“The vision of running Linux on corporate desktops has gained ground during the past 18 months, as full- featured office productivity software has become a reality and improvements have been made to the Linux kernel and to installation and administration tools. But even though the open-source operating system has moved closer to filling desktop needs, nagging gaps remain, said users and analysts at last week’s Enterprise Linux Forum Conference & Expo here. In some cases, they added, the lingering lack of needed functionality is making it hard for IT managers to switch their users to Linux.” Read the article at PCWorld.
What’s Keeping Linux Off Desktops?
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Why did businesses buy an expensive, dated technology such as the IBM PC?
Familiarity.
No-one gets fired for buying IBM, we all know that saying. And why not IBM? The typwriters supposed to be replaced by that brand spanking XT with its green phosphor is from IBM. The paychecks in your company might very well run off an IBM thing. If you’re a sufficiently big enterprise, chances are you’ve got an IBM mainframe or mini in your cellar. Surely, they must be a good choice in desktop computing as well. And it’s IBM.
Only IBM had the power to give Microsoft the power they have.
So you’ve got your PC, XT, AT. Of course it runs MS-DOS. Clones are appearing, and since they run the industry standard applications and MS-DOS, you might buy those instead. They’re still familiar, still IBM. And everyone else does it.
Now, the clones are starting to come down in price, and the employees are starting to buy PC clones, despite rather high prices and low specs/usability. It’s familiar.
Replace IBM with M$ for today’s situation. No-one gets fired for buying MS. Not even big company executives in direct competition with MS. IBM sells computers loaded with MS software. So do HP, DEC and SGI.
Thanks to familiarity, people sho shouldn’t be making such decisions decide that M$ is the way to go, from the PDA to the server. It’s familiar.
Why switch to Linux? It’s a black horse. They’ve never switched to Mac, never to OS/2, only gone from MS-DOS 3.3 to 5.0 to Windows 3 to 95 to XP. Probably thrown off their old Wordstars and Wordperfects along the way to streamline into a familiar M$ environment.
They’ve never used Linux, and have no particular incentive. True, M$ might be more expensive, but so was IBM, and at least it worked. Already dozens of systems have been thrown along the wayside in the neverending march in the trails of Big Blue and the Redmont Empire. Why switch to Linux when they’ve already discarded so many other alternatives? Sympathy? Because it’s said to be cheaper, though entirely untested in their environment?
While plain office tasks might not be the most demanding in the world, its users are about as reluctant to drop their existing setup as HP CAD shops, SUN circuit layout shops or SGI rendering shops are, or for that purpose Mac DTP shops. Their thing works fine, people are used to it, it fulfills their demands. Unless something extraordinary happens, you’re not going to change your system just like that. Even though you might save some money, why put it on Linux just yet?
They’ll look into the saving opportunities in time, but the Linux offer is not all that convincing just yet.
It’s probably been said before, but let me say it again: what Linux lacks is a standardized GUI/Desktop concept.
First of all from the end user’s point of view, but that’s a completely different topic (usability).
But no less important, it must be standardised from the *developer*’s view.
Suppose I want to write a GUI app for Linux. Do I use Gnome? Do I use KDE? Motif? Athena Widgetset anyone? Do I use FreeType? Or maybe the built-in X fonts? 150dpi? 100dpi?
Now I want printing. Ghostscript? lpd? Roll my own?
etc.
What I’m trying to say is that you need to offer developers a single API, at least for the desktop. Once you get this straight, software house will come. Lacking this, Linux or whatever free Unix will never be a hit on the corporate desktop.
Off topic, but as for basic non-GUI oriented server software, it does the trick well enough; it’s just another “flavor” of Unix you need to compile for.
K.
1. People have never seen Linux.
2. People have windoze-based hardware (winmodems, win-scanners, etc.) that won’t work with Linux.
3. People are … stupid. Hmm… Sorry. I don’t meen
_all_ the people around. I mean only those who do not
belong to the above two groups but are still running
windoze. Let’s call them windiots.
Linux, with the noteworthy exception of Red Hat Linux 8 hasn’t a clean, consistent interface. No fussing between KDE and GNOME, unification and standardization, as another poster mentioned is the key.
Useability, consistency, and standardisation.
And patience ..
consistancy, such as Kalimotxo (IP: —.red-80-35-218.pooles.rima-tde.net) aren’t even developers themselves, the majority of developers I know are quite happy with the way things are done on Linux. The developers I talk to haven’t ported it because they don’t see the demand, so there for, they can’t make back the money spent on porting their application.
The defacto standard for commercial development on linux is qt. The majority of users run KDE. Those are two simple facts. There have always been and always be multiple tool kits as each address a failing in another tool kit. The two tool kits in Linux are qt and gtk, qt for C++, gtk for C programming. One only needs to have a unified theme and voila, no problems.
As for usability “guides”, they’re absolutely pathetic. How many programmers do you know who stick to the usability specifications that Apple or Microsoft publish? Even if X11/KDE/Gnome etc at a unified guidelines, it stil wouldn’t make a lick of difference.
Btw. in most cases, commercial developers stick with the KDE usability guidelines.
i think one major problem is that a lot of organizations use two or three esoteric applications that tie into databases and only run on windows.. you either have to try to emulate the applications under wine or convert the applications and databases into a more linux friendly solution. wine might not run the app reliably and introducing new software requires a lot of retraining for the users.
at least, i think thats the biggest deal on the corporate front. i think it’s much more important than desktop consistency or microsoft office compatibility.
Every time Linux users tell me that it’s ready for the desktop, i never use the right distribution… no matter what it’s name is.
That keeps it off. Fragmentation.
The not-so-funny reality is that we _do_ develop (GUI) software for our Linux customer base and I have to admit it’s a pain in the butt.
One is right when saying that there are defacto standards, but simple facts aside, reality is that there are clients that only want KDE and some only run Gnome; we cannot tell someone to start running KDE if they don’t want to. Unless you want to force your customers a GUI down the throat, I admit you can be quite happily using Qt only.
The fact I pointed out is that there is no GUI unity / consistency or even a “roadmap” in Linuxland. This all boils down to the fact that if you offer your software based on the current demands, you might run into trouble later, because today’s defacto KDE might be tomorrow’s defacto Gnome 6.345 or whatever.
I never mentioned usability guidelines, but this would be a next step. Maybe some developers could care less, but in the end of the day it’s the user that stares at that screen everyday, so it better be consistent.
Thats two diferent things!
The first means how easy it is something for use…
The second means how should a branded software work. Meaning specially that all software will be whrestling with the idiossincreties of the writes of the said guide… Which normally don’t have much clues on how a software is built, but which are very much worried about the protection if the brand (meaning they aren’t usability consultants, programmers or anything in between, but marketing or legal workers).
Usability is an universal thing, not related with any environement or API. And AFAIK IMHO is relegated to a very low priority in the majority of the software.
Cheers…
You’re bitching about a nonexistant problem (a real shocker, I know) – http://www.freedesktop.org/
A kernel on top of a console on top of an X manager on top of a pretty graphical shell does not make a usable OS. It does make it plenty slow and plenty complicated (try and install a random RPM for a small taste). Call me back when you supply the end user with a SETUP.EXE
most people couldn’t do a clean install of windows by themselves, so i am talking about larger corporations here where the it department manages all the computers.
the biggest reason people aren’t moving is momentum. people have been using windows and ms for so long that no one wants to foot the training bill. even if there was an equivelant program for linux that met all the requirements to get the job done, what they are using right now works.
the next reason would have to be entanglement. most companies have been using proprietary file formats for long enough that they couldn’t continue their business without being able to continue reading and writing them. the only way out would be to pay big bucks to convert data, and the benefit of switching is not big enough to support that.
I’m not going with Herb here, I’m merely going to use his argument to make a point to those of use who read osnews and want GNU/Linux to be a force on the desktop.
Be careful what you wish for. Many people want to turn the OS into a system that is that easy (setup.exe). This is a nice idea. And people can argue that ease doesn’t mean you have to sacrifice configurability. The reality is you will. Try SuSE and then try Slackware. You’ll see what I mean. SuSE is more fool proof and will automount your window shares, but try telling me out of the box what network services our running without looking.
That is nice for some people, but it’s not what I want. I’d rather use something like Slackware or Gentoo. That’s the benefit of a fragmented community. Control. Lose that, and the only difference between GNU/Linux and windows is how you feel about the people backing your OS.
1. fractured approaches to the desktop
Pool resources and focus into half the “Lets conquer the desktop” projects and I’m sure we’d have something to be proud of in no time. Choice is good, but too many choices means nobody gets anything done because they’re still trying to decide what to use. There are clear choices for server tools, httpd? Apache. We’re in need of one big obvious choice that everyone can accept as a “standard”.
I don’t think you got the original poster’s point (or I read the post differently or whatever). FreeDesktop was built to allow the interaction of apps written for different desktop environments. This could potentially make the desktops look consistent to the user.
However, AFAICS the point was that Linux is horribly fragment from a _developer’s_ point of view, and that remains true. There are hundreds of libraries floating around, and there isn’t really The One Place you can look for APIs. You know, something like the MSDN Library.
Plus, there’s still the problem that fragmented desktop environments are technologically inferior to a single, standardized one. A fragmented desktop environment will never be as efficient as a single, standardized one, and if it’s only because you need to load more than one set of shared libraries into memory.
“We’re in need of one big obvious choice that everyone can accept as a “standard”
———————
Oh, wait that sounds like.. hmm. WINDOWS!
What is Linux capable of that other operating systems cannot do?
