ivwshane
Lifer
- May 15, 2000
- 33,557
- 17,082
- 136
Who is going to define computer basics? And at what point will knowing the basics mean they are good enough to use/understand linux?
If all or a majority of computer users had a basic understanding (basic as in what you and I might define as basic) then we wouldn't be having this discussion. People would either already be using linux or the issues people complain about on windows would no longer be an issue.
The fact remains and will most certainly remain the same is that a majority of people are computer stupid and in order to compete with windows you have to offer something that not only caters to the computer stupid but also gives them a meaningful reason to switch.
The world isn't perfect and if you want linux to be a successful desktop OS it has to cater to an imperfect world.
Great discussion guys. I agree, a distribution of Linux will have to make it cater to the end user before it will ever take off in the desktop segment. Lindows tried to do this, but IMO didn't succeed and made it too like Windows. Nevertheless, it's getting close, very close. Novell has done great things with [Open]SUSE Linux. What I think needs to be done. Note this is for your grandma's OS, not an enthusiant or geek's OS. The latter will definitely want more choice and the following needn't apply to them. They will probably cringe just at the thought of it ("Delinuxization").
Those would be major steps into making it more user-friendly. I don't think any of that would be that hard.
- Don't list 'kedit' in the menu bar, list 'text editor'. Don't list 'xterm', list 'terminal'. Etcetera...
- Remove redundant programs.
- Hide the virtual desktops feature by default.
- Add the purpose of each program after its moniker if it's not obvious. For example, 'FireFox Internet Browser'. 'Evolution E-Mail'. 'OpenOffice' is fine as is.
- Make 'root' 'Administrator' instead.
- Put error messages in dialog boxes, not hidden somewhere in the command line. Joe isn't going to be using dmesg too much.
- Make an easy-to-use device manager like Windows. Allow driver updates via the GUI. This should be a rock-solid, bug-free, stable front-end to modprobe, insmod, and /etc/modules.conf.
- (The updating system is already great in SUSE Linux. They have done very well on that. Just make some of the descriptions more 'English' and rate them in importance (critical, etc.))
- Have one place where you can configure your main programs. I don't mean like a package manager that lists all 5000 that you have installed. I mean like, for example, list OpenOffice, Evolution, FireFox, gaim, and Ark. Don't list the kernel source code, cups, grep, or vim! There should be qualifiers used for each package installed. Vim should be labeled small, and OpenOffice labeled big. And I don't know why cups would even qualify as an application as it's more a printer driver.
- (Power users will hate me for this one) Add temporary balloon tips saying like: Where are my programs? like Windows does.
- Make it so Joe never ever sees the big Linux consoles unless upon request. He will have no idea what to do in there. Always have XWindows running, even if it has to under a default VGA driver.
Those would be major steps into making it more user-friendly. I don't think any of that would be that hard.
ivwshane
Lifer
- May 15, 2000
- 33,557
- 17,082
- 136
Definitely!! Good post!
Those are just little things that make linux more accessible and I don't see how those suggestions would be hard to add. I see nothing wrong with dumbing down an OS and having it's more advanced features hidden or harder to find for your average joe (I think windows actually does a good job of this).
However I still believe that linux will need some kind of singular identity (at least from the average joes stand point) and a default interface that they can look at and say, "that's linux!".
Once the accessibility issues are taken care of it's just a matter of distrobution and finding away to get joe smoe to think about and consider linux.
Those are just little things that make linux more accessible and I don't see how those suggestions would be hard to add. I see nothing wrong with dumbing down an OS and having it's more advanced features hidden or harder to find for your average joe (I think windows actually does a good job of this).
However I still believe that linux will need some kind of singular identity (at least from the average joes stand point) and a default interface that they can look at and say, "that's linux!".
Once the accessibility issues are taken care of it's just a matter of distrobution and finding away to get joe smoe to think about and consider linux.
Nothinman
Elite Member
- Sep 14, 2001
- 30,672
- 0
- 0
What overall project? There is none.
I believe KDE does this already, although in the form of "Text Editor (Kate)".
Distributions like Ubuntu already do this, but inevitably someone complains about their favorite apps not being bundled and you get crap like Automatix.
