“As the economic downturn has taken its toll on GNU/Linux companies, many surviving companies have decided to move away from the desktop market, and focus on the server instead. This escalated to an alarming pace after the death of Eazel, a startup that created the Nautilus file manager, and while the focus-on-server mania has calmed down lately, it is still very much alive. Yet, it seems to me, that focusing on the server will in the end cause these companies to loose not only the desktop, but the server as well.” Find out why at OfB.biz.
Editorial: The Desktop Dilemma
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Desktops have the volume, servers do not.
Most servers sit in a closet or a datacenter.
Your desktop sits… on your desk.
Desktops = mindshare.
Desktops = fun stuff.
Desktops = the long lever of Archimedes.
The desktop is certainly, the heart of the Microsoft vampire. Only by winning on the desktop can Microsoft be killed.
So if Linux does not do well on the desktop, yes, it doomed to be a also-ran down the road sometime.
#m
Agreed!
First of all in response to “So if Linux does not do well on the desktop, yes, it doomed to be a also-ran down the road sometime”
That is just not true. Linux is doing very well right now in the server space and will continue to do so in the future. Its success or failure on the desktop has nothing to do with its future on the server. If gnome and kde are gone tomorrow this would not impact the server side at all.Linux as a server OS will be around much longer than the modern desktop as we know it. I know its tempting but don’t lump linux in with BEOS and Amiga. They never had huge industry support that linux currently enojoys on the server side.
Second and this is what I posted the other day and its related to the pipedream of widespread linux use.
Without linux being bundled by large vendors like Dell, HP, IBM and without ISV’s like Intuit,Adobe etc, and yes MS porting there apps to Linux it will never register on the desktop radar.
Even if Linux becomes the most easy to use OS ever made, without apps like Quicken, Office, and of course all the big games we will never get a large market share.I don’t care how well Wine works, if there aren’t four big isles in CompUSA with shelves full of boxed software for linux there is not much chance people will use it. It doesn’t matter at all if there is a good gpl alternative. GPL means nothing to someone who just want to get their work done. When people think “free” they think “oh yea I download Elf bowling last year, that was a good freebie”. Educating users is avoiding the problem that the apps they have been using for years are not available on linux. You have about as much chance getting them to change as you do the 200 million people who still use http://www.msn.com as their start page.
Let’s also talk about hardware support. Besides Nvidia and a few random companies most companies don’t officially support desktop linux. That is a real problem. Try calling X company and telling them you use Gentoo and can’t get your scanner to work. Unless xonsumer hardware and software products start saying Linux supported on the box, don’t except people to buy them or spend time hunting around newgroups to try to figure out how to install something.
I love linux and its a great OS, but widespread linux use on the desktop is not gonna happen. The ISV’s have said, thanks but no thanks, and the big desktopp hardware OEM’s have passed as well. I’ve been using linux since 1997 and every year someone says wait till X amount of time, or X product reaches 1.0 and then we’ll see some change. Well sadly that doesn’t happen and never will.
In conclusion be happy you have a stable free OS that works well for you, but don’t hang on to the dream that it will ever be a widespread consumer OS.
To win the desktop you need:
* One standard architecture being produced by one parent company so that developers don’t have to worry about compatibility issues with various distributions.
Linux is certainly in a mess in this respect. The solution has often been to compile on RedHat and expect everyone to use that. That isn’t a very good solution, although I can’t say I can think of a better one.
* One standard widget set so that all applications have a cohesive feel, along with which comes one set of APIs that all developers use.
X has always been in a mess in this respect. The desktop is one place where the Bazaar horribly fails. That’s precisely what your desktop ends up looking like, a Bazaar where different people have completely different ideas as to how applications on your system should look, and ultimately it just ends up looking ugly.
* Pleasing aesthetics, one of the hardest goals to acheive. This is a combination of the efforts of three different groups: those who make the display server (proper font handling, antialiasing/alpha channel support), those who make the widget sets, and those who develop the applications.
By far the main place the desktop is lacking. Many of the problems stem from the X legacy. XFree has provided extensions to make up for this but applications don’t use them for a multitude of reasons. Traditional X font handling is atrocious. TrueType support is lacking. Most application developers are amateurs and don’t know much about interface design.
Configuring XFree is a nightmare, especially from the standpoint of an administrator responsible for widescale enterprise deployment of Linux workstations. XFree is still lacking simple features such as a means to choose desktop size (and a monitor resolution accordingly)
The only way Linux stands a chance on the desktop is if X is abandoned and replaced by a superior windowing system which solves all the problems I’ve mentioned.
VERY well said Bascule.
It seems odd to me that Apple had to be the first company to really go somewhere with Unix on the desktop and produce a solid desktop environment for Unix (well, I guess I mean NeXT, don’t I?).
I would have expected that SGI would have solved all of these problems Bascule mentions by now, certainly before someone like Apple decided to do it. SGI workstations are involved in all of the markets where this would seem to matter but the best they could do is IRIX Interactive Desktop (formerly Indigo Magic Desktop) which has some nice improvements over CDE (I guess) but still suffers from all of the standard Unix desktop hobgoblins (AFAIK) and will continue to do so for the forseeable future.
SGI would seem to be the only company (outside of Apple) capable of creating a new Unix desktop standard (to replace CDE) and using its clout in the standards bodies and as a market presence to make it even semi-successful, but they have absolutely no interest in doing this (for whatever reason). If SGI were run by geniuses instead of idiots, they would have bought NeXT themselves and plundered all of NeXT’s innovations smartly.
“The only way Linux stands a chance on the desktop is if X is abandoned and replaced by a superior windowing system which solves all the problems I’ve mentioned.”
The only way Linux stands a chance on the desktop is if someone comes along and decides to do with it what Apple did in creating OS X. It won’t happen. Nobody has the balls (or the MONEY) to try it. And even more unfortunately, Apple doesn’t have the balls to release OS X for Intel. I know so many people who would run it in a second.
Linux is like the Heathkit of OS’s. Actually, that’s not being fair to Heathkit. Installing, configuring and administering Linux is like picking up a copy of Audio Amatuer and building your own stereo. You could do it, but why? Of course, if you like the challange, like to play with your OS more than you do with apps, like to program, like to delve into the depths of your OS, then I understand the fascination with Linux. Otherwise, it isn’t worth the effort.
Satori.
Look at their distribution poll… Debian 40%. Well that’s cool but as it’s completely different as the last polls I saw, this just shows once again how much those website polls are worth.
The only way I feel Open Source community will get somewhere with their desktop is to do what I feel is happening with packaging software. Since RPM is becoming somewhat of a standard (I think it actually is) we should all focus on making it work better. Right now I think that FreeBSD has the best way of resolving this problem, but who said that we can not make it better – maybe better than on FreeBSD or on Debian…
What has this to do with desktop? Well, currently KDE is the leading desktop in among the Linuxoids. Focusing on it as a standard would help. I dont say Gnome is not as good, I like it, its better looking, but _if_we_focus_on_KDE_ we’ll be abel to makit good looking as Gnome.
What else? X question – yes it is problematic, but if we can not fix it we can introduce the successor that would include the benefits of X and add some more. If the compatibility cannot be held we can do as Apple did – allow it to run within the new windowing system.
“Right now I think that FreeBSD has the best way of resolving this problem”
Gentoo works similar to FreeBSD and while I believe it’s superiour to binary packages, it’s just not feasable for every user yet. Maybe next generation. Advanced users can switch to it and have most of their problems solved.
“What has this to do with desktop? Well, currently KDE is the leading desktop in among the Linuxoids. Focusing on it as a standard would help. I dont say Gnome is not as good, I like it, its better looking, but _if_we_focus_on_KDE_ we’ll be abel to makit good looking as Gnome.”
No, it’s not just about looks. You won’t make any GNOME user or develope like KDE by changing the look of it. It just won’t happen live with it. Also the competition between GNOME and KDE is very important and they are different enough to appeal to different users.
Currently you couldn’t make me switch to KDE even when paying me for it.
Also don’t forget that Trolltech holds the KDE monopoly for proprietory software.
Whatever, it just won’t happen but it’s not the death of desktopunix either.
“What else? X question – yes it is problematic”
No really. It’s evolving just like all the other big free projects. First it gets the functionality, then it will get the usability, then the fancy stuff.
We just need auto configuration and resolution changes “on the fly” on the usability side and real alpha blending (which is definetly working, just not finished) on the fancy side as this will allow for smoothed and transparent window borders, shadows, etc.
Also we need to get rid of the ugly strippled X background.
Other than that, there is nothing fundamentally wrong with X for the next few years (maybe Fresco could step in in ten years).
Desktopunix is currently evolving quite fine, thank you. Yes it doesn’t gain in marketshare (yet), but personally I’m more interested in it’s quality and usefullness than pure marketshare to be honest. And quality and usefullness is improving every day.
Quite frankly, I’m puzzled by those commenting how hard it is to install and use desktop Linux. I installed Mandrake Linux 9.0 beta 2 the other day, and was presented with an emaculately polished KDE desktop, as well as a wonderful GNOME 2 desktop. Both were excellent.
The fact that Mandrake is coming pre-installed on Microtel PC’s makes the reality even nearer. Sure Apple has bigger commercial developers and a prettier interface, but KDE and GNOME2 both beat WinXP in looks (IMO), and with things like the news about Productive going Free Software, the GNU/Linux desktop has arrived. Will it suceed? I think so.
RE: X. X is no longer a problem, it is easy to configure, very powerful, and offers some of the best anti-aliased fonts around. All we need is better actual fonts, and for slightly better integration between KDE and GNOME.
-Tim
maybe something like a w3c standards body would help. i would like to see something like that for cross platform and inter platform applications. w3c seems to work pretty well.
i would love to see each os compete in terms of its technical features not applications on it.
…sigh…
‘Cause I have had the alternative discussion again and again, my conclusion is always the same one: the only desktop alternative to Microsoft was BeOS, and it’s buried now in a Palm’s drawer.
I feel like MarkB towards Linux on the desktop, you’ve read my thoughts when you recall that Linux talk that goes “wait till x amount of time, or x product reaches 1.0”, it’s like a mirage, you can never touch it.
All I can’t say in this regard is that although it isn’t a complete solution for me and still has many rough edges (staring at a mirage?), I was very impressed by RedHat Limbo2. I like it. And I like to think that fonts will soon look on RH/GNOME (everywhere) as good as in my W2K. I also like to think their multimedia distribution will be a success.
But at the end of the day, my alternative thoughts are with BeOS. Damn!!!
“X is no longer a problem, it is easy to configure, very powerful, and offers some of the best anti-aliased fonts around.”
X is no longer a problem if you are so lucky and your card is supported well. There are still a lot of things missing in X. It also has a lot of annoyances. Having been around some forums and IRC channels I can tell you that there are tons of people out there who have a problem configuring X. It still does not work that well.
And for best anti-aliased fonts: I couldn’t care less because I think that anti-aliased fonts look damn ugly on my LCD display. I want everything aliased. That’s what I did on Red Hat Limbo – and the somewhat nice fonts went to totally ugly fonts, even the TrueType fonts. Maybe it is better in Limbo 2 (downloading it now).
The article mentions a few things that are critical:
– KDE vs. GNOME: it is a pain for developers
– Red Hat vs. SuSe vs. … : it is a pain for developers
Until those issues are addressed it will take a while for Linux to become “usable” on the desktop.
Dominance of Linux on the desktop is assured regardless of the usual objections from astroturfers here, for the simple reason that governments, organizations, and corporations worldwide will deploy linux on the desktop on a wide scale. That is already beginning. The home user market in America will be the last affected, after linux supplants MS Windows in schools, government offices and corporations worldwide.
<p>
There are many reasons why this is happening. Cost effectiveness, versatility, security, a need for governments and organizations to maintain control over their own data rather than surrender that control to an American monopoly, are a few very good reasons that far outweight the objections raised by Microsoft apologists and sychophants.
<p>
Actually the linux communuty and free software projects, some with corporate support, are devoting a lot of attention to the desktop these days. Some big players like IBM, Sun and Redhat have not been so enthusiastic to date, preferring to focus on the “safe bet” of linux server deployment and services. But even server solutions relying on linux indirectly enhance the attractiveness of linux on the desktop. For example, distributed computing, in which linux desktop computers can play a dual role, using the extra cpu cycles for enterprise level processing. A good example of this is the increasing reliance of the movie animation industry on linux. And that’s not just because of cheap processing power. The animation software being used is very good, and if a particular piece of desktop software is lacking it’s feasible for a large studio to develop it in-house because GPL and LGPL allow and encourage that. The same applies to banking, insurance and other industries although these are not as “sexy”.
<p>
The usual objections about lack of font support, performance issues with X11, lack of applications, lack of a standard gui toolkit, etc., are without merit and irrelavent even if they had merit. Just taking one, lack of a standard toolkit. While most Windows development suites build on the Win 32 API, the manner in which they do so varies widely. Delphi, MS Visual Studio, and .NET are just as different as Qt and Gtk. Because of the open nature of linux development environments, competition helps all of them. It is very easy for Gnome and Kde to borrow code from each other. There is no need to reverse engineer. This is not possible with closed MS application frameworks and toolkits, which are all dead ends because of their closed nature.
<p>
Hardware support for linux is much better than for Windows or any other system. Linux runs on almost any kind of hardware. Does Windows? Does Mac? Does Beos? Try running these on a mainframe and also on a handheld and on a random assortment of “white box” X86 PC’s. While systems like OpenBeos and Mac OSX are more attractive as pure desktop systems for home users, they don’t have the hardware support that linux has and are unlikely to get it unless they can piggyback off the drivers linux uses. MS is not even a player except in the X86 market, and its lack of flexibility to support 64 bit and eventually 128 bit processing will be telling.
<p>
The best news is that IBM, Sun and RedHat are now devoting a lot of attention to the desktop. Look for some announcements about that at Linux World.
<p>
Cheers / John
“Quite frankly, I’m puzzled by those commenting how hard it is to install and use desktop Linux.”
Linux is not so hard to install. And it isn’t so hard to use either. What is hard is when the default install isn’t what you want or need or when the install is somehow broken.
Example: I have a Dual PPro machine with 160M, an old Matrox video card and a 15 Apple Multisync monitor. Over the last few days it has seen installs of Suse, Mandrake, Redhat, Slackware (and QNX and BeOS). First thing I do after I rebbot is type “cat /proc/cpuinfo” and guess what? Some of them loaded SMP kernels, some didn’t. So, I tried compiling my own kernel. That was fun. Do a search through various documents, newsgroups and so on about compiling a kernel. These people can’t make up their mind what the procedure is. Some say do:
make clean
make dep
make bzImage
make modules
make modules_install
and then you have to go to /usr/src/linux/arch/i386/boot and “cp bzImage /boot/vmlinuz” or “cp bzImage /boot/vmlinuz-smp” or whatever you fell like typing that day and then you have to edit /etc/lilo.conf and then you have to run /sbin/lilo. And then you get to reboot and watch your system lock because of some error or another.
Of course, Some people switch the make dep and make clean. Some people don’t use make bzImage etc…. they use zImage or this or that or they just screw that and use make install. And so on and so on. There are about 100 ways (some wrong I am sure) to compile your own kernel and install it. I don’t have time for this sh*t. I got food to eat, work to do, sleep to have, fun to have and I need some time to go to the bathroom too. I don’t have “hours” to f*ck around with compiling kernels and editing config files.
Then we move on to monitor refresh rates. If these Linux gurus are such geniuses, how come none of these fools has figured out a way to let me change my refresh rate as easily as I do under BeOS, QNX, Win9x, NT4, Win2k, WinXP and so on? I have no time and no patience for editing XF86Config files. I have better things to do.
Then there are all of the problems related to missing libraries, improper symbolic links and so on. Oh ya, and you best hope the if you installed GPM that it doesn’t screw your X session up. Again, who has time for this sh*t? Not me.
Then you have to deal with the fact that there are 1000 Linux distros. Why is that a problem? Well, because they all choose to do things their own way. Among other things, file locations are often different and procedures for various tasks vary. And the documentation varies from distro to distro (and version to version). So, when you go looking for an answer to something you don’t find “an” answer. You find 5000 “possible” answers. And all of the “howto’s” and “mini howto’s” are outdated. Again, who has time for this garbage?
Look, I know Linux is sweet when it’s running. The Slackware 8.1 I just installed was faster than any other OS (and any other Linux distro) I tried (except BeOS) and that was “without” SMP enabled. It would have been sweet if my kernel compile had gone OK. It didn’t. I gave up. I simply don’t have a day, or two or three to waste on this crap. Even if I did, I wouldn’t want to.
Back to that Slackware install. I decided to use Windowmaker instead of KDE or Gnome. It was fast alright. The only problem is, many of the programs on the right click pop-up and in the Xfm filemanager did not work, at all. Again, I don’t have the time or desire to try and figure out why. And I shouldn’t “have” to.
If the Linux community wants to keep normal people, and even some power users away, they are doing a good job. Keep telling everyone Linux is easy to administer, troubleshoot and so on. You guys may believe that. I don’t.
I have spent days screwing with Linux distros. And I won’t waste any more time. Why should I spend “days” just trying to get ONE distro to work the way I want when I managed to get Apache, PHP, MySQL and Postnuke running (with no prior knowledge) on NT4 and Win2k in less than a day? The actual OS installs took less than an hour and there was no messing with text files and compiling. The rest took me a few hours over a period of a few days. Would Linux run smoother? Probably. But getting there would be “alot” more painfull.
Satori.
” Gentoo works similar to FreeBSD and while I believe it’s superiour to binary packages, it’s just not feasable for every user yet. Maybe next generation. Advanced users can switch to it and have most of their problems solved.
”
Yhis is actually the best thing about Linux, and everybody knows it – if you dont like it you can change it, reconfigure it, do whatever you want and make it work like you want. When Gentoo came out I was delighted that sombody finally got the message. I was a bit dissipointed when I heard that RPM is accepted as a standard for packaging but not anymore. The thing is, I believe that Linux Community will (It may take some time) understand that they must in EVERY distribution furnish some standard apps and what is more important preferences & default settings.
Think about this: You Like Gentoo? Fine! Lets say people at Gentoo decide their desktop is KDE. The official standard desktop is Gnome – so what do they do? They put Gnome as a “compatibility” desktop for the apps developed for it. This is nice, you can work in either of them. You have a your menu lists setup so you can see both KDE & Gnome apps listed. Until now everithing goes smooth. Then you install a new app. It does not matter if it is Gnome or KDE – but lets say it is KDE – it will show up (maybe) in your KDE menu and you’ll be able to use it easily from KDE. The moment you swich to Gnome (like me, just for a change) you loose continuity since the list in Gnome doesn’t show your beloved app.
What do I want? I want the standard Linux Distribution within every distro (not literally). If I install a new app I want it to install as it was the standar distro. My specialized super fast distro should be able to read the preferences from the standard files and put them in the place or format IT keeps. If I use Gentoo, I want to transparently install BSD style packages and RPM. If it is Deb I dont ask for anything different. You can play with distros as much as you like, make your own if you want but, HAVE A “PLAN B” in case the user cannot find the package in your format or does not know how to do something in KDE because he is a guest on the Comp and he is used to Gnome… I mean , it is so simple people – we just need a fall back solution.
>>Hardware support for linux is much better than for Windows or any other system. Linux runs on almost any kind of hardware. Does Windows? Does Mac? Does Beos?
linux yes, the gui for linux not always. if you do manage to get it to work it might not support optimal resolution for the gui’s which in my experience is greater than 800×600 but im no linux expert. just voicing my findings when i take it for test drives checking its progress.
Hey Satori, what’s the point of all those neat new integrated Linux-based operating systems when people then use Slackware?
Really, I can’t take that seriously. Linux is not Linux! If you run RedHat, you should be aware that you are runnign RedHat. Not just another Linux. It is using the Linux kernel but as a user, you should mostly deal with the stuff at the surface and that is RedHat (or Lycoris or Mandrake or whatever). When you have a problem, ask in the right forums, many Linux hackers will of course not give you an answer that is appropriate for a desktop user as that’s not their world. Or ask your distributions support. That’s why they are there. To solve the stuff for you. That’s what you pay them for!
We will soon see a new generation of Desktop operating systems with RedHat Limbo, Mandrake 9, Lycoris, etc. You will have to evaluate them on their own, not throughing them into one with Slackware. If they don’t choose the right kernel for you, complain! You shouldn’t have to compile your kernel. The fact is, that _real_ desktop users wouldn’t even try to compile their kernel. I mean those like my mother who barely know how to use a webbrowser. Surprisingly, I expect those people to have actually less trouble with Linux, as they won’t try things they are not supposed to do!!
Next time, get the latest and greatest which is supposed to be the prime example for a Linux desktop OS (I think it will be a decision between RedHat, Mandrake and some newcomers like Lycoris or Xandros) and just use the system like you want it to use. When you want to optimize something or install development software, you _can_ go to a console and do some hacking but if you don’t want to and something doesn’t work like it should, complain (friendly of course
)! Complain to those who got your money. It’s their job to fix it for you!
Sometimes you will have to live with some lower performance than possible without getting into the guts but you already pointed out that you are very willing to accept this for less trouble (with Windows) so you should also accept this when using Linux. If you do all this and it still sucks, then we can talk about.
Then you will have _real_ points of which has to be improved with the particular distribution in order to make the system more userfriendly.
</rant>
Second and this is what I posted the other day and its related to the pipedream of widespread linux use.
Without linux being bundled by large vendors like Dell, HP, IBM and without ISV’s like Intuit,Adobe etc, and yes MS porting there apps to Linux it will never register on the desktop radar.
This is either false or a tautology depending on what you mean. Obviously the definition of desktop is success if that the leading vendors sell it, and the leading applications support it so if that’s all you meant you aren’t really saying anything.
OTOH there is no reason to believe that Linux has to triumph outside of a general victory for free (as in freedom) software. There is nothing particularly good about a Linux victory where Linux becomes likes Windows supported by a bunch of commercial corporations offering a dumbed down desktop locking people into commercial software… Even if Linux were to win that way, what would be the point? So in reality I see a Linux victory (in any meaningful sense) as only being possible in a world where Microsoft, Intuit and Adobe support no longer mean very much.
As for Dell, HP.. there is very little PC OEM loyalty for desktop or low end servers. People and companies freely change PC OEMs. These guys are the leaders because they offer a product which meets customer needs at a very good price. If large numbers of customers want Linux and they fail to deliver it other OEMs will spring up and fill the demand. If Microsoft ultimately requires OEMs to sell exclusively their OS all that will happen is that the OEM market will split into “Microsoft” and non-Microsoft OEMs. Its not Dell that is preventing the adobtion of desktop Linux; though Dell could if it wanted to assist the movement greatly. OTOH if Dell had any real opinions about any technological issue they wouldn’t be Dell they’d be IBM.
Dell was initial created to fill a void in IBMs pricing policies. It has sustained itself by creating a method manufacturing unlike any seen before. However now that the technique is known it could be replicated rather easily. They aren’t a technology company they are a manufacturing company far more similar to an automaker than to an Intel or a Microsoft.
Even if Linux becomes the most easy to use OS ever made, without apps like Quicken, Office, and of course all the big games we will never get a large market share.
PCs did just fine without support for PSF, Cobol, JCL, which were the big apps before they came out.
Educating users is avoiding the problem that the apps they have been using for years are not available on linux. You have about as much chance getting them to change as you do the 200 million people who still use http://www.msn.com as their start page.
Users who had been on Wordstar for years switched to WordPerfect and then switched again to Word. Users who had been on Visicalc for years switched to Lotus 1-2-3 and then switched to Excel. Companies that have been on MVS for decades continue to switch more and more of their functionality to PC/NT.
Conversion can be done and it takes years to move the user base.
Let’s also talk about hardware support. Besides Nvidia and a few random companies most companies don’t officially support desktop linux. That is a real problem. Try calling X company and telling them you use Gentoo and can’t get your scanner to work. Unless xonsumer hardware and software products start saying Linux supported on the box, don’t except people to buy them or spend time hunting around newgroups to try to figure out how to install something.
Most hardware companies don’t want to deal with end customers at all. On many of the home level pieces of hardware even one customer support call can eat up the entire profit for the device. If Linux becomes even slightly popular hardware manufacturers will most likely directly support Linux; they don’t want the phone call at all.
I love linux and its a great OS, but widespread linux use on the desktop is not gonna happen. The ISV’s have said, thanks but no thanks, and the big desktopp hardware OEM’s have passed as well. I’ve been using linux since 1997 and every year someone says wait till X amount of time, or X product reaches 1.0 and then we’ll see some change. Well sadly that doesn’t happen and never will.
There are 5 major markets for computer systems:
a) home / small business
b) corporate desktop
c) embedded systems
d) server
e) enterprise class
In 1997 Linux had essentially no presense in any of the 5. Neither Linux nor NT really has offerings in class (e), though both are being (wrongly IMHO) considered for some uses in this class. In class (d) Linux is positioned to become the majority platform, this may not occur but people would have dismissed it as impossible in ’97. In class (c) Microsoft has continued to falter while Linux has become real player. This market is still very competitive and Windows (especially CE) and Linux are both playing for share. In class (b) some motion is occuring. This is a real battlefield for Linux, corporations have the internal knowledge to make the switch and their Microsofts costs are substantial if Linux suites were to pass Microsoft in quality I think we’d see a tidal wave of action on this front. In area (a) the real movement is likely to be on very inexpensive home systems. In terms of small business I think in general GPL business specific solutions are needed much more than a GPL OS.
The X desktop is less developed than most commercial desktops particularly in areas like apperance. Everyone knows that. OTOH network transparency has the potential to be Linux’s “killer app”. The ability to run any piece of software, with any data regardless of where it resides seamlessly is an absolutely huge feature. Unix users have it now; Windows users and Aqua users do not. Further it isn’t an easy thing to add as applications like Citrix Metaframe and Terminal Server show quite well.
OSX is doing a nice job of allowing Unix to experiment on what happens if we abandon X entirely. In any case abonding X should really be a distribution by distribution issue and not a broad Linux issue. For example for Servers I definitely want network transparency much more than I want pretty fonts.
I also should mention that most of the problems people have with X are not X vs. GDI vs. Quartz issues. For example all 3 systems support pretty fonts and all 3 support ugly fonts roughly equally well. Its really a question of what font libraries the systems ship with and getting rid of X wouldn’t solve the font issue.
There are 5 major markets for computer systems:
a) home / small business
b) corporate desktop
c) embedded systems
d) server
e) enterprise class
In 1997 Linux had essentially no presense in any of the 5. Neither Linux nor NT really has offerings in class (e), though both are being (wrongly IMHO) considered for some uses in this class. In class (d) Linux is positioned to become the majority platform, this may not occur but people would have dismissed it as impossible in ’97. In class (c) Microsoft has continued to falter while Linux has become real player. This market is still very competitive and Windows (especially CE) and Linux are both playing for share. In class (b) some motion is occuring. This is a real battlefield for Linux, corporations have the internal knowledge to make the switch and their Microsofts costs are substantial if Linux suites were to pass Microsoft in quality I think we’d see a tidal wave of action on this front. In area (a) the real movement is likely to be on very inexpensive home systems. In terms of small business I think in general GPL business specific solutions are needed much more than a GPL OS.
I couldn’t agree more.
However, I think part of the point has been missed. Palladium has the potential to lock the desktop to the server. If a microsoft desktop does not seamlessly and securely talk to the Linux server then there will be a problem selling the Linux server.
Cheers
David
The only way Linux stands a chance on the desktop is if X is abandoned and replaced by a superior windowing system which solves all the problems I’ve mentioned.
That’s really a very silly statement IMHO. The problems you mentioned are much, muuch easier to fix in the current XFree86 version. What you are proposing is to create yet another windowing system, and expect the world to jump on that bandwagon. That’s simply not going to happen! Instead of wasting precious engineering resources reinventing the wheel why not try and FIX the current shortcomings?
One of the problems you mentioned is the lack of support for switching desktop resolutions. There is the XRANDR extension (X Resize And Rotate) which is being worked on. Why not help those folks?
There is also work being done on automatic XFree86 configuration. The handfull of people (there is probably noone working fulltime on this) could need all the help they can get!
And if you still believe that throwing away X is the only solution, by all means, go ahead and create your own windowing system. Just make sure you support my Geforce4 ti4600. And with support I mean 2D *and* 3D (hint: that means convincing NVIDIA to switch to your new driver API)
-fooks (who thinks 3D a la Apple’s Jaguar is the future for desktops)
You read my whole post and are going to suggest that the problem is Slackware? And what about Mandrake 8.2? And Suse 8? And Redhat 7.3? Tell me, in any of the distros, where in their control panels can you change refresh rate? You can change bit depth and resolution. In a couple you can even choose your monitor, if it’s listed, mine isn’t. And even if it “was” listed, tell me “where” in these control panels you can select your refresh rate from a drop down manu like you do in Winwhatever or BeOS or QNX? You can’t. I don’t want to select “High Frequency SVGA monitor (1024×768 @ 70Hz)” or whatever other generic entry (or non-generic entry) and let X decide what refresh rate I should be running at at a given resolution, I want to set the friggin refresh rate to what I want, when I want, as often as I want. And I want it to be as easy as it is in Winwhatever, BeOS and QNX. Until I can, until people like you will admit that it absolutely should be easy as pie to set your refresh rate (and do many other things), just like other OS’s, Linux is going NOWHERE on the desktop.
I have tried all of the major distros and a few not so major ones. Most of the flashier distros (Suse, Redhat, Mandrake, Lycoris etc…) work decently out of the box but are somewhat slow (understandable, but still, if you do decide to recompile you should be able to find out how without finding 100 varying answers all of which are a shot in the dark). And none allow me to make the refresh rate what I want it to be with ease.
The refresh rate and kernel compiling issues are just two that I chose. The first was chosen to show how difficult it still is to do the most trivial things in Linux sometimes. The latter was more to show what a mess one runs into when looking for information on how to accomplish certain tasks in Linux. Part of Linux’s benefit (flexibility) is also a detriment to it thriving on the desktop. The LAST thing you want to do when promoting a product is “confuse” the customer. And that is exactly what happens with Linux. And that’s even before they have actually installed the thing.
Anyhow, you go back to your dreamworld where Linux is every bit as easy as Winwhatever.
Satori.
There are a couple of things I see missing in Linux before we have broad acceptance:
– good support for softmodems of all kind (including fax support)
– an easy way to install drivers, common to all distributions
– easy printer sharing (without having to read an howto or dig in command lines/package management for hours)
some of those points are already being addressed, but they still have a long way to go before reaching a standard-user usability level.
Many people think X is a windowmanager with a menu just like Windows or Mac. But it isn’t, it does not offer a consistent widget set. Notice how most apps in windows look the same. This is because they all use a similar windows widget set.
X has these widget sets but the application developers have a choice. Sometimes the app you want is written with a different widget set from the window manager you use. But you can still run the program. For instance running Gimp under KDE or konquerer under Blackbox.
Some people complain that Linux is inconsistent because of X because they believe that the GUI is the centre of the OS. In Linux this is not the case. How many people can use Windows without ever using the GUI? I know people that do this with Linux. In Linux the non GUI apps are more important than the GUI ones. Most interfaces are tacked on in front of a console app. This is because Linux is used (not designed) as a server.
Of course Linux can be a desktop system too. All it needs is a consistent set of apps. You can do this by limiting the apps to mainly one desktop environment like lycoris does. A small app that changes the refresh rate and screen resolution would not be hard to make (In fact I don’t know why someone hasn’t done it) an applet for Gnome and KDE would go a long way.
And anyone that complains about Slackware being user unfriendly will be laughed at.
Disclaimer I use blackbox under Slackware 8.1
Why would you want to change your refresh rate? Shouldn’t it just use the best that is possible? Why would a newbie user even know what a refresh rate is? And why should he care? Win2k limited my frequency for Quake 3 to 60hz because of a bug in the OpenGL code and I had to apply some weird patch to my display driver to fix this. Every OS has small problems but one of them won’t make it crap. Just talk about it reasonably and don’t get into overheated flames and maybe somebody will take you seriously. Accusing me of living into a dreamworld is ridiculous as I never stated that current GNU/Linux distributions would be that easy to use. Actually I couldn’t care less.
Even without the refresh rate issue (if it’s even an issue) and without users trying to compile own kernels for some weird reason there are still enough usability issues. More serious ones that should be tackled.
If the naming confuses the user, I really couldn’t care less as unfortunatly this can’t be changed. If customers refuse to accept anything but the hyped “Linux” term, than what should we do against it? Nothing. We can just live with it and make the best of it. If this will make people not use Linux, fine. As I said, it can’t be changed anymore but it’s not a reason to flame.
I still don’t think it’s a big issue though as 90% of the potential “beginner” customers wouldn’t even ever look at a Linux webforum to solve their problems. But obviously at webforums you will only find those who do.
So they aren’t really representative at all. The same applies to OSNews readers btw. In one way or another we are all basically “techies”.
THE CORPORATE MARKET!!!
simple as that…why is windows dominant? b/c it conquered the business world…schools want to teach children how to use the computers that businesses will use…so they follow suit…and home users wanna use what their company uses so they can take work home
simple as that…and this is why Linux will eventually win…it doesn’t have to be the best at anything…just the best overall…think about this…what’s windows best at? Compatibility…that’s about it…but it the best OVERALL
-bytes256
“Why would you want to change your refresh rate?”
Why? Because it always defaults to 70hz. My eyes can’t take 70hz. My other computer has a 19″ DiamondPro that I run at 85hz. I would run it at an ever higher refresh rate if I could.
“Shouldn’t it just use the best that is possible?”
Well, it doesn’t. Part of the reason is that this monitor is not listed in the database. So, I have to choose a generic monitor. And there is no generic setting the resolution and refresh rate I want. So, that means editing my XF86Config.
“Why would a newbie user even know what a refresh rate is? And why should he care?”
Well, a newbie might not, in fact most casual users have no idea they are killing their eyes with their refresh set at 60hz. So what? The refresh rate example was just one of many. And if I, an experienced computer user, can’t get something as simple as refresh rate set (or rather, that SHOULD be simple to set) in XFree86, how well do you think someone less experienced than I is going to do the first time they go to do something that was simple under Winwhatever only to find out they have to edit obscure config files?
“Every OS has small problems but one of them won’t make it crap.”
Absolutely. First, I never said Linux was crap. I just dispute whether or not it has a chance at becoming a desktop OS, a serious contender in the desktop OS arena. The difference though is that with Windows, you can usually find out what the problem is fairly easy. There aren’t 100 variations of win2k with different kernels, bootloaders, window managers, install scripts, included applications etc… Win2k is Win2k, basically.
And forget the kernel compiling or refresh rate issues in themselves. I installed Slackware 8.1 last. I chose Windowmaker as my default desktop. Now, after compiling my new kernel I had to edit /etc/lilo.conf before running /sbin/lilo. No problem. Right click the desktop and pick a text editor from the pop-up menu. Should be a cakewalk. Well, it would have been had ANY of the text editors on the pop-up menu worked. I either got nothing or an error message. Keep in mind, when I did the Slackware install, after selecting the sets I wanted installed (I didn’t install TeX or Emacs) I told Slackware to install “everything”, not to prompt me for individual package selection. Nearly 2 gig worth. So, where are these applications? Do I have to go running around looking for them? Why? Why should I have to do that? The installer should have had it all set up perfectly for me. Luckily, since I installed everything, I was able to find the consolde text editor called “joe” which I am familiar with. But I was already fuming at that point. I did manage to add the new kernel to my lilo.conf and when I ran /sbin/lilo it found the new entry and all seemed fine. That was, until I chose to boot using the SMP kernel and within seconds got an error. I could have spent a significant amount of time tracking down posts on dejanews and readin more howto’s (which inevitably always refer you to yet another howto or two or a dozen). But, I’ve had it.
It’s funny. I have these discussions all the time and people tell me Linux isn’t that hard. And then I read yet another article telling me how Linux is getting easier and easier. And then I see yet another new version of Mandrake come out and it’s supposed to be easier and better. So, I give in an try yet again to make Linux do what I want it to do. And I am disappointed each and every time (for various reasons from poor font rendering (and truly screwy font path problems) to the refresh rate problem to poor tablet support, becuase I could never put my Geforce2 into the bit depth I wanted at the resolution I wanted to….. well, you get the idea). I have installed various distros of Linux (over many years) on 386’s, 486’s, a P90 a P150, a P166MMX, a PPro 200 a P3 800 and now a Dual PPro 200. I bet you I have installed Linux more than 100 times. It doesn’t get any more pleasurable really. the “install” part has gotten easier. But getting an OS “installed” is the easy part. Configuring it and administrating it once it’s installed is another story.
There is only one time that I have enjoyed my Linux experience and that was because I really didn’t have to interact with Linux. I used, for quite some time, a single floppy firewall/router distro called CoyoteLinux. Setup was easy from either a script or a GUI and then I never had to look at it again. I just booted my old 486 and went about my business, though if I actually had to get in there and mess with ipchains and so on I would have again been frustrated. Luckily, at that point, I didn’t need to open any ports and so on. Now I do. But now I use a cable/dsl router and a software firewall. Both are easy as pie to administer.
Anyhow, I really would like to like Linux. I am not married to MS by any stretch of the imagination. I am married to ease of use and applications first, performance second. I would love to see OS X for Intel. That would be about enough for me to stop using Windows. The minute a new OS for Intel comes along that offers the applications and usability of Windows or Mac (I don’t expect the sheer number of applications that Windows has just the type and quality that I want/need)) I am ready to dump Windows forever. I would buy a Mac if it wasn’t for the fact that they are substantially slower than yet substantially more expensive than PC’s.
Linux needs to change for me and people like me. Not the other way around. I know, there are a handfull of new “easy” distros intended to capture Windows converts. They certainly have come a long way. But they still get some of the basics so wrong. The simplest things, that could make a huge difference to people like myself, are ignored.
Satori.
That’s really a very silly statement IMHO. The problems you mentioned are much, muuch easier to fix in the current XFree86 version. What you are proposing is to create yet another windowing system, and expect the world to jump on that bandwagon. That’s simply not going to happen! Instead of wasting precious engineering resources reinventing the wheel why not try and FIX the current shortcomings?
I’ve explained many times the inherent problems with the X architecture. Most of these stem from attempting to solve crippling architectural issues through the use of extensions. The problem is that applications do not make use of protocol extensions, and therefore the extensions are rendered inherently useless.
The first and foremost issue is X’s choice for IPC, sockets. While sockets provide a nicer interface than, say, Mach or SVR4 message queues, their performance is somewhat lacking. This can be solved through the use of shared memory.
X can, of course, use shared memory, by utilizing the MIT shm extension. Unfortunately, nothing uses this extension, and consequently X gets bogged down and lags when some application attempts to send a large number of pixmaps through a socket.
And if you still believe that throwing away X is the only solution, by all means, go ahead and create your own windowing system.
That’s precisely what I’d like to do.
> XFree is still lacking simple
> features such as a means to
> choose desktop size (and a
> monitor resolution accordingly)
I’m not sure what you are using, but Limbo 2 has a “display” app under “System” that you can use to change resolutions and stuff. You will still need to restart the X server, (which is NOT a big issue), but I hear Mandrake already has an on-the-fly solution so I would imagine everybody else would follow suit.
I think some of your last post was rather unfair…first of all…Slackware is a very “hacker” distribution…you are supposed to use config files…that’s part of its appeal…now if you have to go to config files under Mandrake…that’s a different story…mandrake is supposed to be easy to use…i haven’t used a recent enough version to say whether it really is or not…
i use redhat 7.3 (valhalla) right now…and i must admit that i’m very impressed…the default install is pretty darn good and i’ve seldom had to edit config files unless i wanted to for efficiency reasons…there are a lot of nice config tools out there…redhat’s Xconfigurator is really good…easy autoconfiguration…unfortunately my i830m isn’t supported well by XFree (stuck at 8-bit for 1024×768 when it can do 24-bit in windows), but that’s not RedHat’s fault at all…everything else, including the sound card and network card were installed and autodetected flawlessly…
also, i do computer troubleshooting on the side and i’ve noticed a good deal of users who still have their settings left at DEFAULTS!!! do you honestly expect these people to care what their refresh rate is?
as long as you stick to one of the major desktops, KDE, Gnome, WindowMaker…it’s pretty easy to configure the things users tend to care about…themes, desktop backgrounds, icons, screensavers, etc.
also as far as menus go…once again, as long as you stick to KDE or Gnome and use a big distro…Red Hat, Mandrake, SuSE, and stick to the official packages…odds are very good that the app you just installed will find it’s way into your menus
as far as apps, what are you looking for that Linux doesn’t have? AbiWord is SUPERIOR to MS Word for my needs…just enough functionality without the stupid bloat and in-my-way features that MS Word has…spreadsheets? Gnumeric, KSpread, OpenOffice, StarOffice…they all have pretty good spreadsheets…i’ll admit i’m not a hardcore excel user, but i have found more than enough functionality in all these programs
web browsing is in good shape too…Galeon is the best browser i’ve used since Netscape Navigator Gold 3.0…seriously!
to be quite honest, if it weren’t for the XFree problem (movies look like SH*T at 8-bits) i’d use linux almost full-time
linux is ready for at least the corporate desktop and probably schools too…MS should definitely worry
-bytes256
>The only way Linux stands a chance on the desktop is if >someone comes along and decides to do with it what Apple >did in creating OS X.
I for one would be glad to see OS X succeed on the desktop. Is it happening though? Has OS X increased Apple’s share of the desktop? So far, absolutely NO. So there has to be more to desktop success than that Apple has done.
Don’t forget, folks, that the world DOES NOT end in North America. Right now, there are a few places in the world where the use of a linux desktop is routine, and a lot more other places where that sort of thing is being considered right now. Run a modern distribution on modern hardware, and you will see that Linux is definitely a viable alternative for joe user. Whether it will be so used is something we will have to see.
Carpe Diem: I use Linux. I love Linux.
One thing that strikes me as standing in the way a desktop (besides hardware vendor’s reluctance to ship desktop systems with linux preinstalled) is the libraries.
There are many ways of creating applications on Linux, GTK, QT, FOX, etc. This is the beauty and curse of Linux. The beauty is the variety and ease in which an app can be coded but the curse is many Linux apps look very different from each other and use different keystrokes for the same task.
I find it fustrating when I go install something, running into needing this library or that library. or when I want to, let’s say, quit an app. It might be alt-s, alt-x or control-q, etc. You may not even be aware of this, but it is a problem.
No, don’t ask me for a solution. I don’t have one. This is just my 2 cents. Carpe Diem. Your mileage may vary, uh, seize the day, remmember the the titans, etc.
“also, i do computer troubleshooting on the side and i’ve noticed a good deal of users who still have their settings left at DEFAULTS!!! do you honestly expect these people to care what their refresh rate is? ”
So tell me, if ease of changing refresh rate is not important, why did any other OS developer (MS, Be, Apple and so on) bother making it so easy to do? I guess you believe that they just had lots of time/money to waste?
“as far as apps, what are you looking for that Linux doesn’t have?”
Oh, I don’t know. Let’s see:
Dreamweaver
Fireworks
Photoshop
CorelDraw and Photo-Paint (and no, the Linux version that actually runs under Wine doesn’t count – it’s a DOG with huge fleas)
Satori FilmFX64
Painter
and many more….
And that’s just the 2d stuff.
Look, to have a place on my second “hobby” machine, Linux needs none of those apps. But to have a place on my main machine it better have them or equivalents to them. And not just somewhat similar apps (and no, Gimp and Impress won’t cut it), I want color profiles, proper tablet support and so on.
But, of course, no amount of quality apps will matter until I find Linux/X as easy to use as a Windows or Mac OS setup.
Satori.
Libraries are kinda a problem as you say, but I don’t think they are that big of a deal…first of all, even on Windows apps aren’t consistent…that’s not really a problem…in any case there’s usually a KDE and a Gnome version of the same kind of program, so you can use the one that feels like your desktop of choice…
I really think that Linux is damn near ready for the destop…look at the things people complain about now i think that’s the most illuminating thing…
1) “my fonts look ugly”…look back a coupla years…”we need true-type font support”…things are moving on…and there’s a lotta work on this…fonts are pretty much a moot issue right now even
2) “X is hard to resize/change refresh rates for/configure”…remember when everyone complained about Linux being hard to INSTALL…things have come a long way…
3) “Apps are inconsistent”…remember when people used to complain that there were no apps!…i expect that in a few years, QT and gtk+ will be capable of supporting similar looking and feeling themes…if the user wants that, that is
the most impressive thing i find in Linux is that it improves at such an incredible rate…i can’t wait til Linux stops playing catch-up…just think what it’ll be able to do!
i would have to agree that XFree86 is the biggest thing holding Linux back right now…but its being improved pretty quickly…i see no reason to ditch it…most of the problems people have with X are XFree86 implementation issues…and these are improving…
things are looking very good for linux
No offense, but you’re not the typical case…
these kinds of apps are relatively new to WINDOWS, let alone Linux…graphics work has always been traditionally a Mac/SGI stronghold…so yeah i’ll admit that for that stuff Linux is almost worthless…of course you could always run those apps under VMWare…but 300 bucks is a lot of money just to save you from rebooting…hell you could buy a windows box to set next to your linux machine for that…
but anyway…as far as TYPICAL users are concerned…Linux is more than adequate in its current state…and as far as screen refresh rates are concerned…yeah some people like yourself care about that…but avg. joe just wants to check his email, surf the web, and type some stuff into a word processor…ever heard of the 80/20 rule…80% of the people use 20% of the features…linux has about 90% of Window’s desktop capability plus a couple more that MS hasn’t even thought of
i stand by my comments for AVERAGE users…
-bytes256
“I for one would be glad to see OS X succeed on the desktop. Is it happening though? Has OS X increased Apple’s share of the desktop? So far, absolutely NO. So there has to be more to desktop success than that Apple has done.”
That has everything to do with the price of Mac hardware and nothing to do with OS X itself. The real issue, for Apple anyhow, is how come few of it’s users have switched to OS X (I would guess legacy applications needing to run under OS 9 is the biggest reason). But you put OS X on a PC running an Intel or AMD chip and I guarentee you Apple’s market share will multiply. OS X on Mac will definitely win some converts from Linux and Windows. But not in huge numbers. The hardware is just too damned overpriced.
“Don’t forget, folks, that the world DOES NOT end in North America. Right now, there are a few places in the world where the use of a linux desktop is routine, and a lot more other places where that sort of thing is being considered right now.”
And where might that be? In places where they have little choice because they simply don’t have the money? If this is all so true, if Linux easily replaces Windows on the desktop, why do most people who try Linux run the other way as fast as they can? The same cannot be said of Macs for example. I can sit down at a Mac and be working effectively no problem. The first time I actually used a Mac for more than 20 seconds was when my parents bought one a couple of years ago. And guess who taught them how to use it? Me. Imagine that. No experience with the OS and I am able to use it without opening a manual. And I don’t care what you or anyone else tells me, Linux does not offer the same experience. The transition from Mac or Windows to Linux is far harsher than from Mac to Windows or the other way around. That speaks volumes.
“Run a modern distribution on modern hardware, and you will see that Linux is definitely a viable alternative for joe user. Whether it will be so used is something we will have to see.”
I have a P3 800, 512M, Geforce2 GTS and so on. I had to download drivers (two files) from nVidia to get 3d acceleration, install them and then, you guessed it, edit a config file to enable the driver. Of course, that wasn’t too bad, unfortunately, KDE wouldn’t let me set my desktop to 1280×1024 with 32bit color, only 16bit (yuck).
The question really isn’t can “joe user” install Linux and open Abiword. The question is, what in God’s name does he do when he runs into a problem? He will, eventually. That’s when things get hairy. Who does he call? What website does he go to? Does he get the “right” advice or advice from some guy running an older version or different distro? Is that guy running a simple setup, a setup with /usr? Is “Joe user” logged in as root? Is someone going to walk “Joe user” through editing config files, chmoding this or that? “Joe user” will be down at the computer store buying a new version of Windows before it gets to that point. You say, well if he “pays” for Suse or Mandrake or Redhat, he can get support from them. Sure, if he “pays”. Like that’s really going to happen. “Joe user” doesn’t know Linux from molecular science. And he doesn’t care. Especially if it’s going to cost him. “Joe user” doesn’t hate Windows either. He doesn’t know enough to know there is anything better out there or that there is anything “wrong” with Windows to begin with. He is too busy using Win98 to even bother looking at Win2k or Win XP, let alone Linux. And without friends and family running Linux to help him when he does run into trouble there isn’t a chance, not even the slightest, that he will bother with Linux. And then you have Windows “power users”. Even they have troubles with Linux. So, most of them are staying away. If they won’t run Linux, what makes you think “Joe user” will?
Back to Apple. If Apple released OS X for Intel, watch out. Now “Joe user” has something to think about. He’s heard how Macs are easier to use, he has seen his friends Mac and likes the look of the desktop and so on. But he’s never been willing to pay the entry price into the Mac world. Lots of “Joe users” will stick with Windows because it’s what they know, it has huge “mindshare” and they are worried about “applications and games”. But many will take the chance. After all, if it runs on their PC, the worst that could happen is they have to throw Windows back on the machine. In fact, Apple should release OS X for Intel and offer to exchange it for and install Windows if they aren’t happy. I doubt many would take them up on the exchange offer.
As for OS X for Intel and the “power user”, that’s a no brainer. I have tried just about every alternative OS (Linux, FreeBSD, BeOS, QNX, OS/2 and so on). Just for fun. Just as a second OS. Offer me something that can really replace Windows, ease of use, software and all BUT with *nix underpinnings and at the very least I will try it. At worst, it will become my secondary OS. But you better believe I am buying it, like right now. I can guarentee you that OS X for Intel would make Linux’s marketshare on the desktop look miniscule in comparison in very short order. Even Linux users, the ones who use Linux because it’s not Windows rather than the ones who can’t bring themselves to pay for software, will give it a whirl.
Satori
“I would buy a Mac if it wasn’t for the fact that they are substantially slower than yet substantially more expensive than PC’s.”
Wow, I can’t believe someone would ever say that. It has always been my experience that the Macintosh has been significantly faster than x86 Windows-based PCs…. That is ofcourse assuming we exclude the past 3-4 months, as Apple hasn’t made any major processor upgrades to their high-end towers.
“No offense, but you’re not the typical case…”
But I’m an “important case”. Why? Because I’m the guy everyone I know (including clients) calls when they have computer problems. If I can’t run Linux with ease, neither can they. I am also more likely than “Joe user” to be interested in an alternative OS.
“these kinds of apps are relatively new to WINDOWS, let alone Linux…graphics work has always been traditionally a Mac/SGI stronghold…so yeah i’ll admit that for that stuff Linux is almost worthless…of course you could always run those apps under VMWare…but 300 bucks is a lot of money just to save you from rebooting…hell you could buy a windows box to set next to your linux machine for that…”
Photoshop is new to Windows? I don’t think so. CorelDraw and Photo-Paint are new to Windows? I don’t think so. Dreamweaver is new to Windows? I don’t think so. Painter is new to Windows? I don’t think so. FilmFX64 is new to Windows? I don’t think so. Lightwave has been on Windows for ages. Same with 3DS Max, Softimage, Houdini and many others. It has been many years since Windows was a minor player in the graphics arena.
As for VMWare, all I can say is, yuck. It’s still too slow. You can’t get 3D acceleration in VMWare either. And that still doesn’t do anything about the tablet situation or the video card issues.
“i stand by my comments for AVERAGE users…”
I guess we will have to agree to disagree. You see “Joe user” as a guy who buys a computer and “never” needs any help and never needs to configure anything at all. I see him as a person who has a huge support system around him that he depends on to come to the rescue when these things come up. And because of the nature of Linux itself, the size of the community, the number of distros and so on, I can see the problems froma mile away.
Satori.
> Palladium has the potential to lock the desktop to the server. If a microsoft desktop does not seamlessly and
> securely talk to the Linux server then there will be a problem selling the Linux server.
The problem with Palladium doing that is that they’ll hit other systems too. Remember Palladium’s main focus is to allow multi-media content. Corporate America doesn’t care that much about being able to download DVDs of Spidar Man 4 or Debbie Does Dallas 26. OTOH if the system can’t talk well with other servers that’s going to be a big deal. Sure MVS may make changes to support Palladium but what about those older systems that are still in use but unsupported (like HP-MPE)? Either Palladium has little effect on servers or it fails to get onto corporat desktops IMHO.
“Wow, I can’t believe someone would ever say that. It has always been my experience that the Macintosh has been significantly faster than x86 Windows-based PCs…. That is ofcourse assuming we exclude the past 3-4 months, as Apple hasn’t made any major processor upgrades to their high-end towers.”
Well, you are one of the only people who still thinks that way. In the area of graphics, in particular, even the Mac fanatics are admitting it. Here are two articles to look at:
http://www.digitalvideoediting.com/2002/05_may/features/cw_aeshowdo…
http://www.digitalvideoediting.com/2002/07_jul/features/cw_macvspc2…
And please don’t tell me benchmarks don’t matter. They certainly do when you are using graphics applications. They speak volumes. Another example I can provide: On a Cinema4D forum, people compared their render times on an identical scene. My P3 800 rendered it faster than any Mac up until we hit Dual G4 800’s. Those were a little faster than my P3. I rendered in 6 minutes 45 seconds, and two people submitted results for their Dual G4 800’s. One did it in 05:57, another did it in 05:33. Would you care to compare the price of a Single P3 800 machine to a Dual G4 800? I should mention, before you try and say otherwise, Cinema4D is SMP enabled so those Dual G4 renders were in fact using both CPU’s. Here are the final results:
http://www.postforum.com/forums/read.php?f=6&i=49528&t=49528
Oh, and BTW, my ancient Dual PPro 200 with 160M rendered the scene in about the same time as the newer G3 400. Did I mention that I paid about $75 US for that PPro system? Try and buy a G3 400 for $75.
Sorry.
Satori.
…without apps like Quicken, Office, and of course all the big games we will never get a large market share…
You make some good points, but Quicken, Office, etc. does run on Linux now.
Just an appendium to my libraries post: Take a look a Freshmeat.net and tell me how many consumer level (like games, office suites, email, etc) are posted daily. Not many.
Progress is continually being made, but its going to take at least another 5 years before most businesses can use in on primary desktop computers. All of the libraries, desktops, distributions, poor printing configuration and utilzation, and no real standard OLE built in all hinder Linux on the desktop. Yes you can do many things with Linux today, as I do personally, but it still takes a lot of manual configuration even on the most popular distributions. Much hardware still does not run properly under linux, or the hardware has to be tweaked manually in order to run properly. I suspect in the next 5 years what will happen, is more and more distributions, libraries and less popular applications for Linux will drop out of favor. Eventually people will get serious and start dropping the old “DOS command prompt” way of thinking. I mean I do a lot of programming, and I still want to program text based apps, even when GUI apps should be the only thing being worked on. Basically, like I said, its going to take a few more years before the dominant Linux players come out and start leading the pack. I see KDE and RedHat doing this already. These are the people that are going to make the user experience uniform, because despite the fact that they may make choices that other Linux programmers disagree with, they will have enough “critical mass” to unite larger groups with their designs. Want to argue your Linux cause with me? I am not in that business, but just wait and see over the next few years. History repeats itself, and this happened with other operating systems (DOS comes to mind).
Dano.