Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Why choose a Linux Desktop?

By phinance in Culture
Tue Jan 02, 2001 at 11:08:25 AM EST
Tags: Technology (all tags)
Technology

Linux has gotten a lot more user-friendly in the past couple of years because of KDE and GNOME, but I still have to ask, "Why would anyone switch from Windows?"


Linux is more robust -- and this is the most compelling reason for me to use it -- but that's not enough to persuade Windows users. A computer not crashing is not something you readily appreciate, just like you don't sit quietly in your apartment and continually rejoice that there are no loud noises disturbing your concentration.

Linux is open source. When a typical user sits down to write an email or browse the web, they don't really notice this aspect. I (and many Linux users) have been quite pleased to have been able to fix a nagging bug in an otherwise useful program, or to look at some code to "see how it's done", but we don't expect these things to have mass appeal.

As it stands, on the usability front, Linux is still catching up. I can now use my Epson color printer with Linux/CUPS, but that crashes quite a bit. Mandrake installs easily, but doesn't upgrade so well. I can watch video clips on Linux -- theoretically. Fonts will be antialiased soon. Easy andministration tools are, let's say, generally considered to be a good idea. And although I hear one can edit MS Word and Excel documents on Linux, I haven't had much success yet. In general, all of the stuff people want is there, but it's all still quite rough -- with some "shiny spots" here and there (like the extensible and user-friendly Konqueror browser).

So is the plan to just keep polishing? That's not enough. Linux needs something better. It needs something to make it worth the time it'll take the Windows user to install and learn this new operating system.

Are there any projects aimed at creating free, open source applications or technolgies for Linux that are better than those available for Windows? Any ideas?

Here's an idea of mine. Experiment with a "learning" user interface: Try to get the comboboxes to better guess which entries I'm interested in. Make the progress bars better estimate how much progress has actually been made. Auto-place the next window in the right place. Automatically create mail filters: I can drag and drop a message from my inbox to a folder with kmail. kmail could pay attention and learn to organize the messages automatically. These are just a few examples of where learning could help. What is, IMHO, good about these learning examples is that they don't require intense research (so people can hack at them in their spare time -- the learning algorithms already exist), but they are features that other operating systems (as far as I know) don't have.

What are your ideas?

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Related Links
o KDE
o GNOME
o Konqueror
o Auto-place the next window
o Also by phinance


Display: Sort:
Why choose a Linux Desktop? | 49 comments (47 topical, 2 editorial, 0 hidden)
Robustness (2.66 / 6) (#1)
by jfpoole on Tue Jan 02, 2001 at 12:28:53 AM EST

Linux is more robust...

More robust, sure, but do uptimes of more than a couple weeks really matter for a desktop or workstation OS? For example: it's rare that I leave my workstation on for more than a week or so (as I usually power it down for weekends). Plus, people such as my parents who use a computer casually tend to turn their computer off every night, and might only use it for an hour or two. For them, the instability of Win9x isn't an issue. As for Windows NT/2000, both are quite capable of staying up for 2+ weeks. Thus, I'm not sure how this is a win for the desktop.

-j

uptimes (3.62 / 8) (#10)
by Anonymous 6522 on Tue Jan 02, 2001 at 04:16:17 AM EST

I see uptimes not as a way to judge how long I can leave my computer on, but as a way to determine that it will crash while I'm using it. If OS A will usually stay up for a month and OS B can only manage two weeks, I would assume that OS A is twice as sable as OS B. The less I have to cycle the power the better.



[ Parent ]
Re: uptimes (3.00 / 1) (#29)
by jfpoole on Tue Jan 02, 2001 at 03:53:05 PM EST

I see uptimes not as a way to judge how long I can leave my computer on, but as a way to determine that it will crash while I'm using it. If OS A will usually stay up for a month and OS B can only manage two weeks, I would assume that OS A is twice as sable as OS B. The less I have to cycle the power the better.

True, but at a certain point it becomes moot (i.e., the whole idea of diminishing returns kicks in). If I've one OS that can stay up for 1 month and another that can stay up for 2 months, then while the latter may be more stable, chances are I'll reboot for reasons other than OS crashes before one month is up (such as adding or removing hardware, leaving the computer for a weekend, etc).

As a personal aside, while Linux may be more stable than Windows NT, in my experience Windows NT is stable enough for my needs as a desktop/workstation OS. It's rare that I have to reboot the computer due to NT crashing.

-j

[ Parent ]

Re: Robustness (3.00 / 1) (#25)
by CrayDrygu on Tue Jan 02, 2001 at 01:45:36 PM EST

Plus, people such as my parents who use a computer casually tend to turn their computer off every night, and might only use it for an hour or two. For them, the instability of Win9x isn't an issue.

That's odd. I come to the exact opposite conclusion -- my mother doesn't get much time she can spend at her computer, so it's really frustrating for her when it crashes. Sometimes she only has an hour she can set aside to check email and whatever else she does. She can't afford to spend half that time rebooting.

[ Parent ]

Stability (none / 0) (#47)
by Robert Uhl on Sun Jan 07, 2001 at 04:27:09 PM EST

Do uptimes of more than a couple weeks really matter for a desktop or workstation OS?

They sure do. My machine at work crashes about once every other day, generally when I'm trying to do something important. For a few weeks I could only print one out of every two times. This impacted my work, as you might imagine. The only way to fix the problems is to rebrick, but that woul require re-installing all the non-standard stuff I have.

