that.

    The other aspect in the GUI/CLI debate is that GUI will sell better becuase it is *shinier*. Allot of users, both normal people and geeks, like shiny things.
    I don’t know why – it could be magpie syndrome, the shinier it is the more like HAL/tron it is, the fact that it looks like a newer version, whatever. But the fact is we like to buy into shiny, even if we don’t like using it.

    This isn’t an argument against eye candy, I love eye candy, and think its wonderful. Its just that shiny has an advantage commercially.

    Reply  |  Quote
  22. Ryan UNITED STATES Mozilla Firefox Windows says:

    I’ve been exclusively a Windows user basically until freshman year in college, when I learned about Linux. I use Linux as my secondary OS now (after a spell as my primary one). For general background, I’m hardly a CLI expert, but I can find my way around. Likewise, some GUI’s make life easier, while others I have to navigate around for ages to get what I want.

    Linux GUI’s by and large piss me off because there are many redundant interfaces and only some of which work. After trying a slew of distros, I settled on Centos 5 as I was looking for something that wouldn’t have to be updated every six-nine months (I refuse to upgrade a distro between x.0 releases even if utilities are offered because something always isn’t upgraded properly).

    Recently on Centos 5, i was upgrading the fglrx driver to the next release (Red Hat is one of two OS that AMD’s drivers officially support) and ran into a problem where my screen resolution was capped one notch below native res, none of the GUI’s could help me, so I turned to the CLI, but xorg.conf revealed nothing either, while ati’s CLI utility wouldn’t allow me to set native res, so I gave up and decided to revert back to the previous driver. This was a chore as I had to hunt down a slew of files to delete which the install package did not. I successfully reverted to the old driver, but what should have been a five minute upgrade became an hour long fiasco, with the most problem-free distribution I’ve worked with to date. Had I run into this problem 18 months ago I would have reinstalled Centos.

    In this case, the command line saved me as I simply couldn’t perform the tasks I needed to from a GUI (there are three ways to set my screen res from within GNOME).

    Were this Windows, I certainly would have been going through the GUI and the driver uninstall would have been fairly painless (assuming Windows didn’t revert to driver cache and reinstall the driver I just removed, which can be easily averted) because the GUI for the Graphics Driver in Windows has no duplicate and is relatively robust.

    GUI’s are much less flexible because every option has to be displayed and thus reason has to be employed as to what is shown, but they should be able to solve most of the simple problems. CLI’s are extremely useful for tracking down lots of random and obscure problems as well as performing some search tasks (I love grep), but they are overkill for simply changing your resolution because you must know a few things: 1 What executable do I call 2 What are its options? Now, its not hard to figure out 2, but I actually have no idea what I’d call for 1 and would have to do some hunting to find the command (xorg.conf is not the right answer in this case), whereas with a GUI, I can simply pick through the control panel, which has much fewer options. Were I to drop my grandfather in that same situation, he could probably find the option in the control panel for changing the resolution, but for him to find that in the command line would require a much better memory (if he knew what it was called at one point), or some fairly sophisticated searching through *bin/ directories (or google).

    I must echo what has been said above, CLI is great for some tasks, but is overly complicated for others. The GUI can take some getting used to, but its scope is much more limited, so I can learn where everything is much faster and for most one step tasks it seems quicker. No matter how long you’ve used a computer, you’ll always learn something new from using the command line as its simply too powerful and too complicated to know all the ins and outs of every command with every option, but I can say with confidence that I’ve found most if not all of the control panel-esque GUIs in Windows XP (and I’m getting there with Vista) because they are relatively finite.

    I would like to reinforce the point that we generally go with what works and for me the Linux CLI rarely fails me (in fact its never failed me, but there are times when I simply had no idea what to do, what command to call, so I sort of fail it) while the GUIs fail all too frequently. In Windows the GUIs rarely fail me, so I generally don’t have to turn to the command line to solve anything save a couple tasks where I get more, clearer information from the CLI approach.

    Hope you survived this absurdly long response ;).