In Gnome, the only indication that it's enabled is potentially an icon on the taskbar and I'm not sure whether it's there or not by default. But even if it is, there's nothing forcing you to use it and you can remove it just by right-clicking on it.
Again, I think Ubuntu does this already, but I don't have a box handy to verify that.
Irrelevant, they never need to see the account name. In Ubuntu all admin tasks are done with sudo or the graphical password dialogs that ask for the user's password.
dmesg is only for kernel logs. As for each app displaying it's errors properly, that's up to the application and Windows apps are usually much worse about telling you what's wrong.
Synaptic will update any drivers that you have installed via packages already, if you had to install them manually you'll have to update them manually and if you don't like it you can complain to the driver authors and ask why they're not in the main kernel yet. As for loading, unloading modules and managing modules.conf I don't doubt that HAL will be doing that in future, but 99% of the time it's not necessary since the kernel will hotplug any module when the device is inserted.
Now that Novell owns them I doubt English translations will be a problem.
That's pretty much impossible, even if someone did write some filters to hide the 'unimportant' stuff they would get out of date very quickly and I doubt they would be maintained well. And CUPS is an application, it's a daemon that accepts print jobs, runs them through any necessary filters and then pushes them to the printer, the drivers are only a small part of it.
KDE has that tip of the day thing that pops up when you first login. And considering the Gnome menus say "Applications" I doubt a balloon pointing to it would do much to help the person that can't find it. And I doubt power users will care as long as there's a check box to disable it, just like there is on the KDE tip dialog.
Fedora and Ubuntu have been doing that since their inception. Infact I can never find the gnome-terminal icon in Ubuntu.
That will never happen, Linux is essentially a generic term like 'car'. If anything RedHat, Ubuntu, SuSe, etc they have a chance to get some real name recognition and maybe Gnome or KDE. But Linux is not a brand name and probably never will.
Well considering that 90% of the above suggestions are already done, I guess someone should just start working on advertising then, eh?
ivwshane
Lifer
- May 15, 2000
- 33,557
- 17,082
- 136
90% of the above may be done but it's not all done in one package. How many distro's did you have to refer to in the above post?
That's not good enough and the average joe doesn't care whos fault it is. Linux just has to work for joe smoe, if it doesn't he wont be giving it a second chance.
You are missing the point. You are taking what I am saying way too literally. It doesn't matter what the brand name is, it just matters that there is a brand name and when joe smoe hears it he can associate it with one thing. And if that still will never happen then linux will never be a successful desktop OS.
That's not good enough and the average joe doesn't care whos fault it is. Linux just has to work for joe smoe, if it doesn't he wont be giving it a second chance.
You are missing the point. You are taking what I am saying way too literally. It doesn't matter what the brand name is, it just matters that there is a brand name and when joe smoe hears it he can associate it with one thing. And if that still will never happen then linux will never be a successful desktop OS.
Nothinman
Elite Member
- Sep 14, 2001
- 30,672
- 0
- 0
I referred to 3, but I believe Ubuntu already does just about all of the above points, I just don't have a box to verify that and I mentioned Fedora and SuSe for completeness. I did mention a point or two that were done in KDE that aren't done in Gnome, so that means they won't be in Ubuntu by default.
Why does it have to work in Linux when it doesn't work in Windows? How many drivers on your machine are handles by WindowsUpdate?
No, you're missing the point. Linux is already 'good enough'. All that needs to be done is some OEM support so that boxes can be sold with Linux preinstalled and a few companies like Macromedia/Adobe releasing some of their software to get the ball rolling. It's a catch-22, they don't want to release anything because they don't see a market and they don't see a market because they haven't released anything.
ivwshane
Lifer
- May 15, 2000
- 33,557
- 17,082
- 136
Because people are already using windows and it works, if linux isn't going to work better then there is no reason for people to switch. Linux doesn't have the luxury of not working like windows does because people already have windows.
Your argument is that linux wont get the market share it should because oems and software venders don't offer or support it enough, so what! That's not going to change anytime soon. Amd hasn't been slowly gaining market share because their products are "good enough" they have been getting market share because their products are better. And as their product has gotten better the support and adoption rate by oems has grown. And if they want to continue growing and caputuring market share their products have to continually be better than intel and they have to better by a tangible amount.