Meanwhile, my Linux box reached 112 days uptime before a power drop killed it New Year's Day. We have Unix machines at work which have been up for more then a year. We reboot our NT machines once a month, or else they get cranky. OTOH, we reboot a Solaris machine weekly 'cause it's running Java...

[ Parent ]

Aggghhh nooooooo! (4.25 / 8) (#3)
by haakon on Tue Jan 02, 2001 at 12:37:51 AM EST

Can you imagine how infuriating it would be to have a menu shuffle itself around depending on what you select and what you don't. Every time you open the menu, dropbox, whatever you would have to relearn it because your frame of refence would have changed.

Office 2000 will hide menu items that aren't selected often and it drives me nuts.



You can turn that off (4.00 / 9) (#9)
by Anonymous 6522 on Tue Jan 02, 2001 at 04:07:27 AM EST

You can stop Office 2000 from hiding menu items my changing some option, I don't remember where, probably on the preferences screen.

I make it a point to at least try to turn off every part of a UI that annoys me. (Java & Javascript on web browsers, funky menus and paperclip in Office, etc)

[ Parent ]
That you can (3.60 / 5) (#16)
by haakon on Tue Jan 02, 2001 at 09:51:39 AM EST

But is your average newbie going to find it or realise that the behavour can
be changed?

Especially if Office has hidden the Prefence item in the Menus :)

[ Parent ]
A Note (4.00 / 1) (#37)
by pwhysall on Wed Jan 03, 2001 at 03:54:54 AM EST

In Office applications, this behaviour is controlled by the Customize dialogue, which is found under the Tools menu.

And it's never hidden :)
--
Peter
K5 Editors
I'm going to wager that the story keeps getting dumped because it is a steaming pile of badly formatted fool-meme.
CheeseBurgerBrown
[ Parent ]

My god. (none / 0) (#41)
by haakon on Wed Jan 03, 2001 at 08:55:24 AM EST

I guess you can paint me stunned :P

[ Parent ]
and maybe a paper clip would help (4.30 / 10) (#4)
by Friendless on Tue Jan 02, 2001 at 12:44:53 AM EST

I don't think the learning user interface is going to be successful. Windows 2000 makes my Start menu "easier" to use by hiding infrequently used items, which results in there hardly being any entries in it. Attempts such as Microsoft Bob and the Sodding Paper Clip to help people have failed miserably every time. Maybe people just want their user interface to stay the same and to work properly. User interface features I would like to see are:
  • A consistent set of common dialogs, which are at least as good as the Windows ones, remember frequently used directories, are resizeable and remember their size.
  • Editable chrome, as in the Mozilla Cthulthu project, where I can change XML to reformat dialog boxes.
  • Right click on an email to add a "Trash everything from this guy" filter.
  • Netscape on Linux uses Ctrl keys instead of Alt keys
  • Cut and paste always works
  • Every trace of Emacs removed.
  • The ability to turn features off, e.g. I never ever ever want to select text and drag it around in a text editor, yet I do it by accident all the time.
In summary, I think there are a billion rotten things in GUIs on Windows, and Linux can make significant progress just by matching Windows and fixing the stupid things. Look at how much nicer StarOffice is than Word, just by doing things the right way. I agree that if we can think of revolutionary things to do, then put them in, but I think the goodwill of developers is already a major advantage for Linux, and it is only lack of features which Windows already has which is preventing a mass migration.

Some comments on your suggestions (4.25 / 4) (#20)
by YellowBook on Tue Jan 02, 2001 at 01:10:12 PM EST

A consistent set of common dialogs, which are at least as good as the Windows ones, remember frequently used directories, are resizeable and remember their size.

This would be nice, and it looks like it would be something quite easy to add to, say, GNOME. In fact, I think I'll code up such a compound widget in PyGTK just to play with it for a bit. IMHO the current GNOME common file dialog is slightly better than the Windows one, because it has tab completion of filenames in its type-in box, but it could stand to borrow some features from, say, Boomerang for the Mac, which is more or less what you're suggesting above.

Netscape on Linux uses Ctrl keys instead of Alt keys

They've "fixed" this in Mozilla, and I find it sod-all annoying. My fingers want to use the Mac flower-power key combos for closing windows and quitting applications, and that maps to Alt on PeeCee hardware. Under netscape 4, it was right (Alt-W closed a window).

Right click on an email to add a "Trash everything from this guy" filter.

Your mail program doesn't already do this?

Cut and paste always works

Agreed, and it should be extended to work with any data type (just MIME encode the damned things and let the application sort out what to do with it). This is partly an X-level issue. I still prefer the traditional Unix model of active selection vs. the Mac/Windows model of explicit copying to the clipboard.

Every trace of Emacs removed.

Unless you want it there! I would pay good money to have every instance of a GtkText widget replaced by an embedded GtkXEmacs widget (providing a buffer in an already-running XEmacs server process). The embedding is possible today, but the global replacement isn't.

A couple of your points are application-specific; you should really take them up with the application developers. The toolkit stuff is well-taken, though.



[ Parent ]
Star Office (3.25 / 4) (#27)
by kagaku_ninja on Tue Jan 02, 2001 at 03:50:23 PM EST

While I agree with most of what you say... Star Office is lower quality than Word (and I am not talking about paper clips).

I've had Star Office blow up on me far more often than Word. It has a lot of annoying "helpful" features that are hard to figure out how to turn off (OK, most M$ products have the same problem). Star Office also likes to reset my fonts at random times (the most obvious one: position the cursor at the end of a line, then hit the right-arrow key; it will reset to the Star Office default 12 point font).

I am using it in an attempt to (slowly) wean myself away from M$ products...

[ Parent ]
If you want to know about GUI then go see Nintendo (none / 0) (#48)
by fullerine on Fri Jan 12, 2001 at 09:37:50 PM EST

When did removing control from the user become a desirable thing?

The windows 2000 Start Menu took about an hour to annoy me when it started hiding the things it thought I didn't want to use. If I wanted to hide them I'd hide them.

Anyway, how about a default (arcade?) setup to get you quickly into the GUI and then a more complicated (simulation?) set of options for when you want to become efficient.

[ Parent ]

No need to be so complex (3.85 / 7) (#5)
by adamsc on Tue Jan 02, 2001 at 12:51:54 AM EST

I disagree that polishing isn't enough on its own. Windows has a horrible, extremely agravating interface - and it's still better than any of the free ones. BeOS has a much cleaner interface without a lot of the annoyances, but the OS simply isn't there on a capabilities standpoint - it needs more drivers and significant expansion on the . If the opensource projects were polished to the point where they were cleaner than Windows, that alone would be enough reason to switch assuming other software was available (e.g. quality office suites, browsers).

Bear in mind that when I talk about something being "cleaner" I'm not talking about pixels. I'm talking about how well the interface a) matches its conceptual model, b) doesn't get in the user's way and c) works as the average user would expect. This also includes things like making sure you have keyboard shortcuts for everything (= full CUA) and reasonable customization (e.g. being able to disable nagging dialogs).

The main reason why I worry about the idea of trying to build in more intelligent behaviour is simply that a lot of open source projects implement some neat ideas and never progress further. We need more than a few good ideas - they have to be well executed and, more importantly, well integrated so that people aren't left with the feeling that if they could just combine 5 projects, they'd have the 1 program they want.

the problem is incentive and time.. (3.87 / 8) (#6)
by rebelcool on Tue Jan 02, 2001 at 01:03:00 AM EST

there simply isnt enough hours in the day to work on volunteer software. Most people have jobs, whether it be as a developer of commercial software or elsewhere for most of the day because we have to earn a living somehow.

So when we get home we can put maybe a couple of hours into developing. Whereas in a work environment where you are being paid cold hard cash for your work (yes i know, money isnt everything, but you need it to live) and have several hours to refine what you're doing..inevitably the work done for money is going to be of higher quality than the volunteer work.

Now I've read all the things about how the person who wants to work on something will do a better job than a person who's just told to.. but don't most people in the free countries pick a career that interests them even slightly and go for that? That in itself is "wanting to work on something". Getting paid for it doesnt hurt either.

Anyways, to sum all this up. Not enough reward, not enough time in the day to do alot of volunteer work. Nor the raw funds that companies such as microsoft have to throw at research in how people use computers, what they want etc.

COG. Build your own community. Free, easy, powerful. Demo site

"Learning" interfaces (4.55 / 9) (#7)
by Broco on Tue Jan 02, 2001 at 01:08:10 AM EST

Honestly, I don't really like the idea of having a program guess what you want. The problem I have with it is consistency. For me, one of the key aspects of an easy-to-use interface is that it always does what you expect. For example, if a control appears in one location the first time I use it, I expect it to be in the same place next time. If it's anywhere else, I'll be confused and annoyed. Even if it moved to a place where it's more useful to me, I'll still be annoyed, because my expectation is what counts most.

As a concrete example, newer versions of windows hide by default start menu shortcuts that aren't used very often. I hate this feature with a passion. I usually remember where a shortcut is, and if it's no longer where I put it, I have to spend several annoyed moments figuring out where it went.

Your suggestion of automagically creating mail filters would be even more irritating to me, especially since there'd be a high probabilty of the program getting things wrong. Also, without giving it much thought, it strikes me that the AI for such a feature would have to be very complex and resource-eating, considering all the different factors that could affect filtering.

And it would be far worse if a program actually got things wrong and made things less useful. Windows can be like this sometimes. I once decided to modify a Windows 95 .ini file for some reason or another. When I rebooted, Windows, in its infinite wisdom, fired up some sort of registry checker and replaced my "corrupted" registry with a "recent" (2-year-old) backup without prompting me. Gee, thanks. That's an example of how such heuristics can go wrong. Rather than re-install everything, that's when I decided to move to Linux :). Linux never did anything like that to me and I'd like it to stay that way.

As for progress bars that have some semblance of accuracy, I'd be all for that also, but in practice that's much harder than it sounds. Often, the only way to find out how long an action will take is to actually perform it and see. And does it really matter? Nice progress bars are near the bottom of the list of improvements I'd like made to any software.

I think that if something can be automated or done more efficiently, the program should allow users to configure it to their liking rather than try to figure it out for itself through AI. Linux tends towards the former, not the latter, and I definitely don't want this to change.

Still, interesting writeup. +1 :).

Klingon function calls do not have "parameters" - they have "arguments" - and they ALWAYS WIN THEM.