    Reply  |  Quote
  23. Luke Maciak UNITED STATES Mozilla Firefox Ubuntu Linux Terminalist says:

    [quote post=”2177″]How many consumer electronic devices are out there that are controlled by a terminal with a full keyboard VS big shiny buttons???[/quote]

    Not many, but we are approaching a point where hybrid search based systems are becoming more popular. User can just pull up Launchy, Katapult or Spotlight or whatever, type in email and and the correct email client will open up.

    Then they can pull it up and type in “TPS Report” and they will get a pull down list of recently edited reports – clicking on one will open it for them.

    Context aware search will be the next dominant UI paradigm – I’m pretty sure of that. We will be seeing less toolbars and menus and more search boxes. We will see less of Microsoft style nested modal dialogs for configuration, and more of web-page like settings screens complete with exhaustive explanations for each line item (Vista already seems to be going in that direction with some of their new dialog choices).

    So we are really going back to the basics – CLI and config files, but this time around they do have shiny buttons. ;)

    So let me reiterate again: I’m not saying CLI is the king. It can be a pain, and it can be confusing. But so can be a GUI. I’m just saying that GUI is not the perfect choice either despite of what most people think. Having a GUI doesn’t mean that application is automatically easier – it is not always the case as I illustrated in previous comments.

    A well designed GUI does make a difference. Conversely, a well designed CLI can sometimes offer as much or more functionality. And sometimes a GUI can simply be a CLI in disguise with few shiny buttons thrown in for good measure.

    Reply  |  Quote
  24. Johan SWEDEN Konqueror Debian GNU/Linux says:

    About OSX being userfriendly? Whats more userfriendly than the Gnome/KDE-menus? Different categories for Network, Games, Office etc. That’s about all my mom or dad would need.
    “User Friendlyness” is so wornout. To be able to do something without actually knowing what you’re doing isn’t user-friendly.. it’s just plain stupid.
    And btw.. OSX shouldn’t be compared to Linux, since it is proprietary crap that’s only available with some computers.

    Reply  |  Quote
  25. Kiyu UNITED STATES Safari Mac OS says:

    @Ian: I didn’t mean to come off sounding so elitist myself (apology here), but my point was that only a fraction of the total computer user base use keyboard shortcuts. From those who do, remove the people who use Cut/Copy/Paste and Ctrl-Alt-Del (windows dig here) and you have a VERY small group (relatively speaking) – this would constitute some degree of being elite.

    Now, to @Johan: I shall be less polite: proprietary doesn’t equal crap. Some of the best innovations in software were either IN proprietary software or PAID FOR BY proprietary software sales and you know it. Get off your high horse.

    Too many people are WAY too obsessed over the philosophy of software “openness” – this isn’t the global environment or starving kids in Africa – it is an economic and artistic preference left up to the artist who writes the software or the company who PAYS the artist. Would you say that all paintings, songs, movies, books or photographs not freely licensed for everyone’s use are crap?


    Reply  |  Quote
  26. Mackenzie UNITED KINGDOM Safari Mac OS says:

    Lost about a 5 paragraph rant yesterday, so here goes, mapping out the arguement for GUI.

    First Point: GUI is superior in every way too command line for usability, and a CLI is only relevent to tasks where the GUI is poorly designed.

    I’m gonna explain this step by step starting at the most obvious and fundemental differences and what they mean.

    ******Percieved and Recieved information in CLI and GUI respectively******

    CLI is a Percieved language. It relies on computerised versions of written characters, all of which are notified by a certain sound, and meaning, when strung together in the right configurations. If you see something written in chinese, and don’t know chinese, your pretty fucked. Written information is Percieved information, we recognise a character and tag a meaning onto it that we have learnt. A CLI interface is built around Percieved languages- Without knowledge, it’s unlearnable. To learn a CLI, the raw knowledge and syntax MUST be in a textbook, (Or a list of syntax built into the machine) because learning a language from scratch is nigh on impossible. Try learning chinese using a chinese Command Line DOS system. Not gonna happen.