Nothinman
Elite Member
- Sep 14, 2001
- 30,672
- 0
- 0
People already have no reason to switch, otherwise they would have done it already. Most of the advantages are intangible and hard to enumerate, especially for people that aren't computer enthusiasts.
And IMO it already does work better. On the 4 machines that I run Linux I only have 1 non-OSS driver, so virtually all of my drivers are updated automagically when I install a new updated kernel.
Soon is a very vague term. I'm not saying everything will be wonderful 30 days from now, but there is hope. For instance Adobe has been watching Linux very closely for years and they already release Acrobad Reader for Linux. And they even ran a trial of FrameMaker for Linux, but it bombed because they didn't put any effort into it, it was just a port of their crap unix version. If they would have treated it like acroread and ported it to GTK2 and made it look half decent it might have fared better.
AMD's position right now isn't eevn that great and it's taken AMD like 15 years to get here, back when they first started producing 486 clones and the K5 people were scared to use them because they had some really bad problems. Even now, when AMD64 is clearly better than IA64 and the current Xeons AMD is still struggling to get people to look at them. Dell won't even consider them and HP only sells a small handful of Opteron systems. AMD is only really prominent in the gamer part of the market and in the big picture that part of the market is pretty small.
It is a similar situation, the people who know the differences and know that AMD is producing better hardware support the company by buying their hardware and telling others about them. But most people don't care what CPU or OS is being used, they just go to the dealer and buy the cheapest box that will do the job and usually that box is Wintel.
The problem, and this is a positive for AMD, is that it's very easy to replace an Intel CPU with an AMD CPU because they use the same instruction set. Everything you have runs no matter what CPU you've got, the same can't be said for a Windows->Linux transition. You not only have to switch OSes but you have to change out virtually every application that you're used to using and that's not a quick, cheap or easy transition.
gsellis
Diamond Member
- Dec 4, 2003
- 6,061
- 0
- 0
Catch-22 is right (went through this with Team OS/2). The apps that would make people switch are from folks like Macromedia and Adobe. But, to spin up a dev team to port the Apple/Wintel code to Linux costs money with limited sales potential. So, the Linux version would be more expensive for them to make. And folks would not want it to be more expensive to buy than the Wintel/Mac versions. And worse, a bunch of the Linux camp because it is 'free'. They have an adversion to paying premium dollar for software (those supporting it corporately don't count (N0cmonkey for example?). And a company would essentially have to sell it at a loss. Boards of directors do not let that happen for long. sigh. We need competition, but competition has to be profitable before anyone will jump in.
Well there is some momentum in this direction just starting up in the past 18 months or so.
First off you have things like WineLib. WineLib allows developers to take Win32 applications and port them to 'native' Linux applications as inexpensively as possible. Often it's nothing more then a recompile..
As far as maturity of the product goes.. It's entirely possible for you do program, test, and compile a Win32 application only using Linux software and Wine-related items and end up with a exe file that runs directly on Windows. So if that directory (from Linux to Windows) is possible I figure it's entirely possible to get Windows to Linux with minimal hassle for most people. Often it's nothing much more then a recompile... If even that.
Google itself has these 'Google Earth' items and such.. They plan on teaming up with CodeWeaver (commercial wine entity) to make applications for Linux.
With OS X to Linux you also have something along the same lines called GNUStep. OS X Cocoa is based on the Openstep API which developed from the NextStep Unix operating system. GNUStep is a free software implimentation of Openstep. Compatability between both items is fairly high.
Then you have Mono with C#. It's developed in such a way that most inhouse programing for businesses require little effort to go from MS's .NET to Linux's Mono. It's designed so that a programmer with almost no Linux experiance can churn out a effective and complex desktop and enterprise applications for Linux.
This will work for 90% of everything out there.. But the valuable items like Photoshop would take more work. There are probably all sorts of optimizations going on and since it's a very complex product it probably uses all sorts of undocumented items.
It's worth noting that Adobe has membership in OSDL (the group that employs Linus among other people) was/is looking to hire for Linux developers and has had presentations on porting applications to Linux. (Like Adobe reader and such)
Besides the limited Linux desktop market there are other barries for 'ISV's (independant software vendors) in Linux.