But ... (2.00 / 3) (#21)
by Bad Mojo on Tue Jan 02, 2001 at 01:10:18 PM EST

"As a concrete example, newer versions of windows hide by default start menu shortcuts that aren't used very often. I hate this feature with a passion. I usually remember where a shortcut is, and if it's no longer where I put it, I have to spend several annoyed moments figuring out where it went."

But this feature is consistant. It follows specific rules that determine where it `hides' them and when to do that. It can also be turned off. Lack of familiarity with an interface shouldn't be used to judge the interface.

"The problem I have with it is consistency. For me, one of the key aspects of an easy-to-use interface is that it always does what you expect."

This feature of hiding selections from the start menu and drop down menus is consistant. You just aren't used to it.



-Bad Mojo
"The purpose of writing is to inflate weak ideas, obscure pure reasoning, and inhibit clarity. With a little practice, writing can be an intimidating and impenetrable fog!"
B. Watterson's Calvin - "Calvin & Hobbes"

[ Parent ]

A relevant passage... (4.25 / 4) (#31)
by kjeldar on Tue Jan 02, 2001 at 04:57:27 PM EST

A program should follow the "Law of Least Astonishment". What is this law? It is simply that the program should always respond to the user in the way that astonishes him least.

A program, no matter how complex, should act as a single unit. The program should be directed by the logic within rather than by outward appearances.

If the program fails in these requirements, it will be in a state of disorder and confusion. The only way to correct this is to rewrite the program.

The Tao of Programming, Book IV

[ Parent ]
Part of the problem (3.16 / 6) (#8)
by pope nihil on Tue Jan 02, 2001 at 03:59:08 AM EST

In my opinion, a large part of the problem with being able to use a nice GUI environment under linux (or your favorite unix[-like] OS) is X. Now, I know that a lot of people work very hard to make X useable, and for what it was meant (networkable GUI), it is fine.

However, the primitive nature of the X protocol has led to all sorts of problems. If you run a variety of programs, it's nearly impossible for all of them to have a unified look and feel. This is partly because different people wrote the software, and partly because there is no unified windowing toolkit for X. While this leads to a diverse assortment of window managers and GUI toolkits for a programmer to choose from, it doesn't make things easy for the user. No one is guaranteed that there will be a little [x] button at the top right of a window to close the program out. Additional problems arise with X not being fast enough on workstations. This is currently being worked on with the direct rendering stuff, but if I'm not mistaken, each video card requires a special kernel module to take advantage of DRI. What about the BSDs who can't use GPL'd kernel modules? Does X just continue to suck for them? As far as I can tell, yes (please correct me if I'm wrong here).

What we need is a standard windowing system framework: widgets that can be customized, but will be unified throughout all your applications. a GUI application template that has standard menus not provided by some window manager, but by the GUI application framework. Hopefully, this would be developed under a new windowing system altogether. Perhaps this will come about through berlin or some other project. I don't really know. However, X can only be patched so much before it becomes something else entirely or can't be maintained.

I voted.

FUD (4.00 / 5) (#17)
by evvk on Tue Jan 02, 2001 at 10:31:39 AM EST

When will people get this? The problem of inconsistency does not originate from X (X server, Xlib) itself but from the lack of standard GUI libraries on top of X. If there was a central authority that could just say "this is the standard library and all programs should use it" there'd be no inconsistency problems. Motif mostly _was_ such on commercial *nixes but not on the free ones. Free software people will never agree on a standard toolkit, because people like different kind of things and it is, in my opinion, too late now for anyone to say what to use. I know I wouldn't give in to using gtk. (Well, I don't like any widget toolkit or widgets anyway... All the X programs I use are xterm, netscape, gv and xdvi plus the occasional games. I thus couldn't care less about the consistency of some lame widget programs.)

No, X is not perfect and could really use some updates in graphical capabilities (no, not something as ugly as antialias but rather alpha channels and other improvements in image support) and major clean-up and simplification but that has nothing to do with the consistency problem. And I certainly don't want to give up window managers and use something as crappy as the window management is in "consistent and user friendly" systems as Windows. I can't run my xterms in such an environment efficiently.


[ Parent ]
Another Thing... (4.00 / 1) (#34)
by Matrix on Tue Jan 02, 2001 at 07:39:50 PM EST

Having a bunch of widget sets, window managers, etc. makes X far more adaptable to user preferences than Windows. You may like doing things one way, I might like another. Which of us is right, and which should the company designing the GUI system (in the case of Windows) try to please? The X/Unix way of doing things lets you, with varying degrees of difficulty, change a wide variety of things about your UI. Yes, the defaults could sometimes use a bit of work... But generally, the choice available makes up for that.


Matrix
"...Pulling together is the aim of despotism and tyranny. Free men pull in all kinds of directions. It's the only way to make progress."
- Lord Vetinari, pg 312 of the Truth, a Discworld novel by Terry Pratchett
[ Parent ]

Freedom of choice? (3.50 / 8) (#11)
by yojimbo-san on Tue Jan 02, 2001 at 05:43:18 AM EST

"Why would anyone switch from Windows?"

None of your business. What's wrong with allowing them the choice?

"Why would anyone want to drive any car except a Ford?"

Personally, I mainly use Windows, especially where I don't care about the machine, but am interested in the tasks (i.e. I want to word process, read email, surf the web). I use Linux desktop on machines where I want to hack, because I can grok the whole thing better, but that's just me.

So don't diss the choice, especially if your main grounds are "it just isn't as easy for me as Windows is".