    Contrast this too a GUI. GUI relies on visual language, and visual data configurations that relate too real world situations or meanings, so it’s all fairly self explanatory- It is Recieved information, like a picture. You can understand a picture of a dog, even if you’ve never been told the name of a dog. You see a dog in real life, as Recieved information, and you see a dog in a picture, you put 2 and 2 together, a dog is a dog is a picture of a dog. It has no meaning other than itself. This is the beauty of recieved information. What a GUI, and, coincidentally enough, most comic book artists, will try too make a piece of complex information as “Recieved” as possible, rather than percieved. Let’s take an example. “The man was a hypnotist.” This is percieved information, because it relies on characters with learnt meanings, or a CLI, ie, DOS, English, Russian, Etc. However, if you have a picture of a man with spiralling, psycadelic eyes- It’s fairly safe too assume that everyone, regardless of what language, will be able too tell the man is a hypnotist, assuming they know what a hypnotist is. The picture is recieved information, the sentence is not.

    Since your all going “WHAT OF IT?!” the point is this a CLI is written, percieved. A GUI is not. This makes, at a most basic level, a CLI something that has too be “learnt” and a GUI something that is fairly self explanotory that can be figured out by trial and error.

    Lets take an example with a hypothetical CLI and gui which allows a drag and drop function. CLI:


    All percieved information. Unless you’ve learnt that the CLI uses the term “BOX” for directories, and you know the FUNCTIONMOVE command, your screwed.

    Not lets compare this too a picture file of a ball on a GUI, where the picture is displayed as a thumbnail on the desktop, and the directory as a picture of a box on the desktop. Simply click, drag, drop. No prior knowledge needed at all.

    The final outcome of all of this is that GUI is universal and discoverable, where’s CLI is language specific, and requires a bank of knowledge.

    Second major point: *********hardware interface***********

    Much more too the point on this one, now I’ve gone over the groundwork.

    A CLI relies on a percieved information on a plethora of keys too type in strings of code. Again, requires a bank of knowledge. Where’s a GUI relies on a point and click mouse interface- The simplest of tasks can be done by anyone using a few clicks, following self explanatory visual langauge (Ie. Cross for close, a folder for a directory, an icon for piece of software, a thumbnail for an image file, etc) once again, file extensions are percieved. The GUI makes them recieved and universal, and therefore manageable by anyone using a mouse. A mouse requires about 3 seconds of discovery- Left, action, right, menu, middle, scroll.

    Of course, you could easily run a mouse interface using a CLI- But it would be pretty obscure and meaningless. CLI relies on syntax, so drag and dropping functions into sentances is pretty weird, and pointless, considering those functions can be explained visually.

    One step up from this, and relying on mouse control, is the window system. The window system allows users too allocate different tasks and programs in different panels, effectively allowing efficient multitasking- Something that’s nigh on impossible with a CLI.

    Once we’ve got our windows, and want too start using more complex functions easily, we have the pull down menu system, making a large number of tasks accessible under each category. It’s the mouse that makes this work, and the visual idea of a pull down menu, almost like a page with different characters on it. Using a menu system through CLI is much slower and clunkier, because it’s all textual, there’s no point and click. With point and click, the text ceases too be percieved, and becomes pretty obscure logos and icons for pathways too other functions, since it’s made into a visual button by the menu system, rather than an typeable/speakable percieved command. A CLI that uses a mouse isn’t really a CLI. It’s an absurdly badly designed GUI.

    Epilogue (Lol)

    So, we’ve learnt basic differences in means of communication, and how these work with hardware. It’s pretty obvious that a CLI is now obsolete in all but the most obscure and complex of tasks, that arise from an unstable GUI or from a GUI that isn’t flexible enough too achieve the task. There is nothing that cannot be changed into a logo or self explanatory visual icon. One look at the plethora of buttons on internet explorer will explain this.

    For those of you who remain convinced theres still a place for CLI since the advent of visual communication (GUI, Album covers, Logos, Street signs, TV, Illustration, Etc, etc, etc) check out Peter by Gemtree for windows. Neat little coding engine where EVERYTHING is compiled using algorithm trees of icons representing functions. Then go pick a copy of “Understanding Comics” by Scott McCloud, and check out the Copy/Paste logos on Microsoft word. Or even Mac OSX’s Automater, too create workflows. Or maybe a copy of Neal Stephensons “Diamond Age” where language has been almost totally forgetton except by elite clans- To be replaced with functional logos and icons.

    GUI is the universal medium.

    Reply  |  Quote
  27. Ian Clifton UNITED STATES Mozilla Firefox Mac OS says:

    Fair enough; it’s definitely true that most users (sadly) do not take advantage of keyboard shortcuts. Apology accepted and I apologize if I sounded harsh.