For instance for servers and such there are standards and certifications for applications. You can usually expect a peice of software made for a server 5 years ago to work fairly well on Linux today. Binary APIs with the Linux kernel and other core items are VERY stable, possibly more stable then Windows. Applications built in the 2.0 days (or even earlier) will more then likely work on 2.6 kernels for instance.
However for the Desktop there isn't realy something like that. Right now it's a very fast moving target and is rapidly evolving and developing. When a person developes a application for Linux what can they use? What parts of the desktop can they depend on not to change? What toolkits, what programming languages?
So normally you can make a intellegent _guess_ for what to aim for.. For instance Gnome 2 series has had a very stable API and make severe pains to avoid backward compatability. For KDE QT 4 will break API with 3 stuff, so maybe it would be smarter to choose Gnome... But not all Linux distros may not have Gnome.
So there needs to be standards. Something that can be printed on the box and labled 'Redcommended Linux requirements'.
So that needs work and people are working on them. They are figuring out a programming model for Linux you build something now it can be depended on to work 5 or more years from now. It's maturing...
As for the rest of the 'Catch-22' part.. The market share items.
This is a huge problem. No applications, no customers.. no customers, little market share.
The goal to break out of this cycle is to target businesses. This is to emulate Microsoft's rise.
Businesses have technical staff and as Linux becomes easier and more widespread... Now most corporations everywere are using Linux for one thing or another. It's going to very rare to find situations in medium to large businesses were you don't have people that know Linux and are using Linux in real-world situations. Weither it's for 'infrustructure' stuff like email, DNS, file shares etc etc, or more enterprise type stuff with databases and such. (for small businesses not nearly so much)
So the idea is to take that and sell people on the superiority of Linux and Linux-based solutions for business desktop. Improved security, improved stability, reduced cost of deployments, reduced costs of software and software developement, reduced costs of maintanance (hardware and software) etc etc. That sort of thing.
Once you sell the businesses on Linux desktops then regular people get familar with it at work.
They want to do work at home. They want to use the same applications and same enviroment at the work place to get work done at home.
That's were you have things like VMWare's Player and such. If you use Linux at work and have people that create images for desktop deployments it's just a short hop skip and a jump away to give somebody a vmware player application so that they can run Linux on their desktops at home, even though it's in a virtual enviroment. It'll 'just work'. No problems with hardware, no problems with installing or giving tips on how to install Linux or use the command line to install drivers or whatnot...
When you get your laptops from work for doing business at home or while traveling they will be running Linux.
So 'Joe Blow' gets the desire to use Linux and familarity from what they use at work. It's just right now the job of people like Novell to sell the Linux desktop as a business solution.. then after that matures then they will be able to sell linux as a home desktop solution, eventually.
Joe Blow isn't going to want to run Linux in a VM forever. Especially when that is what they use all the time. He'll want Linux on his hardware, but he isn't going to want to install it and support it himself.
He'll just pay OEMs to deal with that sort of thing. OEMs are going to carefully select hardware that 'just works' with Linux. The OEMs aren't going to want to have different hardware configurations for Windows and Linux.. When it's easy to find hardware from vendors that 'just work' in both operating systems why would you want to buy hardware that only works in one or the other?? So eventually it will come to the point were if hardware vendors want their stuff to be sold AT ALL, it will have to work with Linux.. even if Linux is only 10% of all the computers.
So right now...
Step 1: Linux in the server room and enterprise and other enviroments.. (done)
Step 2: Linux to become a substantial minority OS in the business desktop arena... (working on it)
Step 3: Linux to gain a foothold in the home desktop arena..
Step 4: World Domination. (er.. profit or something)
On the flip side of things technology is changing.
Mobile computing is now as powerfull as PCs were a few years ago. Your mobile phone will probably play Quake faster then a Pentium 300 MMX machine can by the middle of next year. Linux in embedded computing is fairly popular and is going to get more popular as embedded computers stop being so 'embedded'. Most people will end up using Linux weither or not they know it. Your house has probably a LOT more computers then you think it does. Computers for modern stoves, your cofee maker, your TiVO, your whatever. The traffic lights, your car, your car's ABS system. Traffic control, Security items, Air conditioning controls. All sorts of stuff.