Quick wafting zephyrs vex bold Jim
yo (2.80 / 5) (#12)
by daevt on Tue Jan 02, 2001 at 08:21:07 AM EST

there must be some kind of project for a learning intereface some where, if not i think the idea is a good one and would be a big feather in any cap to say, "you dont have to learn our system, as our system will learn you".
yo
Linux more robust? (4.08 / 12) (#13)
by joto on Tue Jan 02, 2001 at 08:47:44 AM EST

Is Linux really more robust? That might be true for server applications. Well, come to think of it, Linux definitely is more robust when it comes to server applications. And the Linux kernel is definitely more robust, but there are other things that can ruin your day:
  1. Your application crashes (neither KDE nor Gnome has reached anything close to the stability I expect under Linux, in fact I have been accustomed to a lot more under Windows as well. Nutscrape is also fun...).
  2. Your window-manager screws up (This happens mostly because some other application does something wrong. And then something weird happens, e.g you never get focus. Very annoying).
  3. XFree crashes (ever tried to play some 3d games under XFree. Maybe you were lucky. I wasn't. I've even managed to crash XFree just by switching between virtual consoles).
  4. Not having the apps you need to be productive. I still know a lot of people who would miss MS Word no matter what Linux afficionados tell me. Not to speak of all the other apps your company has "standardized" on, or some of the funny games out there.

In conclusion, I have no problem understanding why people don't run Linux instead of Windows. The amount of bloat, early beta-code, random program crashes and so on actually makes the desktop user-experience much worse than Windows. And although Windows has a bad reputation when it comes to stability, it has improved a lot the last years, and the user experience when used as a desktop OS is actually much better than that of GNU/Linux/BSD/Gnome/KDE.

I use linux because that is where I find my killer apps. But with Windows also being in progress (Win2k especially looking good), and the progress of CygWin (they even have XFree running under Windows) there is less and less reason for anyone accustomed to Windows to switch. Unless Linux can improve more, of course...

In my experience, yes. (3.50 / 4) (#22)
by phinance on Tue Jan 02, 2001 at 01:10:24 PM EST

I use Linux every day at work and at home. Each machine only comes down for an occasional kernel or hardware upgrade or a building-wide power outage. I'm not comparing 30-50 day Linux uptimes with 10-day Windows uptimes here. I really comparing to Windows uptimes of less that one day.

A guy down the hall from me -- who won't switch to Linux because he likes Excel -- has his machine crash most every day. Each time the Windows IT guy comes by and replaces a piece of hardware, certain that this time the machine will be stable.

My Mother takes a class on line and complains that her machine often locks or crashes -- or even shuts itself down without being asked to (saying, "It is now safe to..." ) -- while she's submitting her homework or discussion comments.

Stories like this abound.

This doesn't happen to me on my Linux systems. The only thing that crashes on me is the latest Mozilla -- but I was well aware of that risk when I installed it ;) (Besides, I'll use Navigator 4.7 for more "important" web functions.) KWin, Konqueror, kicker, konsole, and Corel WordPefect, etc., don't crash. In all honesty, kmail does occasionally crash -- less often then once a week -- but nothing, not even a the message I might have been in the middle of composing -- gets lost or corrupted.
Read, annotate, and discuss open source documentation.
Andamooka: Open support for open content.
[ Parent ]

Re: Is Linux really more robust? (3.50 / 2) (#24)
by dlc on Tue Jan 02, 2001 at 01:20:35 PM EST

Is Linux really more robust?

As always, it depends on what you do with it. Running buggy software on a solid platform is naturally going to provide a frustrating experience, overall. But the real test is how much of the machine crashes. X is notoriously unsolid, Navigator is, at best, a necessary evil (thank the gods is can be (almost) reliably replaced now), and the various and sundry desktop/window manager combinations are often sketchy, but you can always Ctl-Alt-F1 and killall xinit to fix it. That is definitely not a Linux problem.

I use linux because that is where I find my killer apps.

Bravo! Rate: 5!

(darren)
[ Parent ]

Linux is not an end user system (3.66 / 15) (#15)
by unstable on Tue Jan 02, 2001 at 09:20:12 AM EST

I dont want to make linux easier to use.
I dont want it to be a bundle of pretty graphics.
I want my command line.
I want my /etc/files
I want my MTV
I want my linux to be..... linux

let the end users have their windows. Those that demand more from a computer will use *nixes, BeOS, etc.

Let the users play in word and excel, while the DB that they pull from is a linux box running mysql or something similar.
Let the users surf the net useing Explorer, while the firewall protecting them is a BSD box locked down tighter that any NT box can dream of.

I see linux not as a windows 9x killer, but as a win NT killer... I want it to take over the servers and firewalls while the users sit in there cubes and are oblivious to the "revolution" that is happening on the "back end"

The day linux becomes as easy to use as windows... is the day I find another OS to learn.

and as for the stability problems of X..well maybe its time for an overhaul... everything I have heard of X programing is bad, "its too crufty, its over complicated" etc etc.. maybe instead of "polishing" what we have, we should start new and carve a new X windows system... make it smaller less complex and cleaner... dont make it do "everything (tm)" just make a base that the windows enviroments can build off of and work faster and better.

but thats just one geeks opinion.