    You have some interesting points that remind me of Saussure and others, but I think you are neglecting some things. Visual language is culturally based. An “X” does not universally mean “close” (in fact, it has a different meaning in OSX compared to Windows). A dialog asking if you want to “Save” or “Discard Changes” is not visually intuitive despite being a visual interface. You can add colors to it, say green for save and red for discard, but those are culturally based as well. Red is significant in all (or nearly all) cultures, but the meaning is very different. The same is true of all colors (such as black and white). Is there a universal shape that says “save?” One might say a disc, but what about people who are growing up with the technology now. Hell, if their experience is based on a MacBook Air, they will have no idea what a disk/disc is!

    Shapes are essentially the same thing as letters: visual representations of something. I don’t think that a “+” would be universally known to maximize (or increase) window size. Even in a culture where that does mean “add,” a plus usually means combine two (or more) things rather than just increment.

    The window system allows users too allocate different tasks and programs in different panels, effectively allowing efficient multitasking- Something that’s nigh on impossible with a CLI.

    I disagree (and not just because you can multitask very well on a CLI). In Windows, most applications start up maximized, so the user has to learn how to change Window size (assuming the user doesn’t think that the other thing that was open is closed now). In some cases, one dialog is more important than another, so you can’t switch. In general, though, I would say it’s easier to multitask in a GUI (and I think that’s one point where OSX really shines), but how much multitasking does the computer-newbie do?

    Once we’ve got our windows, and want too start using more complex functions easily, we have the pull down menu system, making a large number of tasks accessible under each category.

    Which is based on language just like a CLI. You never click on an octagon that brings down a square, triangle, and a squiggly to pick what you want to do.

    Using a menu system through CLI is much slower and clunkier, because it’s all textual, there’s no point and click.

    Pointing and clicking is a LOT slower than pressing a key on the keyboard. Most menus (assuming you use them and not direct commands) have a key associated with an option (GUI or CLI, but especially CLI).

    There is nothing that cannot be changed into a logo or self explanatory visual icon.

    Can you give me an example of a GUI interface for simply moving files… but one that allows you to be specific (e.g., “move all files,” “move all files with ‘blah’ in the name,” “move files ending in .blah,” etc.) using icons? Would it honestly be self-explanatory?

    Reply  |  Quote
  28. Luke Maciak UNITED STATES Mozilla Firefox Ubuntu Linux Terminalist says:

    @Mackenzie – wow, very insightful post. I have one counterpoint though. You talk about using DOS in Chinese. How about using WinXP in that same language? Can you honestly say you would feel comfortable using a localized Chinese version of Windows? Sure you would probably be able to drag and drop things on the desktop, but not much less beyond that. Especially if your keyboard would have Chinese characters on it.

    Here is a real life story from one of the blogs I follow – K. Mandla recently moved to Japan and had major issues configuring the router because it’s web based UI was all in Kanji. It is a graphical interface, and you interact with it by clicking on buttons, typing in numbers or checking boxes. Still, you can’t really understand what is going on because the meaning of these buttons and boxes is still tied to particular language.

    So when discussing different types of interfaces we probably should assume they are in the same language just to be fair.

    Also, a lot of GUI conventions are intuitive to us, really make no sense.

    For example, why do you press Start to shut down your computer? How come “Print Preview” is usually under File menu and not under View menu? How come I need to click on the “Floppy” icon when I want to save something to my hard drive?

    Why do I need to double click on desktop icons, but single click on toolbar and taskbar icons? Yesterday I talked with a guy who would always open 3 or 4 sessions of IE because he would double click the icon in quick launch.

    Few years ago I worked with a lady who would open My Documents by right clicking the icon on the desktop and choosing “Explore” because she never could get a hang of double clicking.

    Both types of interfaces require you to develop certain set of skills. A CLI requires you to learn a vocabulary of commands. A GUI requires you to develop set of skills, and learn a set of conventions (ie. you double click these things, you click and drag by holding the left mouse button, you open context menu with the left and etc).