Is there a main “selling” point other than it’s free?
What can the average user do with Linux that he cannot do with other operating systems? Please notice that i mentioned AVERAGE.
Linux is simply turning into a free Windows clone. Why WINE if you are not?
Hi
I’m a developer myself but I’m also a Linux user.
Linux won’t success on the desktop. Not in the form it has now.
Why ?
Simply because it suffers many simple problems that could be very easily fixed but won’t because some developers don’t want to.
An example:
What sound daemon should you use ? OSS ? ESD ? ARTS ? something else ?
If someone uses KDE, he’ll have ARTS running and programs using ESD or OSS will fail to make sound. If someone uses GNOME, maybe he’ll have ESD launched, maybe not. Anyway, ARTS apps won’t work. Imagine explaining your little sister that she needs to go to xmms->options and select the right sound daemon…
Linux desktop needs ONE sound daemon (not OSS as it doesn’t have mixing capabilities), ONE widget set (GTK or QT), one file dialog, one printer system, one print dialog, one configuration place, a way to choose an internet browser and to use it from any app that displays links, a way to choose one email program and have it open itself when someone clicks a mailto:// link from any browser, one and only one packaging system that always work, on any ditribution (debian apt-get, RPM apt-egt or URPMI) etc…
Linux doesn’t LACK anything to succeed on the desktop, it just provides too much incompatile choices.
I don’t care there are 3 or 4 great video players because users can choose the best one, but I care when there are multiple underlying and incompatible libs because users don’t care at all about that choice and it makes them run into unexpected problems.
I know it will never happen but I like dreaming anyway.
Maybe in 10 years…
The question is: What’s keeping Linux off of desktops. – not what is prefectly right
Linux is for power users who can dig into code. They want it clean and they want full control of their systems.. that ain’t gonna fly on the desktop. Not everyone can code and not everyone wants full control. Some people just want to USE their system. I don’t see these two worlds coming together anytime soon.
… yeah … thats all i am waiting for. the ability to run photoshop7 natively under linux.
If you are a programmer you know that Borland have the best solution to developping software to desktop and servers linux.
Yes, Kylix is a good programmers tool to develop software to LINUX. If you don’t already try it go to Borland website and download it. Borland also have an open-source version of KYLIX.
The most interesting thing that borland is doing is making their tools multi-plataform:)
get games on linux == get more software and support…
so what to do, try and make linux a better gaming platform,
then the desktop will follow…
Some game company’s have tried and they left us with some
great stuff, like SDL for example…
The biggest thing I hear from the Linux community is “choice”. You always have a choice to run whatever you want however you want to run it. That’s the problem with getting more mainstream products onto the Linux platform: too much choice. It seems some users forget that developers don’t want to spread themselves thin on multiple platforms, especially if they are not going to get any return on investment. There’s another reason: the push for free (as in $0) software. The Linux platform isn’t going to get the respect it wants if the big developers can’t make any money on it as there seems to them that Linux users don’t want to pony up the dough. Laugh or sneer if you want but after Sun announced that Star Office 6 (which garnered great reviews) was going to sold instead of given away, many a Linux user posted on boards protesting and going to Open Office “because it’s free”. Then if they COULD make money on a product, which distro do they write it for? Which windows set: KDE? Gnome? All? Just One? Until the number of popular distros are widdled down to a couple and users start showing the greenbacks, THEN we’ll see the big boys start bringing the good stuff to Linux.
Me? I’ve got choice: I still use Windows 98SE and Office 97 at home. Got it patched, protected, and runs everything I want.
What is keeping Linux of the desktop? The fact that many people have fully working WindowsXP desktops and are quite happy with them.
Hmm
Your comment is typic.
Windows is the model linux has to follow on the desktop. Why ?
Simply because it’s the best from a conceptual point but not the best in both the philosophical ay and the technical way.
Linux’ goal IMO is to provide at least what Windows already provides but using a free license (free as in free software) and a superior technology.
The lack of a standard desktop on linux is a bad bad point because it makes it harder for users to use it and for developers to create applications. So I don’t see why it’s so good for you.
> Linux doesn’t LACK anything to succeed on the desktop,
> it just provides too much incompatile choices.
There we are.
I am a Software Engineer, I have used (and administrated) AmigaOS, Win98, WinXP and – Linux. I am no dummy user.
Of all those, Linux is the most difficult to administrate, because of *choices*, because of the very flexibility Linux advocates boast about.
If sound doesn’t work on Windows, it’s either the hardware, the driver (see System Manager), or the settings (double-click that speaker icon). With Linux, who knows?
Installing SuSE or RedHat with “standard” options gives me half a dozen MP3 players installed. What the heck…?
Linux is, so they say, trying to become “as simple as Windows”. Dudes, that doesn’t suffice. I have the “easiness” of Windows already on my system, and it drives me nuts. To beat MS’ ass, you have to be easier than Windows – much easier.
In the meantime, kudos to RedHat for http://www.cygwin.com – because with Cygwin under Windows, I have all the functionality I would want in a Linux system without having to cope with that ever-changing, ever-morphing choice monster.
The bad part about it is, I am still a Windows user, although I hate MS’ guts. Thing is, Linux is simply no alternative for me, and won’t be for years to come unless it finally starts to settle, and they start to realize that choice and simplicity doesn’t go together.
(And *please*, don’t give me that “you didn’t get the right distro” stuff. That’s another choice that’s scaring other people away, and there hasn’t been a lack of different distros being tried on my hardware. The only Linux related thing that installed properly so far was Cygwin…)
First of all…
> Linux desktop needs ONE sound daemon (not OSS as it doesn’t have mixing capabilities).
OSS is included in 2.5.x dev kernel thus will be in 2.6 when out so I must assume that some people think it’s good enough as a standard of some sort, I have no idea about mixing capabilities though but OSS works OK as standard in my eyes atleast.
Linux aren’t dominating the desktops because it will always be compared to other OS desktops and no matter how funky it looks (bluecurve does look smooth) it will always be sub-par on older machines (around 500Mhz) cause of the slow redrawing native to XFree.
Put X on a P4 3GHZ w/ HT and it may even outperform or atleast equal performance to others, but I won’t be having such a system in a looong time and can only dream back to the days of fluid desktop motions on even a old S3 Virge card on my old BeOS box.
> > “We’re in need of one big obvious choice that
> > everyone can accept as a “standard”
>
> Oh, wait that sounds like.. hmm. WINDOWS!
YES! Exactly! And that is “what’s keeping Linux from the desktop”!
Face it, the OS is a *tool*, not a goal.
People that install an OS don’t want to choose between KDE and Gnome, they want an out-of-the-box install that gives them an intuitive desktop with *one* word processor, *one* mail proggy and *one* way to explore their file system.
I’d say, create a standard install accepted by *all* distributions, and keep the choices for later – but…
…users expect the system next door to work more or less the same.
(How about having a *standard* install that can be customized later, but that’s always available as a one-click fallback?)
On an open system you cannot avoid choice. If someone doesn’t like a sound daemon, he is free to write another one. That’s the way an open system works. In the end, thingg will settle down and then you have one or maybe two standards that everybody uses.
There are dozens of kernels, but Linux won. There was a dozen of window systems, but X11 won. There are dozens of toolkits, but Qt and GTK won. There are hundreds of webservers, but Apache won.
There is a consolidation, but give it some more time. In 5-10 years there is a good chance that either KDE, or Gnome, or maybe even both are obsolete. But as long as not even one of them is mature or complete, you can not expect that the market decides for one of them.
BTW you shouldn’t say something like “Linux needs one desktop environment”. Both KDE and Gnome or not Linux specific, and there is no Linux beside the kernel. Who says that you are not allowed to use more than one desktop on a kernel? You don’t say that MacOS and Windows should be merged to stop the fragmentation, so why should KDE and Gnome?
I am a Software Engineer, I have used (and administrated) AmigaOS, Win98, WinXP and – Linux. I am no dummy user.
If sound doesn’t work on Windows, it’s either the hardware, the driver (see System Manager), or the settings (double-click that speaker icon). With Linux, who knows?
If you don’t know how sound works in Linux then you don’t know enought about it to say you have administrated it. I would say that this comes from using those newbie distros like Redhat or SUSE. If you had learned Slackware it wouldn’t be an issue because you would understand what is going on in your box instead of letting the GUI do all the work for you.
In fact, you’re partially right.
I mean, of course Linux is just a kernel and nothing prevents you from creating any OS on top of it.
But the problem is that OSes like RedHat, Debian, Mandrake, SuSE etc… are not consistent OSes as they come with 2 destops. I think that, said like that, it’s ore clear. In fact, it’s distributions that should make choices. RedHat should provide GNOME only, Mandrake and SuSE, KDE only etc… Then, you’d have software for RedHat, software for SuSE, drivers for RedHat, drivers for SuSE etc… but the problem is that, as long as KDE and GNOME will both work on XFree together, people will want both because they think tehy are compatible when they are just GRAPHICALLY compatible. Their sound daemons aren’t compatible, their drag and drop is hardly compatible, etc… and the saddest part is that apps like OpenOffice, Real Player or Mozilla are fully compatible with KDE and GNOME neither…
I really don’t know what the solution is. Maybe one idea:
As XFree is the standard graphical interface on Linux, create an XFree widget toolkit and port GTK and QT to it. Then, KDE and GNOME apps would always look the same. Then, standardize ALL apps on a sound mixer (Maybe ALSA which will be in the next kernels and which supports hardware mixing). Finally, create a file dialog and a printer dialog that EVERY app and widget toolkit will use.