Reverend Unstable
all praise the almighty Bob
and be filled with slack

Curious.... (4.00 / 2) (#28)
by Malk-a-mite on Tue Jan 02, 2001 at 03:52:37 PM EST

"The day linux becomes as easy to use as windows... is the day I find another OS to learn."

Why?

I won't begin to guess as to the reasons, trying to pigeon hole you into a stereotype. Instead I ask you to explain why making the user interface easier would cause you to use another OS.

Also, would you use another OS even if linux was still the most secure, most powerful? Would the ease of use drive you away from something that worked? If so, please explain in more detail.

Thank you,
Malk-a-mite

[ Parent ]
easy of use (none / 0) (#46)
by gromgull on Fri Jan 05, 2001 at 11:52:03 AM EST

Because as easy of use goes up, the possibilities and the power goes down.

Every time you introduce that new fantastic feature, there will be things your new intelligent system cant handle.

(struggling to come up with a good example from the world of windows, but failing, and thus completely "voiding" my argument)
--
If I had my way I'd have all of you shot

[ Parent ]
Linux IS ready for the desktop (4.00 / 1) (#36)
by pwhysall on Wed Jan 03, 2001 at 03:47:37 AM EST

Just not the home desktop.

In a corporate environment, where there are skilled (or hey, even not so skilled) techadmins on hand, where there's an infrastructure in place for things like standard desktops, Linux is a reasonable choice of desktop, if it runs the apps you need to run.

A little example. One of the developers here was installing Helix GNOME and while the installation went fine, and gdm ran up nicely, he didn't have GNOME in his list of sessions.

Much headscratching on his part later, he called me. I went over and had a look in the sessions directory, and the GNOME session file wasn't executable. chmod 755 gnome and he was away.

The point I'm trying to illustrate here is that some problems that are very serious for an unsupported home user become virtual non-issues in a corporate environment.

It's not sensible to say "Linux isn't ready for the desktop" without at least some qualification.
--
Peter
K5 Editors
I'm going to wager that the story keeps getting dumped because it is a steaming pile of badly formatted fool-meme.
CheeseBurgerBrown
[ Parent ]

Have your cake and eat it too... (none / 0) (#45)
by Brandybuck on Wed Jan 03, 2001 at 03:33:48 PM EST

  • I want my command line.
  • I want my /etc/files

As far as I know, no one is planning to take them away from you. The value of the command line has been proven, so much that even that Mac is going to have one. It won't go away. /etc is going to be around for quite a while.

So don't worry about it. No one is going to make you use a GUI admin tool. You can always drop into /etc and edit to your heart's content. Hell, you can do all of your computing off of a daisywheel terminal if you want. But don't deny to GUI to those that want one. If you don't want KDE or GNOME, then don't install them! But if you even think you're going to touch my desktop you have another thing coming...

[ Parent ]

Why a Linux desktop? (3.71 / 7) (#18)
by djkimmel on Tue Jan 02, 2001 at 10:43:06 AM EST

Why not another Unix-like system?

In particular, what is wrong with the BSDs when it comes to running a Unix-like system on your desktop?

FreeBSD would be good for older systems since it is very fast, faster than Linux on the same hardware, but not good for a network of hetrogeneous systems since it currently only supports two or three platforms.

OpenBSD is a fair bit slower, but much more secure, than FreeBSD, and it supports more platforms. This would be the choice if security is of utmost importance.

NetBSD has the advantage of supporting more platforms than any other OS that I know of. I've never used it, so I can't comment further.

I'm kind of tired of everyone talking about Linux this and Linux that when what they're saying can, and in some cases should, apply to the whole Unix world, not just a small subset.
-- Dave
Because it's pointless. (5.00 / 1) (#35)
by cameldrv on Tue Jan 02, 2001 at 08:19:20 PM EST

FreeBSD may very well be slightly better than Linux for some things. However, Linux is much more widely used, and precompilied binaries are much more available and it is generally much easier to use because of its popularity. Furthermore the similarities between the two otherwise are too small to be of much consequence except to people who just have to be different because they think they're cool.

[ Parent ]
Not pointless (none / 0) (#43)
by djkimmel on Wed Jan 03, 2001 at 12:31:28 PM EST

There are many similarities between FreeBSD and Linux, but the differences are certainly not pointless.

As an example, I find that IP Filter is a better firewall for my needs than IP Chains/IP Tables/ipfwadm. IP Filter is available for the BSDs, but not for Linux. This is a fairly trivial case as the Linux firewalling system can accomplish the same thing, but serves to illustrate my point.

I only bring up the BSDs because the two desktop environments mentioned in the article, GNOME and KDE, run just as well on the BSDs as they do on Linux. It is unfair to imply otherwise.

The similarities between Linux and the other Unixes are many, but the differences are non-trivial.
-- Dave
[ Parent ]
"just keep polishing" (3.40 / 5) (#19)
by Michael Leuchtenburg on Tue Jan 02, 2001 at 10:57:40 AM EST

Perfection is a bad goal. I've noticed that when my goal is perfection, I won't take perfectly good opportunities to make it better, and will instead wait for the "perfect" one. Thus, I end up with an inferior product to what I could have had, because I rarely find that "perfect" solution. "Better" is a much more effective goal at producing a good product.

So yes, just keep polishing. Or rather, keep on looking out for something to make it better, don't just make what's already there less buggy, etc. Take a look at each possibility, compare it to the current situation, and if it's better, go for it. If it turns out not to work out, go back. The only loss is some time.

Obviously, if you have more than two choices at once, you should compare all of them. You may have better luck comparing them in pairs, ie A <=> B, B <=> C, C <=> A. Then look at your results and pick your path. But don't just wait for the "perfect" solution. You'll never find it.

[ #k5: dyfrgi ]
[ TINK5C ]

Why I chose a Linux desktop (3.00 / 4) (#23)
by dlc on Tue Jan 02, 2001 at 01:13:01 PM EST

To be honest, I have a problem with the Linux-on-the-desktop vs. Windows discussion. It's a non-issue. Linux, and Free Software in general, are not going away; they definitely do not need "market share" or anything so vague to maintain an existance. The acceptance of Linux on the desktop is of little real importance, in the big scheme of things.

Rather than make Linux more user-friendly, why not make the user more Linux friendly? I don't think it is at all unfair or presumptuous to assume that someone using a tool as complex as a computer should be required to understand how it works, and why. A little education will go a long way, and I dislike the current trend (i.e., the last 10 years of "popular" computing) of dumbing down the interface and making the computer accessible to the average, uninformed (as far as computers go) masses. Why should a piece of software need to anticipate the needs and wants of a sentient and, presumably, intelligent being? Software should react, not act. The minute a piece of software starts rearranging itself, or offering helpful suggestions, I stop using it.

Yes. Linux is still catching up -- on the desktop. As I have already pointed out, I feel that the desktop is incidental to the real value of Linux. As a development platform, Linux has almost no rivals -- as attested to by the sheer number of programming tools and server software that not only run on Linux, but are designed specifically for it. Aside from the BSD's, there is no other hardware /software combination that comes even close to the bang-for-the-buck-ness (yes, I just made that up) as GNU/Linux on x86. That is why I choose a Linux desktop.

I want to be in control of my software, not the other way around. That is why I use Free Software--because I control it, not the other way around.

P.S. I don't mean this as an attack against the article or the author, or the author's point of view; please don't take it as such. The article was well-written. I disagree with most "Linux on the desktop" evangelism.


(darren)

Re: Why I chose a Linux desktop (3.50 / 2) (#30)
by jfpoole on Tue Jan 02, 2001 at 04:34:55 PM EST

Rather than make Linux more user-friendly, why not make the user more Linux friendly? I don't think it is at all unfair or presumptuous to assume that someone using a tool as complex as a computer should be required to understand how it works, and why. A little education will go a long way, and I dislike the current trend (i.e., the last 10 years of "popular" computing) of dumbing down the interface and making the computer accessible to the average, uninformed (as far as computers go) masses.

Why should people need to sit down and learn how to use a computer? Ideally, an inexperienced user should be able to sit down in front of a computer and be able to use it without resorting to documentation or another computer user. I can't think of any other tool that exists (which is how most people view computers) that requires so much education just to use, let alone use successfully. As for the specific proposition of making users "more Linux friendly", keep in mind that for the average user, sitting in front of a Windows machine isn't much fun -- I can't imagine Linux would be any better (if not much worse)!

Sidenote: For an interesting examination of current computer user interfaces, take a look at The Inmates are Running the Asylum by Alan Cooper.

-j

[ Parent ]

Why Not? (4.50 / 2) (#33)
by Matrix on Tue Jan 02, 2001 at 07:35:51 PM EST

You have to learn to do everything else in your life. Prepare food, drive a car, read, and write. Why should software have to be usable without any learning curve at all, and try to guess what the user wants it to do? Yes, its behavior should be consistant and friendly (comfortable to use - I doubt many would drive cars where they had to do EVERYTHING with their hands), but users shouldn't expect it to do everything for them. I often finds this gets in the way of what I want - for example, I actually prefer browsers without autocomplete, as I generally know what the URL I want is, and the browser's guessing gets in the way.


Matrix
"...Pulling together is the aim of despotism and tyranny. Free men pull in all kinds of directions. It's the only way to make progress."
- Lord Vetinari, pg 312 of the Truth, a Discworld novel by Terry Pratchett
[ Parent ]

Not true (2.50 / 2) (#38)
by evvk on Wed Jan 03, 2001 at 04:00:02 AM EST

> I can't think of any other tool that exists (which is how most people view computers) that requires so much education just to use, let alone use successfully.

For many people, even some VCR's and other similar equipments with lots of features are too hard to use. (DVDs are horrible with all the menus...)

[ Parent ]
Why should people need to sit down and learn... (5.00 / 4) (#39)
by dlc on Wed Jan 03, 2001 at 06:29:32 AM EST

Why should people need to sit down and learn how to use a computer?

You are re-phrasing my original question: Why shouldn't they?

I can't think of any other tool that exists...that requires so much education just to use

That is exactly my point; that is why people should educate themselves about how computers work. As more and more of our lives become dependant upon computers (banking, air traffic control, heck, everything), why wouldn't you want to learn how they work? I would even go so far as to say that relying on something as powerful as a computer without understanding it at all is irresponsible.

Treating a computer like a tool makes this even less excusable, at least for people who use computers in their jobs. If a construction worker came to a job without knowing how to use a hammer, he would be mocked and fired, yet people regularly use computers without having the foggiest idea how they work or why. Of course a computer is more complex than a hammer, but training a new user in basic computer concepts--enough for them to get over the basic ignorance about how they work in general--takes a week, tops.

sitting in front of a...machine isn't much fun

No, but then again, many things in life are not fun. Did you really enjoy high school? Can you honestly tell me that you are not a better person for having been educated, whether you liked high school or not? Learning to understand and use a tool does not require that the tool become your passion (I guess at that point, it's no longer just a tool) or that you enjoy every minute of it. Whether it's "fun" or not is beside the point.

Ideally, an inexperienced user should be able to sit down in front of a computer and be able to use it without resorting to documentation or another computer user.

I've heard this argument thousands of times, but I have yet to hear a good, sound reason for why this is (allegedly) the case. Use of a computer is not a natural right! If you can't be bothered to take the time to learn how to you use it, why are you using it? Oh, it's required for you to do your job? Then your employer should provide training. See above.

Then again, I'm the guy who, whenever possible, reads the code for new software before I build it. Never trust code you haven't compiled yourself; never trust a something you don't understand. jfpoole - sorry for mixing up the order of your post in my reply...


(darren)
[ Parent ]

The Inexperienced (5.00 / 3) (#40)
by pete stevens on Wed Jan 03, 2001 at 07:09:59 AM EST

"
Ideally, an inexperienced user should be able to sit down in front of a computer and be able to use it without resorting to documentation or another computer user.
"

Ideally, an inexperienced pianist should be able to sit down in front of a piano and be able to use it without resorting to a book or another experienced pianist.


I bought a computer - why do I need lessons on how to use it?

I bought a piano - why do I need lessons on how to use it?

.... the Flat Earth Society announced in 1995 that their membership was global
[ Parent ]
Simple answer (none / 0) (#49)
by dennis on Sun Jan 14, 2001 at 09:14:02 AM EST

People also use cars without knowing what happens under the hood. Doesn't make them less effective drivers, though it does tend to cost them more when something breaks down.

Learning is best, but some things are more worth learning than others. Some software is hard to learn just because it's poorly designed. Ideally, software should be easy to learn, without getting in the way of experts. Even better, the gui should help you learn.

Eg.: it's nice having text configuration files that you can edit. It wouldn't be so bad though to have a gui app that handles the most common tasks. It would be real nice if you could hit a button in the gui app and have it point out what changes it made.

It's easy for a programmer to say that users should learn about the computer, but nonprogrammers tend to have plenty of work to do, and their time and training costs money. That's the answer to "why shouldn't they." The computer is a much more effective tool if it's not too hard to learn.

[ Parent ]

Freedom vs flexibility (3.50 / 6) (#26)
by Demona on Tue Jan 02, 2001 at 03:15:32 PM EST

No matter what standards are established by fiat, they'll only be adopted if they make sense to those who've followed long-established custom...

From a programmer's point of view, a standard API would certainly do wonders in terms of encouraging development of a multitude of desktop interfaces. If this is what's meant by a "standard desktop environment", I've no objection, but this should never imply a single GUI and user-level interface. A consistent set of interfaces is important for both users and programmers, but the Apple route (or insert your favorite fascist hardware/software) is not a model worthy of our emulation.

Want to use raw garlic and code in straight Xlib instead of using a toolkit? Go ahead! If it's good enough, maybe we'll get an entirely new API out of it; if not, it'll fade as you tire of supporting it. Either way, you had the freedom to make it, and the community had the freedom to use it. If I were a programmer coming into the world of free software, I'd be ecstatic at the unlimited possibilities, rather than bemoaning the plethora of choices like an 80's Soviet immigrant in a North American supermarket.

I would no more insist that every user have the same interface than I would every person wear the same shoe size. Down that road lies the sort of hell we glimpsed when the web stumbled; a one-size Procrustean nightmare that says My way, or the Highway. And down at the bottom of the blinking marquee is the dreaded word

SUBMIT

To which I reply, "Never!"

OS Ambiguous (3.00 / 3) (#32)
by mattyb77 on Tue Jan 02, 2001 at 06:21:23 PM EST

Is the idea to have a better GUI the result of a desire to make Linux more used and acceptable? Is that really what Linux needs in order to become a "better" operating system than we we have now? Can Linux be everything to everyone?

--
"I bestow upon myself the `Doctorate of Cubicism', for educators are ignorant of Nature's Harmonic Time Cube Principle and cannot bestow the prestigious honor of wisdom upon the wisest human ever." -- Gene Ray, the wisest human ever
Linux on a desktop: I use Unix on a desktop... (3.00 / 1) (#42)
by bscanl on Wed Jan 03, 2001 at 09:51:46 AM EST

Personally, I've never found much difference between Solaris/FreeBSD on a desktop compared with Linux. Same software works the same, aside from badly written Linux dependant stuff, of which there is a suprising amount of. Some free software coders just can't write good configure scripts...

I think people are underestimating StarOffice. StarOffice 5.2 was written pre-Sun and Open Source hitting it. StarOffice 6 is going to kick ass.

I hate that (none / 0) (#44)
by djkimmel on Wed Jan 03, 2001 at 01:02:11 PM EST

I'm always surprised when certain programs which strike me as fairly generic can't compile and run on anything but Linux.

Of course, my faith is restored when I see a program like cdrecord, which has to interact with the hardware on a fairly low level, that compiles and runs on many different Unix flavors.
-- Dave
[ Parent ]
Why choose a Linux Desktop? | 49 comments (47 topical, 2 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!