The future of mobile computing is a teter-totter between Linux and Windows stuff.
It maybe that in 2-5 years time you'll end up doing the vast majority of your computing stuff away from your notebook or desktop.
Web based items are also exploding. Applications that are not performance critical (the vast majority of them) can more easily be deployed as web based items now more then ever before. Things like XULRunner (from Mozilla.) and whatnot will allow this. (Realise that applications like Thunderbird and Firefox render _themselves_. It uses the same engine as what they use to render web pages...) Look at Flock Browser as a for instance. http://www.flock.com/ If your a blogger this thing is pretty much were 'it's at' in terms of usability. This is also another browser based off of the Mozilla stuff like Firefox is. The interface is programmed almost exclusively in XML and Javascript.
And for something that is completely online... EyeOS: http://www.eyeos.org/index.php?section=Whatis It's kinda silly, but shows how actually complex items online can be right now. Obviously very beta. Shows some of what is possible and what some of the limits are.
After you realise that Firefox is rendering itself.. Realise also that the Cairo Library that is bringing vector based graphics to Linux desktops is cross-platform. That is eventually Firefox will have vector based graphics and full SVG support as well as have optional OpenGL accelerated (and in Windows they will possibly make a DirectX backend for it.. it's very modular) rendering.
If it's webbased (and not using ActiveX) then it's irrelevent what OS your using. Linux or Windows; what does it matter? You just choose what you want.
These sorts of things open source and Free software are at the for-front of things. The truth is is that as your desktop becomes more and more irrelevent Linux will possibly gain a substantial advantage over Microsoft's stuff. Microsoft desperately needs you to KNOW that your using their OS, that only certain things can be done using their OS, and that you need to pay to use it and you should be happy to do so given all the functionality that you gain. They're whole business model is based around Windows. As more and more items are desktop independant then Linux will become more and more usefull and it's inexpensive-ness with shine more and more as time goes on. I think that it's possible that Microsoft will be releasing distant versions of MS Office with native Linux ports.
The future is such a unknown aspect especially with software and technology.
kamper
Diamond Member
- Mar 18, 2003
- 5,513
- 0
- 0
Exactly. Why is everyone so obsessed with the notion of linux becoming a drop in replacement for windows? The vast majority of linux users don't give a damn about displacing microsoft, some just like to make fun in stupid ways (like the op, for example).
I, for one, cringe at the thought of linux becoming a windows clone. The job probably can't be done better, it needs to be done differently. The only reason I see to specifically target windows users is maybe for this crowd that seems so intent on seperating themselves from the evil empire before vista comes out, but they need to learn to adapt. Making stupid, draconian statements like "Linux needs to have only one desktop so as not to confuse the microsoft people" is not going to do the world an ounce of good.
Bottom line is that linux will become whatever the people writing the code want it to be and that's what should happen. It's a completely chaotic and decentralized process and nobody controls it by saying that linux needs to gain market share.
That's a fine attitude, but then people shouldn't whine that Linux isn't being widely adopted on the desktop.
Changing the GUI to make it objectively 'better' in some sense is one thing. Changing it because you personally don't like it, or just for the sake of changing it, confuses people needlessly. Completely neglecting the GUI because you think that only 'power users' should use Linux (not as common an attitude anymore, but it still seems like an afterthought in a lot of Linux systems) is definitely going to hurt your chances of it becoming a widespread desktop OS.
How many years do you expect it to be before there are enough tools avaliable in Linux to successfully take the market over by providing the proper compatabilty?
kamper
Diamond Member
- Mar 18, 2003
- 5,513
- 0
- 0
That is exactly my point. People should stop whining unless they are down in the trenches contributing. I'm not saying people shouldn't try to make linux appropriate for the desktop. You don't have to be a developer to help either.Originally posted by: Matthias99
That's a fine attitude, but then people shouldn't whine that Linux isn't being widely adopted on the desktop.
ivwshane
Lifer
- May 15, 2000
- 33,557
- 17,082
- 136
If this represents the attitude of linux developers then we don't need to be having this discussion at all.