That would be huge step and it’s not even hard from the technology point of view but developers have to change their mind, radically.
Isn’t that amazing that OpenOffice is to be ported NATIVALLY to MacOSX while Linux users have to support those alien widgets ?
//If you had learned Slackware it wouldn’t be an issue because you would understand what is going on in your box instead of letting the GUI do all the work for you.//
Yet another thing keeping Linux off the desktop: arrogant techie dweebs who think everyone else is an idiot.
Typical Penguinista.
The only thing keeping Linux off the desktop is all the people writing articles and arguing about what’s keeping Linux off the desktop!
Come on -How many articles can be written about the same thing? We’ve been reading this same diatribe for the last 2-4 years now it seems.
Perhaps if everyone who loves to argue and write about what’s keeping Linux from becoming a desktop OS actually sat down and did just a little bit of what they’re arguing about, it’d happen.
I’ve said it before and I’ll say it again -I personally believe that the Linux community as a whole is what’s to blame for its lack of inroads into the desktop.
Too many people pointing fingers and saying “This is what’s wrong”, rather than sitting down and changing it.
Too many people saying “RTFM” rather than making Linux easier to understand and work with.
Too many people belittling people who point out Linux’s inadequacies rather than working to turn these shortcomings into strengths.
And to many people who are hapy with the status quo, and don’t want or need an easier to use OS. They like typing in a console, and editing their config files. And that’s cool, but that’s not what people want in a desktop OS.
Basically, there’s just too many people talking about the problems rather than cohesively working on a solution.
For some people it’s already desktop ready, and there’s others who don’t want to see it homogenized into a consumer-ready system (read as “Joe Computer User”, not “Eugene Imageek”). That’s cool and all, but in order for Linux to branch out from the Ubergeek world of servers and clusters, it’s gotta happen.
If it’s not going to, then everyone should realize this and stop writing the same types of articles about the issue! Man… Even Slashdot’s finally getting over its “Linux on the Desktop or bust” mentality.
It’s either your desktop OS now, or it’s not likely to ever be with the current players who are writing software for it.
Conversely, if Apple or another “new” company came into the mix and offered a Window Manager which was able to embody the simplicity of OSX over the foundation of Linux, they’d win the ballgame hands down. KDE, XFCE, and Gnome would be officially relegated to “Geek” WM’s, while this new WM would instantly become the consumer friendly one which all distributions would default to.
I’m really surprised someone hasn’t done this yet actually -KDE’s very nice, but it’s not quite there yet for a desktop OS, and while I love Nautilus, Gnome as a whole leaves me kinda wanting. XFCE is very cool and easy, but without a taskbar (which I’m not sure if XFCE4 will bring this to the table or not!) it’s not going to win mass acceptance as a desktop OS.
Either way… Get over it. Linux needs changes now (yesterday even!) if it’s going to get into the desktop ballgame. Until then you’re all writing different articles about the same things. We’ve seen it all before -You’re just writing from a different angle about the same tired issue.
some very good posts here. momentum and IBM – PC – Windows = a logical chain of events do explain quite a bit.
You need some spark/energy to cause a change. Right now, its MS’s licensing practices. I doubt that will get better. Linux might miss out on some of those opportunities now due to lack of applications but the apps are coming and so is acceptance of linux.
It is always difficult to unseed a dominant market force but its starting to happen. It going to take more time but give it 10 years and linux will probably outnumber MS in the corporate desktop. Now as i always add, i wish apple would address some of those opportunities as well.
1) Linux is too hard to use
2) X-Windows/KDE/Gnome are too slow
3) Linux is too hard to use
These are the facts of the matter.
The same neurological characteristics which alow a person to enjoy programming also usually renders them utterly unable to design software from a human-usability point of view. I’ve got no faith it interfaces & desktop-systems designed by programmers (well most of them anyway, some of them ‘get it’ though).
Solar:
“In the meantime, kudos to RedHat for http://www.cygwin.com – because with Cygwin under Windows, I have all the functionality I would want in a Linux system without having to cope with that ever-changing, ever-morphing choice monster.”
RedHat didn’t write Cygwin; they inherited it from Cyguns Support when they bought them out. Agreed that it is a nice set of tools.
Right now the best alternative to Windows on the desktop is Mac. As far as Linux getting a foothold on the desktop major coroprations should be the first target. Companies should sell Linux as a turn-key solution; sell the PC, software, and support for $X/year/desktop. When $X is much less than they currently pay to support their Windows PC’s they will consider Linux. For home users, Linux should be sold as an internet appliance. It should come pre-installed with a simple interface, ISP service, and simple software to do typical things like surf the net and email.
For me personally, I am a very application-centric. I can work through almost any OS (including DOS and Win 3.1) if I have the proper apps.
From this point of view, there is functionality that either can’t be duplicated at all under Linux or can’t be duplicated well. And until that changes, I will not be switching.
The fact is that there is still a need for Windows, because there are just somet things in Win32 that can’t be done (yet) with Linux. I know some of the zealots just hate to hear that, but it’s true.
I don’t know if I’ve ever said this before, but if someone can find me functional equivalents to the following applications:
* MS Streets & Trips 2003
* Propellerheads Reason
* Cool Edit Pro 2.
.. I will switch this weekend. Just these three applications … that’s ALL you have to do. If you say that Linux is the end-all solution and there is no more need for MS or Windows, then put your money where your mouth is!
The one thing that is keeping people from switching to Linux or any other alternative OS is the fact that they have no reason to switch.
There is no one overwhelming reason or app or advantage that is worth the trouble of installing any alternative OS.
Sure we can discuss what would make linux a better desktop OS. But, if you want to talk about what is preventing people from switching droves here it is.
I work everyday in Unix-centric world doing sys admin work for unix programmers. I live and breath Solaris and linux everyday. I use linux at home and at work. I have the reason. You want cheap reliable web servers you don’t have to pay license fees for. Try Linux or BSD. You have a reason on the server side. Are you a Unix programmer, sys admin or network admin who pines for the power of Unix without all the cygwin nonsense that never seems to fit right in a NT world. Try Linux. You have the reason.
Do you have a reason?
The linux world has to stop merely copying other OSes and come up with some OS advantage or application that makes switching worth the trouble. It is possible.
People seem to forget the nature of GNULinux, which is that it is Liscensed under the GPL and free to use and modify. As such, if someone doesn’t like an idea, they can fork it. This leads to change and competition, both of which are good things.
However, Linux is never ever going to be exactly the way you want it to be unless YOU or a Linux distibution company decides to make it that way. We’ve seen many efforts by Lycoris, Xandros, Lindows, etc. to unify the desktop and present the user with only one of everything.
As for the developer’s perspective, each commercial GNULinux distribution is creating its own operating system for all practical purposes, including its own, standard (depending on the distribution.) set of drivers and apps. Saying that every Linux distribution should have the exact same toolkits is like saying that all commercial BSD based distributions should have the exact same toolkit because they’re based on BSD. It’s not a reasonable expectation to expect distributions to standardize because they are trying to set themselves apart from the competition, not look more like it.
As for the graphical toolkit, I would agree that you are right to say that this needs to be standardized, but it is each individual distribution that needs to pick one toolkit, not GNULinux in general. Xandros, Lindows, etc. all got it right when they chose to only pick one desktop environment. In my opinion, Redhat should go solely with Gnome (since that seems their inclination) and Suse should go with KDE. And then you pick a platform to write on, whether it be redhat or Suse or whatever else. It should not be expected that any app should automatically work on BOTH redhat and suse, because in reality they are not the same OS, it’s like saying that something should work on both IRIX and HP-UX because they are both BSD derivatives.
In conclusion, yes some of the big guns need to pick a preference as to GUI toolkit and sounds daemon, but once this happens (Which I believe it will sooner than most think as Lindows, Xandros sell better and better encouraging other distros to follow their lead.), it’s up to developers to pick a platform.
take care,
Red
X11R6 is slow.
>> I am a Software Engineer, I have used
>> (and administrated)AmigaOS, Win98, WinXP
>> and – Linux. I am no dummy user.
>> If sound doesn’t work on Windows, it’s either
>> the hardware, the driver (see System Manager)
>> , or the settings (double-click that speaker
>> icon). With Linux, who knows?
> If you don’t know how sound works in Linux
> then you don’t know enought about it to say
> you have administrated it. I would say that
> this comes from using those newbie distros
> like Redhat or SUSE. If you had learned
> Slackware it wouldn’t be an issue because
> you would understand what is going on in your
> box instead of letting the GUI do all the
> work for you
But that’s just the point! An “average computer user” should NOT have to fully understand the complex inner workings of Linux in order to use it and perform routine maintinance of his system. The average user should not have to take a university course in computer engineering in order to use his Linux desktop with the same level of ease that he uses his Winows desktop. Say what you will – the thing that scares most people off is the total lack of consistency in the Linux world.
I’m an “above average computer user” myself. I started tinkering with 6502 assembler on a Commodore64 back in the mid 80’s. Did a spot of Lattice C on my Amiga in the mid 90’s and got myself a PC in the late 90’s and played around with Delphi. (there’s a point coming, bear with me)
For about a year now I have been flirting with Linux. And as much as I hate Micro$oft I still don’t see Linux as a viable option for a full time desktop. There’s just too much “messing around in the guts of the machine” for me to REALLY bother. It’s Gnome/KDE, it’s dependency-hell, it’s “makefile this, makefile that” it’s “root/bin/lib/x11/slib/user/crap/whatever” ad nauseum.
My point is this: If a guy like me, who has about 17 years worth of computer tinkering experience, finds Linux muddled and confusing how the h*ll do you expect the average boob to ‘not’ cry in his beer a few weeks after trying to get something as simple as NetHack to install and work properly on his Linux system?
Diversity is good, but we don’t have diversity here – we have complete and utter fragmentation. Divide et impera. I want standardisation. I want compability. Essentially I want “KNOME”. And THAT, my friend, is what’s called “usability”.
…just my 2 cents worth…
For about a year now I have been flirting with Linux. And as much as I hate Micro$oft I still don’t see Linux as a viable option for a full time desktop. There’s just too much “messing around in the guts of the machine” for me to REALLY bother. It’s Gnome/KDE, it’s dependency-hell, it’s “makefile this, makefile that” it’s “root/bin/lib/x11/slib/user/crap/whatever” ad nauseum.
Amen! Down with Linux!! And down with Micro$oft too!! Somebody buy me a Mac
lol
its X
we need a new gui more like quartz in os x but fast but add X layer compatability for old apps
we need vise style installers for everything and afix to the dep hell we all face when we first get into linux
and make it hard for normal home desktop users to do development stuff they dont know about before they break their systems
http://www2.fresco.org/
I hope it’s the future !
As some here might know, I am a longrime Slackware user. I flirted with many other distros, including Debian, but I kept going back to Slack. I even had a lab machine at work with Slackware. Well, one day I decided I wanted to upgrade KDE on that computer. I thought I’m a big boy now, I don’t need to upgrade to the new Slackware just to get the new KDE. So, I decided I will:
– check the packages the new KDE depends on
– remove the old versions of those packages
– install new version of those packages
– remove old KDE packages
– install new KDE packages
Well, halfways through this process I learned that the new KDE needs the new Xfree86! “f* f* F*!” I thought, why do these opensource programmers always have to use the newest and latest? Is nobody out there thinking about backwards compatibility? Is that little new feature really worth the swithch to an entirely new Xfree86? Oh, nevermind… so I tried to do the same with Xfree86 as I planned with KDE, only Xfree86 depended on a lot more libraries than I imagined, and I didn’t have the packages for all those libraries, no matter how hard I searched the Internet.
Oh well, so I decided to extract those libs from an RPM package and put them where they belong, manually. Now, when I installed the new Xfree86, the X doesn’t want to start at all. Reason? The xinit has changed. OK, so dive into the whole script chain that gets the X up. Then I tested X (you can use X even without any window managers, just to display a remote desktop, with “X – :0 -query hostname -once”) and as it worked, I proceeded to install KDE. KDE never worked on that computer, but I jst ran out of patience, and never finished investigating the reason. I decided I will install ICEwm instead. Which I did, but I got, as a result, a horribly slow ICEwm, something horrendous.
That was the last time I was adventurous on Linux. My attitude is now 100% “if it ain’t broke, don’t fix it”. Kinda like Windoze.
Slackware is still my fav distro, though, because I have experienced much more often problems with RedHat than with Slack. But this was NOT my first or last problem with dependancies in Linux. The most recent one was with installing Freeciv (www.freeciv.org), but that’s another, gory and R-rated story.
I work for a company that developes windows software for doctors offices. I have tried in the past to create an intreset in linux. The development tools we use are avaliable and most of the office buy new hardware before we install our system. The one things that I have noticed is there is a stigma concerning any *nix system. Alot of the office are moving from SCO or some other unix system to ours for several reason:
1. They use windows at home.
2. They view UNIX as a legacy system (Dumb terminals, main frames ).
3. They think the administrative cost of a unix system are to high.
…
I am sure there are some other reasons also.
Linux needs to make a break from UNIX. I am not saying that any of the POSIX stuff needs to go. What I am saying is that it needs to make a break from xterm, xclock, CDE, …
That is one thing that Mac OSX has going for it, that abstract layer that keeps *NIX from feeling like *NIX.
I have to agree w/ Julien on a his first post.
One problem is the fact that there are too many choices for different daemons be it the printer, the sound, etc. Overly, the problem is with a standardized GUI though I’m not blaming KDE, GNOME, etc in this case. When saving a document (and this prob. depends on the toolkit used, I don’t know) you get confusing interfaces to choose where to save files and they’re not the same in every app. Toolbar menus in each app are much different.
Apple published a 60+ pp pdf with their GUI requirements which are by far the strictest I’ve seen but it greatly benefits the end user. While reading it, I recognised many places where MS could improve upon.
For example – Do you want to save this program?
On the MAC its usally Save – Don’t Save – Exit program
In Windows its usually Yes – No – Exit
Less room for error if you ask me.
Must really burn you guys that BeOS *still* kicks Linux butt.
I mean, how many iterations of KDE and Gnome does it take to get to a consistent, speedy, useable interface that is well integrated into the OS? Do these developers have any honor? ;o))))))
It’s true, BeOS just works.
I don’t see having different desktops on one OS a problem, except when the desktops don’t look/act in a consistant manner across distros. If KDE/Gnome in distro X is drastically different than in distro y, you have a problem. Best case senario going this route is that one distro is going to come up with a killer desktop configuration that will leave other distros in the dust, and everyone ‘standardizes’ on one distro (which will probably happen if/when Linux goes totally commercial). Worst case senario is that this distro ‘tweaking’ is going to create huge incompatability issues amoung distros, and I think that’s already starting to happen. Take the file manager in Xandrose for example … does this thing work in other distros? Can you make it work?
3. People are … stupid. Hmm… Sorry. I don’t meen
_all_ the people around. I mean only those who do not
belong to the above two groups but are still running
windoze. Let’s call them windiots.
Yup, and you just proved yourself a moron by judging intelligence of others because they have different opinion. Once again the Linux geek charisma is sinking, gratz.
“If you are a programmer you know that Borland have the best solution to developping software to desktop and servers linux. ”
True, Kylix is available for developing Linux apps, but it is only officially supported on a handful of Linux distributions; none of which I personally use.
I’ve been developing Windows apps with Borland’s DElphi product for 6 years and would like to use Kylix but it does not work correctly on MY chosen Linux distribution – Slackware.
Iggy Drougge: Only IBM had the power to give Microsoft the power they have.
True true. But they didn’t give it willingly, Microsoft power came from IBM’s mistake and Microsoft ability to maximize on it. But from your analogy, are you implying that only Windows clones that unseat Microsoft?
Iggy Drougge: So do HP, DEC and SGI
HP and DEC are like IBM. Their usage of Windows is for machines for a totally different market. Their competitiveness against Microsoft is for the server market. SGI on the other hand stopped selling Windows NT workstations. It wasn’t profitable. It was a stupid business decision.
Kalimotxo: Do I use Gnome? Do I use KDE? Motif? Athena Widgetset anyone?
I would say either Qt or GTK, whichever suits you best. However, if you app is multiplatform, you can’t go wrong with Qt. If you want a low-cost port of a Windows app, try Wine.
Kalimotxo: Do I use FreeType? Or maybe the built-in X fonts?
FreeType is used by XFonts and XFT2, and not a competitor to it.
mX: 2. People have windoze-based hardware (winmodems, win-scanners, etc.) that won’t work with Linux.
Winhardware is only used in the consumer market. The article is about the ENTERPRISE.
mX: 3. People are … stupid. Hmm… Sorry. I don’t meen
_all_ the people around. I mean only those who do not
belong to the above two groups but are still running
windoze. Let’s call them windiots.
1) Stupid people aren’t the ones switching. Trained sysadmins would be doing it. Remember, the enterprise?
2) Windows users use Windows not because of the applications don’t really see any reason to waste time and effort switching to Linux for little to no rewards.
3) Those who do depend on Windows-only (or Windows/Mac-only) applications don’t really have much choice. Between Productivity and Inproductivity, most would pick the prior.
In any case, Windows users aren’t windiots.
Anonymous: Linux, with the noteworthy exception of Red Hat Linux 8 hasn’t a clean, consistent interface. No fussing between KDE and GNOME, unification and standardization, as another poster mentioned is the key.
If this was the only reason, Linux would be popular on the desktop right now. It may make Eugenia scream in disgust (no offence again), but frankly it isn’t the thing keeping people from using Linux. Read above.
Charles E Hardwidge: Useability, consistency, and standardisation.
If the first two was the only thing keeping back Linux, I wonder how Windows got the hold of the market when their product was neither consistent nor usable at that time. Maybe standardization.
Matthew Gardiner: One only needs to have a unified theme and voila, no problems.
It takes much more than a unified theme to make them feel and act the same. On Red Hat 8.0 for example, drag and drop behaviour between GTK+ and KDE apps is very different.
Matthew Gardiner: How many programmers do you know who stick to the usability specifications that Apple or Microsoft publish?
Many. Most OS X apps follow the Aqua HI Guide, for example. Microsoft’s HI guidelines is very, very basic, it isn’t as strict as Aqua, GNOME nor KDE guidelines, hence the inconsistency between applications.
Matthew Gardiner: Btw. in most cases, commercial developers stick with the KDE usability guidelines.
I know a few ISVs that have apps following KDE’s HI guidelines. The biggest two ISV for Linux desktop, Netscape and Sun, don’t have anything that looks and/or feels like a KDE app.
H-kon: Every time Linux users tell me that it’s ready for the desktop, i never use the right distribution… no matter what it’s name is.
If each distribution merge into one product, it wouldn’t take Linux closer to the desktop, in fact may even be hindering it. Each and every distribution takes a different approach to the desktop. Try out each one, see which you like best. If you don’t like any, Linux isn’t ready for you yet.
One is right when saying that there are defacto standards, but simple facts aside, reality is that there are clients that only want KDE and some only run Gnome; we cannot tell someone to start running KDE if they don’t want to. Unless you want to force your customers a GUI down the throat, I admit you can be quite happily using Qt only.
1) I really want to know what app you make for Linux. I really do.
2) Having a Qt app doesn’t mean GNOME users can’t use it. I’m using GNOME 2 right now as my main Linux desktop on Red Hat 8.0. Yet I use Opera, KOffice (specifically KWord and KSpread), Quanta, Kate and KDevelop Gideon. I don’t see any problem. When I switch to KDE 3.1 when the final version is released, I don’t see any forseable problem using GTK+ apps I grew fond of like XChat, xmms, Evolution, etc.
Kalimotxo: The fact I pointed out is that there is no GUI unity / consistency or even a “roadmap” in Linuxland.
Quite on the contrary. Red Hat managed to make unified themes for both environments, others are bound to follow.
Herbert Gone: It does make it plenty slow and plenty complicated (try and install a random RPM for a small taste). Call me back when you supply the end user with a SETUP.EXE
The way Linux is designed isn’t why Linux installations are hard. You probably won’t find setup.exe, but for stuff like StarOffice, there are .bin files that have a nice wizard ala Windows.
Plus, installing RPMs is pretty easy. On Mandrake, open it in either Nautilus or Konqueror, a wizard comes up which would get dependancies from the Net for you. Red Hat, I heard, could do that, but require Red Hat Network (which I would get as soon as I get DSL working on Linux).
Besides, since when installation wizards mattered for the enterprise? Call me when it does.
Nicolai: Choice is good, but too many choices means nobody gets anything done because they’re still trying to decide what to use.
I think the enterprise couldn’t care less about the desktop. They would use whatever their consultants/distributions recommend.
H-kon: What is Linux capable of that other operating systems cannot do?
Terminals connecting to centralized servers? Could you hook up 200 terminals using OS X or Windows? Largo could only do it with Linux and XFree86.
What can the average user do with Linux that he cannot do with other operating systems?
Please note that the article is about the fucking ENTERPRISE market. NOT THE CONSUMER MARKET.
If someone uses KDE, he’ll have ARTS running and programs using ESD or OSS will fail to make sound.
Little sisters aren’t the main concern by the article. In the enterprise, sysadmins would have already selected the sound daemon on xmms before the users knowing the problem existence.
Linux desktop needs […]
I notice everything you said would only hold true for the consumer market. For the enterprise market, there are sysadmins, therefore these “incompatible choices” aren’t a big deal.
Linux is for power users who can dig into code.
I use Linux a lot, yet I can’t code more than Python, HTML and XML. But Python, HTML and XML doesn’t count (it took me a day to learn each). Yet I’m a Linux power user.
Quazion: so what to do, try and make linux a better gaming platform,
then the desktop will follow…
Gaming is completely different from the enterprise. Having all the hardcore gamers using Linux wouldn’t make Walmart offices switch from Windows to Linux.
Solar: Installing SuSE or RedHat with “standard” options gives me half a dozen MP3 players installed.
I don’t know about SuSE, but Red Hat only installs XMMS by default. Which doesn’t play MP3s out of the box, BTW.
Cyco: it will always be sub-par on older machines (around 500Mhz) cause of the slow redrawing native to XFree.
I got XFree86 4.2 and XFce 3.8.18 working on Mandrake 8.2 on a 120MHz Pentium with 32mb of RAM. More responsive than Windows 95, IMHO. (Of course, that machines still boots BeOS by default).
The problem is that Slackware has in no form a proper package manager like Debian or Red Hat. On Debian for example, if the debs is available, all I would need to do is apt-get kde<version>. It would install the new version of XFree86 as a dependancy. It is similar on Mandrake and on other RPM systems using apt-get for RPM.
The problem is with Slackware. So your whole post is implying that Linux is bad because of it, but it isn’t. BTW, there is a version of RPM for Slackware, you could have used that.
NOT because of the killer applications like Office (which the opensource community is only now starting to address) – it is all the small properiotor software that runs on thousands of corporate workstations in the world.
Unless there is a way for emulation of Windows Apps (at no cost) to linux with little to no noticeable speed impacts – conversion will never change.
With this in mind and linux not having a solution – Linux needs to target the new startup companies so that they START with linux and not the evil trap of Microsoft.
Also, I wont to stop the post before it happens. VMWARE is NOT a solution to this. It IS a great product but it won’t address this problem for 2 reasons:
(1) Why run windows apps under vmware emulation of windows when you could just run windows
(2) We need a Windows emulator that doesn’t require installing windows – so that we can install the win .exe files natively on linux. This is a huge problem because of such things like Win32 and other libraries.
riiiight…. like I never had dependancy problems with RedHat (or with RPM for that matter).
I have to say that I was startled by the fact that I actually agreed with all the points in your post 54. I thought hell froze over or the sky is falling ;o)
Good that you scolded me for my choice of distro, without even looking at real problems that Linux may have. Always the good ol’ rajan r, after all.
In my opinion it is the easy of installation of the modem what´s keeping home user (me for instance) on windows. Yesterday I downloaded and installed Mandrake 9.0
Everything is installed OK except, of course, my Alcatel USB ADSL modem. I use my PC mostly for Office needs and OpenOffice 1.0.1 is good enough, I haven´t tried KOffice) and surfing the internet. When my modem isn´t recognized I run back to Windows where all is working well. On the internet there is a site which explains how to install it but it is 5 pages editing files ect. To much an effort. I know it is lazy but Windows just is easier to use and most user just want an easy to use system.
When Linux makes it easy (including the installation of the modem) it makes a change.
Greetings,
Jeroen Roeterd
Two of the biggest problems with getting Joe the point-and-click computer user to use linux:
1) No standard central GUI for choosing settings of the system, daemons, UI, and programs.
2) No standard install format that works all the time with all distros.
It’s possible to create solutions to these, but it would require a)Support from all distros. b)A linux standard for the file layout. Over the last few years the hurdles of creating a slick GUI, and providing an easy to use and understand install system. Even the problems with fonts and printing are just now getting wrapped up. But until these afore mentioned problems are solved, Joe I-like-windows-because-I-don’t-have-to-think is not going to touch linux.
No it is not slackware, however, from the story above it does not sound like you were using all stock slackware packages from the source — Slackware.
Listen, if you stick with Slackware packages from say Slackware itself and maybe one other trusted source then you would be fine.
Let me tell you a story. There was this guy who installed Redhat and he was just trying it out. He used rpms from Redhat Network and freshrpms. He did not run into dependency hell because he did not download random rpms from all sorts of places (not saying you did — you will see the point in a second) or try to install lots of things from source. He used apt-get and RHN and he was fine.
Then there was this other guy. ME. I installed rpms from RHN and mozilla.org and freshrpm and Nyquist’s site for gtk2/gnome2 apps and then I mixed in some Havoc’s Gnomehide site. I had dependency issues and I expected it. These were not vendor support 3rdparty apps. These were not apps from my OS vendor. These were apps and packages gotten from all corners of the web and I had dependency issues and I dealt with it.
I ran into dll hell on my NT box when I started downloading freeware and shareware from all over the places and ending up having to install one thing to get another thing and it happens with free software. Get over it.
Part of the problem with these debates is that most posters are from America. Linux is doing much better on the desktop in areas like the 3rd world and Europe. For example the article’s assertion, “But many efforts to migrate have faltered because of a lack of collaboration, calendaring, and scheduling software for Linux”. That’s actually being addressed by the German government for KDE. I’d expect the American corporate market to be one of the last to fall:
1) Microsoft is an American corporation and thus Microsoft understands the needs of American corporations very naturally.
2) The target market for Microsoft products (until this year where Microsoft has changed focus) has been American corporations and thus their products represent a “best fit”.
Now stard addressing internatiional issues:
How many applications work properly with text going right to left (like Arabic or Hebrew)?
How many applications work properly on thin clients?
How many applications work properly on very inexpensive hardware?
Also a lot of the other barriers disappear. 3rd world office workers are much better educated and on the whole more intellegent than Americans in similar jobs. Ease of use / training is less of an issue.
OTOH issues like cost become much more serious.
My guess is that the American corporate desktop and the American home / small business desktops will be some of the last markets to fall. It seems pretty clear that over the next 5 years Linux will be deployed heavily on 3rd world desktops in countries like Peru and China. Germany will follow soon thereafter. So lets say something like 2010 for major US deployments.
That’s an easy one: investments. Why should a company ditch perfectly legal Windows/Office licences to make room for Linux? If it does the job then why replace it? Or better: if it ain’t broken why fix it.
ahron: its X
we need a new gui more like quartz in os x but fast but add X layer compatability for old apps
Well, even if we manage to get something like Quartz, maybe PicoGUI or Fresco, it wouldn’t make Linux successful. X11 currently works, and if something else comes in and works better than X11, I’m all for moving for it. But Linux’s success wouldn’t be hindered by the usage of X11.
Besides, Quartz is actually fast. What isn’t fast is all the stupid useless effects Apple insist on. If you try duplicating all this stupid useless eye candy on, say Windows, X11 or BeOS, it would be way slower. It is because of Quartz that Aqua is quite usuable. Now all we need is someone to convince Apple that eye candy doesn’t matter, responsiveness does.
we need vise style installers for everything and afix to the dep hell we all face when we first get into linux
No we don’t. I don’t know how many times I have to help someone install a app because they don’t know how to install an app. And that app happens to use vise. Try thinking up something better. Maybe something like Apple’s idea.
Julien: I hope it’s the future !
We all do, but unless something changes, it probably won’t. The development speed is quite slow because there isn’t much manpower. Plus, because the design is new, it is completely unproven. How would we know if it is feasible at this point.
guess what. the tool kits are irrelivent, the developers use what they want to for there project…the problem is the Libraries. we have the LSB, but as long as developers compile against the latest and greatest Libs, the programs will not work without dependancy resolution.
get the developers to start using the LSB and we will have a nice system from both ends.
QT runs in a gnome environmnet and GTK+ runs in a QT environment.
standardise your tool kit that you develop with for god sakes. any company that focuses on stupid crap like toolkits are stupid themselfs. “oh, it doe snot look like the pretty littl crap” please..there is a unified menu system, and a unified clipboard. you just need the toolkit installed on the computer to run the app. I worked in amny environmnets and guess waht….none of the homegrown apps for company use even looked like the environment they were used in.
who in the hell uses sound in business applications that are farmed out?
and besides, OSS works with aRTs. no one uses ESD…except maby the developers them selfs casue they run E.
@ Red:
> Linux is not a company
> People seem to forget the nature of GNULinux, which is
> that it is Liscensed under the GPL and free to use and
> modify. As such, if someone doesn’t like an idea, they
> can fork it.
But that is part of the problem! That’s how that terrible mess came into being in the first place!
If RMS were a little less interested on his vision of “freedom” (and the FSF’ control over it, I might add), and a little more focussing on how to make that “free” software world of his a world that’s worthwile for people less geeky than him…
That’s also why my very own Open Source (sic!) OS project is explicitly not using the GPL. Open Source != GPL, and Open Source != free (as in both beer and rights).
> However, Linux is never ever going to be exactly the
> way you want it to be unless YOU or a Linux distibution
> company decides to make it that way.
From our initial proposal:
“We would have to change so many things in Linux (facing considerable resistance from those who actually like cryptic options, mystic directory structures and arcane configuration files) that it´s probably the best to leave it all behind, keep the lessons learned in mind, and start from scratch.”
@ Brad:
> True, Kylix is available for developing Linux apps, but
> it is only officially supported on a handful of Linux
> distributions; none of which I personally use.
See the drift? “Supported [only] on a handful of Linux distributions”… hell, that’s supposed to be one OS! Why does Borland have to support distros instead of GNU/Linux period? Because GNU/Linux is a mess!
@ Rajan:
> If each distribution merge into one product, it
> wouldn’t take Linux closer to the desktop, in fact
> may even be hindering it. Each and every distribution
> takes a different approach to the desktop. Try out each
> one, see which you like best. If you don’t like any,
> Linux isn’t ready for you yet.
I did that with Mandrake 8.0, Debian 2.I-don’t-know-anymore, RedHat 6.1, RedHat 7.3, RedHat 8.0, SuSE 7.0, and SuSE 8.1. Of all those, SuSE 8.1 was the worst experience and RedHat 7.3 the best – so much for the “yet”.
All in all, trying Linux again and again did cost me more time than all Windows bluescreens and scandisks (and Amiga gurus / SW failures…) so far combined.
And I am a pro, god dammit (sorry). How do you expect Joe Average User to try different distros? And, btw, Joe Average User doesn’t have a 2.5 MBit connection either, so “trying” Linux does cost him about as much as a Windows license. Uh, less, since he got that one inlcuded…
> You probably won’t find setup.exe, but for stuff like
> StarOffice, there are .bin files that have a nice wizard
> ala Windows.
Ah-uh… and pray tell, where do I find them? /bin, /usr/bin, /usr/local/bin, …? Uh, yes, depends on your distro / version / setup…
> But Python, HTML and XML doesn’t count (it took me a
> day to learn each).
Now, now, now. Either you are the super genius of the year, or you are boasting like a madman, or you have a funny definition of what it means to “learn” a language / markup.
But that’s OT, now is it?
> I don’t know about SuSE, but Red Hat only installs XMMS
> by default.
And mpg123 last time I looked in earnest (7.3) – in the “minimal” install, which isn’t really useful if you want to do “real” work. So you install Office, and KDE and Gnome just to be sure because you don’t want to make up your mind before you tried ’em, and voila…
And please, don’t dismiss valid criticism with “this disussion is on corporate desktop” or “this discussion is on consumer desktop” respectively. Most of the people here are pointing out weaknesses because they want to help (well, I am for one)…
@ Anonymous:
> Unless there is a way for emulation of Windows Apps (at
> no cost) to linux with little to no noticeable speed
> impacts – conversion will never change.
Just like the Amiga dudes thinking that running on a PPC won’t place them in competition with Windows, simply because it doesn’t run on the same hardware – this argument is severly flawed.
What you imply is that Windows users will only switch to Linux if they can continue to use Office, MSIE and Photoshop. If they continue to use the very same software, then why should they undergo the effort of switching at all?
You cannot win anything by emulating. You have to surpass your competition, instead of imitating. Gimp has to surpass Photoshop. KOffice has to surpass MS Office. Mozilla has to surpass MSIE. KDE has to surpass Explorer. RPM has to surpass InstallVise.
And choice, reliability and “freedom” are only a part of the equation. Usability, consistency, compatibility are others, and the list goes on.
>> “The two tool kits in Linux are qt and gtk, qt for C++, gtk for C programming.”
May I point out to you that Gtk is not only for C programmers?
I’m using Gtkmm2 to build an UI for a pretty large app and I feel very comfortable with it.
(more than any other UI toolkit like wxWindows, Qt or even MFC)
Gtkmm2 is actually just Gtk2 (ie. gtk_object_function() becomes Gtk::Object->function()), but you don’t have to take care of all the irritating things from the C Gtk API and it’s completely ‘C++-ified’.
1. They need to advertise it, people don’t know about things they never hear about & if some of the bigger distros start advertising, do it on none geek sites and stations if you want normal users to hear about it.
2. I do agree some things need to get easier, like the packeges, I mean I can install from a .rpm, but a windows like installer might be good. The autopackege thing I read about here a while ago might be what is needed.
3. More programs, games would be good.
Linux programs should all have standard interactions. Tab should, like in most windows programs, move from one promt to another. Alt should take you to the (file/edit/view) menus at the top… etc.. Who wants to spend 20 mins learning a new programs unique way of doing things every time they test out a new program?
Power features I miss in Linux are: The ability to make tables in wordprocessors, such as how Word does. The ability to open a word document w/ tables & not have it look wierd.
The ability to auto-convert a word 2000 document into an html page. The ability to clean up word’s extranious html code… such as how dreamweaver can. Dreamweaver MX cannot be WINED.
These are real issues that I have ran into when using Linux & it is the reason I’m using that POS w98 right now.
I wish the people that I work with did not write all their documents using word 2000, but that is how business is mostly done these days. & I cannot do any practical work w/out the powerful tools that I have only found in Windows. If such tools exist for Linux please enlighten me.
This whole debate of audio problems on Windows and Linux is ridiculous.
Someone said that, with Windows, if audio doesn’t work it’s either driver or hardware, and that on Linux “who knows”. That’s as maybe, but the fact is that your Johnny Desktop-User really is not able to fix either of the operating systems. So the point is completely moot.
And I claim that Linux is always fixable if you know what to do, but there are situations where reinstall is almost the only thing that fixes your Windows driver problems (even if you know what you are doing).
I’ve been a Linux user since the 0.98 days, and it’s *never* been a suitable (or really even very usable) desktop environment. It’s not bad as a server (although BSD is a better choice in most cases), but to be honest, it’s really a pretty substandard desktop.
In general Linux desktops only appeal to those that *want* to engage in endless twiddling and mandatory customization – the people who are anal enough to go tweak all the X configuration settings, themes, and such. Linux has finally gotten to the point where it is *marginally* usable as a desktop alternative, but really, unless you’re being driven solely by a political agenda, I can’t imagine why you’d choose it as your primary desktop.
Finally, there are just way too many good Windows apps for which there are no suitable Linux alternatives (Visio comes to mind (and no, Kivio, Dia, etc. are NOT in the same league), as does Six Degrees, which is good enough that it’s got me, a Netscape/Mozilla Mail fan since day one, seriously pondering a switch to Outlook just to get its benefits.) The reality is, as much as Linux folks would like to pretend otherwise, that Linux on the desktop lags WIndows in terms of functionality by two years or more, and in ease of use (including configuration and installation) by light-years.
So really, many of us need an environment that can do *both* Windows and Unix/Linux: But it’s FAR easier to add Unix/Linux functionality to Windows through something like U/Win or Cygwin than it is to try to make Linux do Windows.
The reality is that running Linux as a desktop is a pain in the butt, and a whole lot of things just won’t work easily anymore. (This is true even for some open source software on Windows – I just wasted 15 minutes getting Real Player to work correctly with Mozilla on W2K – there’s no excuse for that, especially since it works with the substantially similar Netscape 6/7 out of the box…)
That’s why Windows will likely remain my desktop of choice, and Linux will continue to go begging, even for the desktop of this 17-year Unix bigot. Just because I *can* plow through all that crap doesn’t mean I *want* to…
There was a dozen of window systems, but X11 won.
Won what? I mean, it sure works, and as far as a “network transparent” graphics system it’s probably unbeatable, the thing is that for a desktop role it’s lousy, you don’t need a server/client system and all the trouble that comes with it if your’re only using one client. This is one area where you don’t have any choice.
As was said upthread about 20 or so posts ago, there’s too much RTFM instead of making it work.
In the end it all boils down to useablity.
If it cannot be used, it WILL NOT be used.
The systems department has enough to do where I work. I’m sure they’d LOVE to spend even more hours just trying to get the printers to run, the video cards to work, and all the software installed and playing nice.
And then there’s little things like … hmmm, this video conference I’m supposed to watch is streaming in QT or RM.
a) Will I be able to find the software I need to watch it?
b) *IF* found, will I be able to install it?
c) *IF* successfully installed, will I be able to quickly and easily find and open it?
See, right now, I can do all of the latter at work (and home) without needing to get an Alpha Geek from systems.
As I load the new RedHat today, I’m once again reminded that a large portion of my hardware is useless in a Linux environment and that much of it, like my excellent Xerox M940 Workcentre scanner/copier/printer will *never* be able to work in Linux.
I like Linux in a lot of ways, but as I stated earlier, it’s a desktop best suited for masochists.
<quote>
The systems department has enough to do where I work. I’m sure they’d LOVE to spend even more hours just trying to get the printers to run, the video cards to work, and all the software installed and playing nice.</quote>
It takes me maybe close to an hour of install time (with Redhat I just start the Kickstart and leave) if it is acting really slow and about thirty minutes of configuration time (setting up the user’s closest printer and getting all their favorite network shares into LinNeighborhood) to get a Unix programmer up on Linux.
RM streaming content? Install Realplayer. We do here by default along with java, flash and all the goodies we can find.
There are better examples of gaps in linux software out there. This ain’t one of them.
QT? mplayer is coming and crossover plugin is here now.
All the stuff I install is sitting right there in Applications menu.
The fun prob with Linux is it does not come pre-installed on computers set up to run out of the box. The installation and initial setup is a bitch but then again I watched a friend try to get his network card running in XP and that sounded like one of those linux horror stories every other person keeps listing. MS says it should work, he screeched.
I laugh.
I still say that until there is one killer app or benefit to using linux that outwieghs the minuses of using an alternative OS then you are not going to see linux or any other OS take away space from MS. period.
I watched a friend try to get his network card running in XP and that sounded like one of those linux horror stories
Ok, this is a little OT, but I laugh at people who spend hours trying to get a network card to work, when you can go down to your nearest computer/electronics store and buy one that works out of hte box for $10-$20.
The ability to clean up word’s extranious html code…
Have you tried htmltidy? http://www.w3.org/People/Raggett/tidy/
Two things are needed:
First – Easy installation of programs/packages, standardized for all distros, and backward compatible. Apt-get is nice as far as it goes, but would need to include *all* the latest packages.
Second – A file system/file management where it’s easy/intuitive to know where to go to achieve.. whatever, and also to know where packages are installed, and what goes where and why.
All in all it just needs to feel and operate more simply and intuitively. And for the rest, there will always be slackware.
Sheesh….why all the bias against Slackware?
I’ve been using it for years, and find the Slackware package system to be very intuitive and easy to work with.
The FreeBSD ports system is by far the best in existence IMHO, but the Slack packs are a close second in my books.
From the linuxprinting.org Xerox mailing list this post:
I have the Xerox WorkCentre M940 printer and it works PERFECTLY for me.
(At least I haven’t found anything it doesn’t print.)
This is what I did:
– Download CUPS and ESP Ghostscript from http://www.cups.org
– Installed those two packages (remember to install the font packs or it won’t work)
– Came to LinuxPrinting.org and got the PPD file for “Xerox DocuPrint M760 (PCL3)”. I know that this is a different printer but the M940 is PCL3 compatible just like the DocuPrint M760 so I tried it and found that it
works GREAT.
– I then basically followed the CUPS and LinuxPrinting.org Help file found at http://www.linuxprinting.org/cups-doc.html. I installed cupsomatic AND the foomatic-gswrapper just like it says to.
– I set the “my $enscriptcommand =” command to “enscript” in the cupsomatic script, but I don’t think that this matters.
– I went to localhost:631 to config CUPS and added a new printer. The PPD file for the M760 is listed there, now, and I selected that one.
That’s it. This printer works great for me. It prints color, graphics, everything with no problem. I think that the resolution might not be what the printer is capable of, but that’s no big deal. Text is very clear on
printouts. I’ve successfully printed from: Mozilla, Acrobat Reader 5, GEdit, GNotepad+, and many other applications.
This is my question, does anybody know how I can get this printer added to the database on linuxprinting.org? Also, it would be nice if this printer had its own PPD file. I believe that we could just copy the PPD file for
the M760 but rename it for the M940.
A big “Thanks” to everyone on this website (linuxprinting.org) and in these mailing lists for helping me to get this printer working! This is so exciting, I was afraid that I had a paperweight until now.
Later-
Michael
Ok, this is a little OT, but I laugh at people who spend hours trying to get a network card to work, when you can go down to your nearest computer/electronics store and buy one that works out of hte box for $10-$20.
I agree. If people would just do a little research before buying a machine or any hardware, they would never have hardware issues when running either Linux or Windows.
Who the hell wants to go through all that just to install a damn printer? Am I missing something here? I like the Windows way … hook it up, insert the CD when it asks, and go. hell, if it recognizes the printer, a CD is not even necessary, no dialogs … no config files, no drivers, no nothing.
“””Who the hell wants to go through all that just to install a damn printer?”””
Last time I installed a printer in Linux I did the following: install the CUPS binary (would’ve done it when I installed in the first place, but alas, my printer was in storage and I didn’t think I need it…); connect to the cups daemon (631) with my web browser, go through the add printer sequence, select it from a list. Twas done.
This is if you do not have CUPS or ESP or anything.
For SuSE you configure the printer under CUPS choose the PCL generic (or M760) driver.
As in like Windows –> Add Printer –> Detect Printer –> Which driver? –> PCL –> Go.
I wanted to post something from a site proving that it was possible that is all.
Jeez/
I wasn’t being elitest about saying Slackware users know more about Linux and that if you used Slackware you would more likley know about how sound worked. I was trying to make a point about false Authority syndrome.
Solar said he could admin Linux and used this to make it seem like he knew that Linux Sound was complicated. When in reality he should have checked wether it was his own lack of knowledge that was the problem.
A newbie may not understand but a admin certainly should.
as long as i aim able to break my sound, internet connection or apps that i like, use and that worked before i clicked some button or opened another app, but didn’t worked anymore afterwords and i was able to do that with all the most popular distros within an hour and as long as the install is different for every distro, it won’t be ready for me.
Still i will try new releases of Redhat, Mandrake, Knoppix, Yoper and so on and hope i’ll be lucky on day soon. Windows works just fine for me, but i don’t wanna MS to become more powerfull then they already are and i don’t like their vision of the future. So i keep trying linux, but i don’t wanna learn the system, just use office (it’s there for me), photoshop or paintshop pro, dreamweaver and listen to music (with sound) and watch movies all over the internet.
The shop I work in currently uses Windows. But not ALL of the stuff that comes on the installation CD. The PC Support staff ask all of us what we need before the new OS is standardized, then they give us some fraction of it.
They would do the same with Linux. There would be one desktop, one office suite and one browser, NO MATTER HOW MANY CAME ON THE DISTRO CD SET. The end user would not be bothered with these choices any more than they are with Windows; they would use what they were given just like with Windows.
The end user would not be bothered wwith all those config files, either. They would be setup before the user saw the PC and they would probably be fenced out of them or a new set would get pushed over the network if they called for service. Pretty much like what happens here for Windows.
What’s keeping Linux off of our desks even during our current budget crunch is management fear. They are familiar with Windows and fear Linux precisely because they are not familiar with it. They know what works with Windows and what doesn’t, but they have no clue what works with Linux. Since MS-Office is so prevalent, management also fears that if we get off Windows (or even just Office), we won’t be able to communicate with the outside world.
If we DID go to Linux, our average user would be productive in about the saem amount of time they are with a new version of Windows or Office. Most of them are self taught on Windows and they would be self taught on Linux. Most of our vertical applicaitons are written in-house, so we have the design docs and the code.
I don’t think it would be a big deal for the users to switch. The Support staff would go nuts for a while, but they do with a new version of Windows, too. Looks like a wash to me.
I know I had to do things the dangerous wa, that’s true, but that wasn’t my point. My point was:
– why did I have to do this crap in the first place? Why are there apps that actually need to run in the newer KDE? Not many, but there are. For me, KDE 1.1.1 was exactly enough.
– why did they make the new KDE depend on the new Xfree86? I mean, what is that killer feature that got KDE on each desktop in the world, that really needed the new Xfree86? To me it looks a bit like developer egomania.
– why could I not find all the neessary packages? NOTE that I had similar problems with other distros, too (except with Suse, but only because I wasn’t using it that long), not dissinkg Slack here!
– and finally, why isn’t there just a huge package called KDE_verxyz.pkg (or .tgz, or .deb) that has all this shit inside, so I don’t have to worry? With the f* windows that I totally loathe, the only thing you end up having to install, if anything, is some version of DirectX. If you tell me thatsuch a package is not possible because KDE depends on way too many things that change way too often, well, then there you have the Linux problem. Every developer group dicking with their pet project changing it and releasing a new and improved version every week. The kernel, Xfree, gcc and a whole pile of libraries that all those little developerheads are so proud of, changes too frequently. No wonder I could not find the necvessary packages, who in the world can keep up with the pace? By the time you collect all the packages you need to satisfy depedancies you go silently mad.
And if you think that the solution is, let’s say, use RedHat for distro and only one source for packages (RedHat), then I would not be able to install half the apps I need (not me personally perhaps, but some of my colleagues). I am very glad that it all worked OK for your friend, but I need some net analyzing tools from time to time, and why not, play Freeciv.
Now you have the opportunity to admit that something is wrong with how Linux is developed. Come on, fess up, let me see some objectivity in the Linux world.
Johnathan Bailes: “Are you a Unix programmer, sys admin or network admin who pines for the power of Unix without all the cygwin nonsense that never seems to fit right in a NT world?”
Dub Dublin: “So really, many of us need an environment that can do *both* Windows and Unix/Linux: But it’s FAR easier to add Unix/Linux functionality to Windows through something like U/Win or Cygwin than it is to try to make Linux do Windows.”
Dub, you should check out Microsoft’s own Services for Unix product for a fully Unix compliant sub-system under Windows that is full-featured (incl. NFS capability) and works properly (not merely a hacked together little DLL like Cygwin):
http://www.microsoft.com/windows/sfu/default.asp
See some of the *nix software that has been ported to SFU/Interix here:
http://www.interopsystems.com/tools/warehouse.htm
…is two things (of highly various complexity!):
(Note: Assuming the “Linux desktop” is based on GNOME)
– A good file selector
– System Policies
After that, it’s up to the apps…
I did not just say get your rpms from Redhat but the combination of redhat with apt-get on freshrpms makes a pretty powerful combo in terms of selection. Freeciv comes RH8 BTW. :->
I fully admit that the packages for things like KDE and Gnome are split into too damn many subsections. This is a pain. It is true. I compiled Gnome 2.0.1 by hand for my SuSE box (I guess I was in a masochistic mode) and it is amazing how much many subpackages are needed.
Still, that is why you stick with your distro and upgrade it on a fairly consistent basis if you want new desktops and such. Nobody complains that they have to install XP to get the new XP look and feel for example.
On the depends relating to cutting edge packages. They are developers for goodness sake and they do this for the most part in their free time as their own project. The kernel and clustering and device support gets big corporate support (from Redhat, IBM, SuSE and others) but the desktop core gets amazingly little of this. These are people who did it in their spare time. If they want it to depend on XYZ their attitude is its their puppy if you don’t like fine, a file a bug report. But no one does on Usability issues or issues relating to too many depends.
It is their pet project. If the distros and other big boys pushing linux on the desktop don’t like it they can start by committing some programmers to the cause. Ximian has a big voice on the Gnome Foundation because of this.
Corps standardize their desktops. Some even download almost the entire OS to the users’ desktops every time they log in (at work I live with login times varying from 20 minutes to more than an hour!).
Everything is completely managed. Linux can do that magnificently with its powerful shell scripts and open nature. Corps can standardize on KDE or Gnome or even roll their own quite easily since the standard office desktop only contains a small set of apps. Theme those (it’s open source, everything’s themeable given enough effort!), and you’re set.
Prolem is with apps already developed and deployed in-house. Most of this is Windows stuff and I’ve seen some REALLY expensive apps for archive management (you know, climate-conditioned rooms with lots of boxes containing real paper documents). This stuff costs BIG money to port over to Linux.
Johnatan, what you say is true, but it doesn’t make Linux’s situation any better. You did not provide an argument by saying “it’s their pet project, if you don’t like it that’s your prob lem”. It’s not my problem, it’s Linux’s problem.
Now, have a look at this. Honestly, I wasn’t hunting for it, I am just very much into composing on the computer and everything thatrelates with music software (be it on Linux, BeOS or Windoze) attracts me:
http://slashdot.org/comments.pl?sid=47498&cid=4857243
See what this guy is saying? One thing I’ve noticed is that other software packages I’ve downloaded elsewhere don’t always get along with the libraries, probably because they were written with older distributions in mind.
It’s stuff like this that makes me barking mad. I just don’t put up with it anymore.
Freeciv is a work in progress. It’s still missing some pretty basic features, but I like it because I can configure some neato stuff, like map generator type, fog-of-war, exact map size and water/landmass ratio etc.
But recently I discovered that the Win32 port of Freeciv is more usable than the version that natively is built for Linux.
People keeps saying “I got choice” and go on to mention their choices consist of Windows 98se, Office 97, etc. You made the choice to buy one of the most expensive word processors. Why buy a computer when all the choice you want is to edit, print, and email out letters. You had that same choice 50 years ago. Why does it take so long for people to make stupid decisions.
So many people buy computers these days when all they really need is a typewriter and a cell phone. A computer allows you to do things like store all your data and content on a network which can be accessed from anywhere in the world, even wirelessly. Your data and content consists of your bills and finances, movie, music and book collections, home movies, family pictures, artwork, links to information, etc. The choice we are talking about is your freedom to store, process and manipulate your data and content. This has nothing to do with choosing to use Office 97 or Office 2000. You are a moron if you think that’s choice.
An en masse switch won’t happen immediately no matter
what is nor isn’t holding back Linux Desktop now.
Overtime as more people are ready for linux ,it will be more ready for them.
Don’t like Linux, wait a while. Like the weather it will
change but only better.
As business starts to clue into advantages of being free of the expense and hassle of msft licenses that force upgrades on you and limit what the user can do with the software. And with them will come the users.
Needing work:
Installs: Not much work needed here.Linux installs are
pretty well nailed these days. Yes, any install on any distro can still hit a snag. XP is no different maybe
worse, if you try and instal it yourself.
And you can reinstall Linux. xp only comes with a restore
disk, and if you change your hardware configuration you
have to beg.
Fonts: Xft2 is making its way into the Distros and qt 3.1.1 supports this now, so this is almost licked.
Cut and paste between gtk and qt apps. Works now but could be better. Get it together.
Menu cooperation. Baah. More a problem of organization. and cooperation. This could be fixed like.. last year.
The good and bad news that this is as complicated as two people agreeing on how to organize a Medicine Cabinet. I only wish this was the most complicated technical “problem” on the Desktop.
Too much choice: Well Lindows and Xandros are giving more
focus by limiting choice. There is no need for this to
be an either or propositon. No one is stopping Distros from giving users an install and desktop that can default and be adjusted to whatever level of complexity the _user_ wants., ie from hand holding simplicity to the total flexiblity for the expert.
Does anyone really install 4000 programs, just because they are on the CDS ?
I am hoping there to be some healthy
cross fertilization of approaches amongst the Distros.
The Dispaly Managers: xdm, gdm, kdm .
Linux remember we are about _playing well with others._ Installing a new display manager should never result in your previous window mangager bieng mysteriously nixed. It should always be there as an option.
X11 speed. Faster hardware is slowly making this less of an issure. Perhaps hardware drivers giving more of a nod
to X would help as will programs being compiled on newer
gcc. I am not sure. Use icewm if your resources are meager,esp if you are transitioning from Windows.
Apps: hmm. A lot of people will find all of the apps
they need on Linux. Many people will have most of the apps but be shy one or two, that may be key to them.ie
ie ,Adobe Illustratror. But for most people most of the apps are there, now.
There is always the Dual boot solution. But don’t upgrade
to a new Microsoft.
Remeber what the signs at the zoo say:
Please do not feed the Beast.
Most of the advantage of MS apps is compatability with
themseleves, not superior features. This is artificial
result of their monopoly. As more people “get off the
Mircrosoft” this will be less of an issure.
Mail,news,spreadsheet,wordprocessor,music,burning, text editor… check, check, check,…
Browsing ? Don’t make me laugh. I have more fingers
on my left hand than the number of times I have ever used
IE in my life. Even when I was on w32 I used Netscape and
Opera. IE is a pathetic excuse not to switch to Linux.
All the custom apps will come as Linux gains propularity.
Finally,as a rank newbie ,fresh from Windows, installed
Corel Linux. Then I went onto, Libranet, Debian, Mandrake Suse and even Gentoo. And I had to learn a few things along the way (back then) like partitioning, which a computer professional would already know.
So for them it should be easier as they could focus on what was truly different about Linux.
It wasn;t that tough for me so how could it be so tuff
for them?.
I am sorry,but imho if you are a msft professional computer guy of many years and you can’t install Linux or even if you find it really dificult installing Linux then it seems to me you are proving the point of all those who say being an MS administrator does not give you a deep understanding of computers, it just makes you a point and click chicken.
So then when it comes time to install Linux all you
can do is go buk,buk,buk.
Would you prefer to walk into your favorite store only to find that it only carries one model of TV, VCR, DVD Player, etc. or would you prefer to have a wide range of selections for your entertainment?
Peeple need to educate themselves of what choices they have so they can make an informed decision of what best for them. The same can be said about Linux and choices users have.
Jonathan B says:
The installation and initial setup is a bitch
—
Yeah, and the systems people just want more bitchiness right out of the box.
And here’s the thing I don’t get. Everybody keeps saying, oh, Linux is so easy, but I read 2 or 3 posts down and there’s all this garbage about going to the command line and mucking about with sys configs.
Not to mention all the things I keep reading about “packages” for downloaded software. On windows and the mac, I double click, it unzips, I tell it what folder to go live in, I click on the icon and I’m off and running. If it’s the same thing in linux then what’s all this bother about needing to package an app? What do I do when I download it and I can’t just click and go? Will I have the permissions to “package” the app? (And I’m sorry, why isn’t it ready to run right off the bat?) Or will I have to place a call to an IT person and get the problem fixed in a couple of hours?
Look, at the end of the day, the systems folk only care if it works or doesn’t, and all too often, it doesn’t under Linux for the desktop at this point in time.
Every minute spent trying to make things work is time spent not working which is why Linux is on the servers but Windows or Mac is on the desktop.