    And for the record comparing a complete CLI novice to someone who have been exposed to many GUI interfaces over the years and already developed a set of skills is not fair. To really compare the interfaces we do need to use the “flaky aunt Sally who never used a computer” example.

    Otherwise you are comparing a seasoned used to absolute novice.

    @Ian – oh hey, I was thinking among the same line on the language thing. :mrgreen:

    @Everyoone – I seriously envy those of you who hang out with competent computer users all the time. I actually teach introductory computer class at college so I get to see how people use computers during the lab session. Sometimes I wonder how these people tie their shoes in the morning…

    I mean, don’t get me wrong – they are all on Myspace, and they all use IM. But I usually spend 10 minutes at the beginning and 10 minutes at the end of every lab for the first 2-3 months showing them how to zip up and submit their work (ie: hilight, right click, choose send to zip archive). These are undergrad college freshmen and sophomores. You’d figure that this generation would know more about computers than aunt Sally does… And they do – unlike aunt Sally they know how to use Myspace, Facebook, Youtube, AIM and perhaps limewire (I’ve been told bittorent is “like confusing”) or some other shitty, spyware ridden p2p app. Nothing beyond that. They know how to use certain tools, but have no concept of how the operating system works at all.

    Reply  |  Quote
  29. Mackenzie UNITED KINGDOM Safari Mac OS says:

    To continue the sparring- The GUI’s we use everyday are not really GUI’s. Their CLI’s with pretty skins, and a desktop base. Hence all the dropdown menus are textual.

    I do graphic design at college so I have too figure out this shit for my studies. I often like too use the viewpoint of somone who has knowledge of the basic physics and such of the real world, but an inability too read. For example, I had too design a logo for a “Here’s an idea” campaign for the start of my A levels. The Brief was pretty vague, so it thought I’d come up with a logo for the communication of information that was pretty self explanatory. Overall, I produced a banner with a rounded speech bubble quiff at one end, and ending in a blocky arrow, to show the direction of spoken word. I’ll post if I find the image file anywhere.

    Anyway, back too my main point. What we called GUI’s are really pretty skins, and aren’t purely “Graphical.” I figure for something that’s totally accessible, depth has too be sacrificed, until designing a neat little means for combining functions.

    Ie, The mouse is normally an arrow. To me, a picture of a hand seems more intuitive.

    Gonna look at my prefs list on safari and see what I can come up with up with off the top of my head..

    “About”: The universal (i) Tourist information logo.
    “Private browsing” AOL has a neat little All Seeing Eye Logo for it’s MSN function. Perhaps an eye with a red NO circle/line through. And such.
    “Block popups” Window icon, again with cross through.

    Some of these, like “Hide safari” are routes too other functions, like the cross in the corner.

    new window: Windows has a “New tab” Logo for that already.
    Open file, Print, etc, all of these, already have logos. As with Undo, copy, paste, Search, send, History, Refresh, Home, Zoom, Help. Gettin’ the idea?

    It’s just a simple idea of replacing the conventional Pull down windows with a pull down Button, (a little down arrow, hm?) which brings up a grid of function icons. Of course, the more complex the task, the harder it is too communicate visually- But that’s why percieved information was invented in the first place, arising naturally after self awareness.

    Another neat idea is combining functions. We already use this sometimes: “Copy” button, then drag and drop too duplicate an item. Or dragging and dropping a picture of an Orange onto the magnifying glass icon on the desktop. Which would bring up a whole load of text files with the word “Orange” or any other pictures of oranges. Or you could bring up a File type grid. With a picture of a painting for pictures, and the Aa logo for word files. Or cinema logo for movie files. Etc, etc, etc, etc.

    All of this is just ideas for an instantly accessible OS for basic stuff. Going in depth like configuring a router means the recieved must become percieved. It’s all very well having an internet icon (Lets have the BT one for the hell of it, two arrows rotating round a dot, it appears on their Homehubs) but without raw numerical info you can’t set it up. Though, to be purist, could be communicated using tally charts or roman numerals, but that would be a step backwards.

    As for Zipping and Unzipping- I always just drag/drop into stuffit expander, which has a cool icon of an arrow coming out a box, too say, “Unpacking”

    Everything can be done like this. For example, on mac, when dragging around a drive icon, the trash bin suddenly becomes an eject icon.

    On a bit of a tangent, I liked Ovolabs Desktop Earth program a helluva lot, as well as iTunes cover flow. Both were pretty basic, sure, but the idea was neat. I like having a Photo Album Africa. Or all my text files in Russia. Or whatever. Or refering too CD’s by what colour they make when they flash past. Infact, one of my favourite ideas is from the novel SNOWCRASH- It’s mentioned that one of the characters has his virtual space arranged with 3D folders. Instead of having a list or a grid of icons like conventional computers, his virtual space is arranged so that different files he’s categorised too different topics all follow the vertices of 3D shapes. For example, all my files in a certain folder would be arranged in 3D according too the vertices of a dodecahedron. The number of sides would increase with the number of files. Each can be dragged and dropped out the orbit, or out the form of one shape, and placed into another. And so on. Use your imaginations.

    And yeah.. I had too wiki bittorrent. I didn’t know what the hell they were talking about until I saw the animated gif and understood how it’s broken down into packets of info, and each packet sent individually, etc.

    (Lol, just realized, it’s best too assume too assume a worldview opposite too “ELIZAs” for graphic design- No knowledge of grammar or semantics, but large amount of real world knowledge.)

    Reply  |  Quote
  30. Johan SWEDEN Konqueror Debian GNU/Linux says:


    Way to obsessed you say? What’s wrong with wanting to secure fair competition and use? How we store all of our information is extremely important. The ability for poor countries to develop and infrastructure of their own is greatly improved from already existing code under non-restrictive licensing.
    And to take an exampe a bit closer to everyday: Apple selling music that’s only playable on their own devices, is that something good?
    There is a reason for wanting to keep things open. And it has nothing to do with my high horses. Too be fair though, I was in a bad mood when writing my last comment, so I might not have been the best diplomat, and for that, I apologize.
    I do know that some nice applications and innovations have come from prorietary products/companies. However, that really has nothing to do with if we should use those kinds of applications or not.

    Reply  |  Quote
  31. Luke Maciak UNITED STATES Mozilla Firefox Ubuntu Linux Terminalist says:

    @Mackenzie – Interesting ideas. I guess the other polar end of usability research is the area of voice and touch activated control which is really an extrapolation of CLI.

    For example, you sit down at a computer and say: “Open Browser” and browser pops up. You then touch the screen to scroll around the text and click links. So the control is command driven with a set vocabulary, and some smart search features – just like CLI.

    If you say help it will show you a list of common commands which you can expand. A lot of stuff like that is already done for mobile and hand held technologies.

    I don’t know which one is more intuitive, but both should be pretty accessible to people who had no prior experience with computers. I guess the UI of the future will be voice + multi-touch with an on-screen keyboard. So navigation will be a hybrid of spoken command as well as dragging, dropping and sliding. Because let’s face it – like you yourself said, some operations are hard to conceptualize with just icons and drag/drop and click operations.

    Reply  |  Quote
  32. vacri AUSTRALIA Mozilla Firefox Ubuntu Linux says:

    Mackenzie – whether you have English language text or an abstract symbol is irrelevant to the function. You have to understand the meaning of the icon in exactly the same way you understand the meaning of the text. A new user wouldn’t even understand what compression or archives are.

    Anyway, whether your stuffit icon is a box with an arrow or a text label saying ‘extract’ or ‘stuffit’ is irrelevant to the workflow. Changing dropdown menus from text to icons wouldn’t really make a gui ‘purer’, more graphical in anything but the most superficial sense – the workflow would be identical.

    You are right though, guis are skins for cli, but no amount of elite design is going to change that fact. You perform an action in a gui, the gui interprets that action into cli, and passes it on to be actioned.

    Reply  |  Quote
  33. Kiyu UNITED STATES Safari Mac OS says:


    I feel that the current state of competition IS fair. Apple’s policy of allowing buyers to share the song onto up to 5 other computers for 99 cents IS fair. Their goal has never been one of dominating their market segments – it has only been to provide a superior level of service and functionality to the people who buy their products. They sell iPods and provide the iTunes store for the people who bought them – their goal was never to build the biggest or best music distribution site. Meeting this vision required them to lock down the songs – that requirement was by the music industry. The music industry is now having a change of heart not in small part due to Apple’s desire to distribute DRM-free music. For those companies/artists which allow Apple to do so, they have higher-quality, DRM-free (higher cost) songs.

    The reason closed source software exists is because in large part, people won’t pay for something they can have for free, and most open-source software can be had for free. Apple is a business and the point of having any business is to earn profit. Making money off of authoring open-source software is really hard; among commercial open-source ventures, there are way more failures than successes. Commercial companies which contribute the most to open-source software generally have a financially sound reason for doing so.

    Apple (and pretty much any successful company) just asks: “What would people PAY for?” and that is what they make. As we all know, making good software or hardware is EXPENSIVE. The best programmers don’t work for free – they have to eat too. This is the cycle of a strong economy. If you take away the strong economy, nobody will be thinking about whether their software is open or not.

    So, I come back to my original position: all of the people posting comments here are fortunate enough to live in places and in time in human history where things are mostly very good. We are healthy and well fed, we can practice whatever religion we want, we have reasonably good governments – compared to most of the rest of human history, we live in the lap of luxury. To me, there is art that is free, and art which, if we want to use/experience it, we must earn it. Thanks to economies of scale, we all have pretty inexpensive access to essentially unlimited amounts and varieties of entertainment (music, pictures, songs, books, etc.) which took their creators months or years and anywhere from hundreds of dollars to hundreds of millions of dollars to create.

    Software is art, and I am thankful that I live in a time where people can invest themselves into making great software and giving it away. But I am not unaware of the fact that it is the art I (and others) have to EARN (pay for) that supports circumstances (strong economy) which allow people to make great software for free.

    Now, just in case you are going to say, “It isn’t about the price, it is about better, more secure software!” I will say this:

    Quality: With a few notable exceptions, when I go looking for a piece of software to do some task, and I find 10 different products, rarely is the open-source one the best.

    Security: Knowing that those who would break your code have unfettered access to it DOES encourage one to be very careful and thorough. However, I don’t think that closed-source teams *rely* on the fact that they know nobody can read their code – they still write it as tight as they can. So if both teams are writing code as tightly as they can and one codebase is readily available to those with malicious intent and one isn’t, which codebase is more secure? As far as “the public” knowing that their software is secure… they don’t know – even if it is open-source. The people who wrote it don’t even know for sure that it is secure or not! There is a false sense of security either way!

    Developing nations: they don’t *need* open-source software either. If they want it, they take it.

    The future: I look forward to that Star Trek society where nobody has nor needs money; where everyone works to better themselves and they work in teams to produce the best software ever known. But the path to that society will be built on a foundation of innovation for the purposes of commerce.


    Reply  |  Quote
  34. Mackenzie UNITED KINGDOM Safari Mac OS says:

    Vacri: That’s the whole idea. It translates it into CLI. The whole point of GUI is too make commandline recognisable as icons.

    That’s what it boils down too-Icons. Or, more specifically, Iconography. Something that can be explained pictorially, a simplified picture of real life. Of course, for “Put your trash in the trashcan” and “Emergency Exit” this is all pretty simple, but for computer language it’s massively complex.

    Whether a GUI is pure or not depends on the amount of stuff which can be Recieved Iconography as opposed too Percieved language. The workflow is totally irrelevant too all this- It’s just another action too be describe pictorially.

    Luke- Your entirely right- It would be equally intuitive for somone who spoke that language, because voice command imitates conversation. Speaking is the most natural thing in the world and for computers too incorporate this would be a massive step for usability. It’s ideal for anyone unused too a type/mouse interface, or who’s unused too the whole Computer idea. Like Dr.Whateverhernameis shouting at Will Smiths JVC stereo too “Turn Off! Deactivate! Shut down!” to no avail, it’s the ideal means too easy computer command.

    If you speak that language.

    Cross language needs to be pictorial, but it cannot explain complexity well at all. Even the idea of saving too a disk is massively hard thing too communicate with a simple icon. But we already compromise and have flags too distinguish between various language interfaces- Another small step towards the most intuitive interface.

    Reply  |  Quote
  35. Luke Maciak UNITED STATES Mozilla Firefox Ubuntu Linux Terminalist says:

    [quote post=”2177″]Cross language needs to be pictorial, but it cannot explain complexity well at all. Even the idea of saving too a disk is massively hard thing too communicate with a simple icon. But we already compromise and have flags too distinguish between various language interfaces- Another small step towards the most intuitive interface.[/quote]

    Yup – this is why we have localization. No matter what you do, UI will always have to include text and/or voice controls in a language that the user understands. Otherwise it’s an exercise in pictorial abstraction which is ok for very simple things only.

    It’s like that project they had to design pictorial warnings to prevent archaeologists thousands of years in the future from digging out nuclear waste which haven’t decayed yet. We could assume that these people would probably know as much about us as we know about ancient Egypt – ie. a lot of what we know is speculation and theories. I think they had a whole show on Discovery about this project, and the conclusion was that it was nearly impossible to design pictographs that could be instantly interpreted by people who may not speak our language or know anything about out culture.

    Similarly, creating an UI that is intuitive for anyone regardless of language is really, really hard hard unless you are doing something like the classic Play/Stop/Fast Forward/Rewind VCR buttons.

    Reply  |  Quote
  36. Alphast NETHERLANDS Mozilla Firefox Ubuntu Linux Terminalist says:

    Well, I am sorry to say, but the main problem with an OS is not per se its user friendliness, as Luke perfectly demonstrated, but how it works flawlessly on a standard machine. Windows is not as good as OSX with this, but still pretty good and a lot better than Ubuntu, for instance. Not that Microsoft is doing a good work at it, but simply because every hardware vendor develops its drivers for it. Linux on the other side (and Ubuntu in particular) has to wait for drivers to be developed by its community. This is, in my opinion, the main problem for new Linux users, not the GUI/CL approach.

    For example, I am using Ubuntu both at work and at home, along with Windows XP and 2000, all on different machines. Well, I can’t use the sound properly on the Ubuntu machines. And I found out after two months of following support forums and other sites that my only option was to waut for a new versio of ALSA to support my (antique) sound cards. No need to tell that these cards are all working perfectly under Windows…

    Reply  |  Quote
  37. Luke Maciak UNITED STATES Mozilla Firefox Ubuntu Linux Terminalist says:

    I think Apple is in a slightly different league when it comes to drivers and hardware support. I think someone mentioned it already here but Apple only supports a very limited set of configurations. You can’t really build a Mac from scratch yourself, and for the most part you can’t start swapping parts in and out to upgrade it. Since they only have 5 or 6 different hardware specs to work with they can make sure that everything works out of the box and that it works well. Furthermore they can optimize their software for these specific hardware profiles.

    Windows and Linux on the other hand are supposed to work on generic PC architecture. This means millions of different devices and components. Amazingly enough, more often than not Linux manages to get things working out of the box – mainly because they can’t rely on hardware vendors to provide them with drivers. I can’t tell you how many times I installed Windows on some box only to find that half of the hardware including the video card, audio card, network controller, wifi card and etc siply were not recognized. I would then take the Ubuntu Live CD and laugh as it detected and configured all these devices out of the box.

    So I would venture out and say that hardware support in most linux distros is mind-blowingly awesome. It’s just that some hardware just won’t work. Windows has piss poor hardware support, but every hardware vendor usually ships with a driver which may or may not be very poorly designed.

    Reply  |  Quote
  38. Alphast NETHERLANDS Mozilla Firefox Ubuntu Linux Terminalist says:

    OK. I get your point there…
    I am probably just extremely unlucky to have both sound and video card totally unsupported by Ubuntu on only two machines… :-(

    Reply  |  Quote
  39. JuEeHa FINLAND Mozilla Firefox Mac OS says:

    I use OS X and I think that it is pretty good OS, but there is one thing that I hate about it. You can’t really customize the GUI behavior. I am not saying that OS X’s default GUI behavior is bad or anything, but it is quite restricting for me. I bought my iBook because I want to use as many operating systems as possible and now I kinda regret it. It isn’t really that bad after you get used to it, but I really miss my GNU/Linux+MWM setup I had on my old computer. I hope I don’t start another flame war on this subject, but I think that any UI I have used except MS Windows is better than OS X’s. Even my old System 6.

    Reply  |  Quote

Leave a Reply

Your email address will not be published. Required fields are marked *