The topic, as we have discussed, is why linux isn't more common on the desktop. I as well as other brought up points as to why it isn't and how to get it there.
If you, as the linux user and developer, don't want linux to be a successful desktop OS then we should close this thread and end the discussion.
The last really dramatic change was Win3.1->Win95 (at least IMO). Win98 was nearly identical. WinNT/Win2K had some minor changes, but works pretty much the same as Win95/98. I don't have too much experience with WinME, but it looks a lot like Win98 on the surface.
WinXP tweaked a bunch of things, but with the new graphical stuff turned off is pretty damn close to Win2K. I'm not crazy about everything they did, but some people like it, and it takes all of about three mouse clicks to revert back to the old mode if you don't. That's about as convenient as you can make it while still making substantial changes.
We'll have to see what they do with the final version of Vista, but I'm guessing there will be a fallback mode that's relatively WinXP-like (if not Win2K-like).
Now Office... that I wish they wouldn't screw around with as much. At least you can make that stupid paperclip go away.
Cerb
Elite Member
- Aug 26, 2000
- 17,484
- 33
- 86
Kamper ~= everyone else.
I'm using Windows for gaming, my soundcard (it kind of works with new kernels!), and that I've got a nice tweaked desktop that I'm used to.
Linux already is a viable dekstop OS, and on anything except my main desktop, I would consider nothing else--well, I lied, I might have to try PC-BSD, but dual-booting with a nice Linux distro!
Ubuntu is the only real company out to make it truly successful. But really, that's fine. As long as the communities remain string, there is nothing to worry about.
I'm using Windows for gaming, my soundcard (it kind of works with new kernels!), and that I've got a nice tweaked desktop that I'm used to.
Linux already is a viable dekstop OS, and on anything except my main desktop, I would consider nothing else--well, I lied, I might have to try PC-BSD, but dual-booting with a nice Linux distro!
Ubuntu is the only real company out to make it truly successful. But really, that's fine. As long as the communities remain string, there is nothing to worry about.
kamper
Diamond Member
- Mar 18, 2003
- 5,513
- 0
- 0
As Nothinman pointed out already, nobody can really say what represents the attitude of linux developers (if you're lumping all the desktop oriented developers together as 'linux developers'. Their software runs on all sorts of operating systems in reality). I know some of them are targetting the desktop and I'm sure some of them constantly compare their work to windows. I disagree with them, but heck, at least they're hacking so they get to do what they want. On the other hand, there's obviously lots who dislike ms and lots who couldn't care less.
That's fine, I'm in that discussion too. I'm just countering the notion that "linux needs to this" and "linux needs to do that". People treat it as if it's a business that will die if it doesn't make money. These are lovely ideas for companies that are making a specific brand of linux, but not for linux itself (and here I'm referring to linux and all the software that runs on linux: kde, gnome....).
For the record, I'm not a linux developer and just an infrequent user. I never said I don't want it to be successful. I just said I don't want it to win by trying to be a better microsoft. À la Lord of the Rings, it'd be dethroning one evil lord to put another in his place (analogy only, I'm not (necessarily) calling microsoft evil in general
).
ivwshane
Lifer
- May 15, 2000
- 33,557
- 17,082
- 136
By successful I don't mean over taking MS either, the more OS's the better (provided the quality is still high).
I also don't believe linux needs to take on a single indentity but it does need an identity joe smoe can identify with. Whether it's all the linux developers coming together to create that identity or a single developer, it doesn't matter.
Average joes associates microsoft with one identity but MS has more than one product and caters to more than just the average joe. Someone needs to step up and create that identity for linux. I'm not saying linux needs to do this, I'm just saying that if it wants to be a successfull desktop OS ("it" as in whoever has that goal) then "it" would need to do this.
I also don't believe linux needs to take on a single indentity but it does need an identity joe smoe can identify with. Whether it's all the linux developers coming together to create that identity or a single developer, it doesn't matter.
Average joes associates microsoft with one identity but MS has more than one product and caters to more than just the average joe. Someone needs to step up and create that identity for linux. I'm not saying linux needs to do this, I'm just saying that if it wants to be a successfull desktop OS ("it" as in whoever has that goal) then "it" would need to do this.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 23K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter