How? To illustrate, let's compare two users, Abe and Bret, who have purchased hardware and seek to install an OS. Abe buys a copy of Windows and proceeds to do a simple installation. When it finishes, he has a working setup, just like every other Windows user in the world. His out-of-box (OOB) experience with the OS includes the most basic tools he needs, and lets him install additional software for his specific needs. These he chooses after due research or based on peer input. Of course, there is cruft too, which he can choose to let be or replace or remove.
Now let's catch up with Bret, who is part of the growing minority deciding to install a Linux distribution on their machines. Before he can begin, however, he must decide which distribution. Fedora? Debian? Mandrake? Gentoo? Xandros? Bret is, understandably, in a quandary. He spends considerable time scrounging around for information -- browsing forums, asking for opinions, visiting distro Web sites. Eventually he picks one and installs it. Either he likes it and retains it or he goes back to hunting around.
Unlike Windows, it is not enough just to choose Linux. One must choose which Linux. An average user, when confronted with more than a thousand distributions (or even just the 10 most popular) and asked to make a choice at the very outset will, in all probability, take the easier route and stick with Windows, where he can do work without having to do research beforehand. Having to research and choose a distribution may keep a considerable set of users away from Linux.
How might Linux developers solve this problem? Let there be choice, but let there also be a basic consistency for those who wish to choose a little later.
Today, each Linux distributor offers a certain feature set, and each seeks to make this set as unique as possible. One would assume that they bundle the basic Linux kernel, the neccessary run-time libraries, the toolchain, the set of console programs, and the X Window System, and then add packages and software above this or extending these. Sometimes, however, distributors make changes with implications that affect the basic packages I've listed above. One distributor adds a security feature to the Linux kernel, another compiles his distribution with prelinking enabled, another adds a beta of a given component instead of the more-popular stable release. This tends to magnify differences between the distros.
Now imagine an understanding between the various distributors, akin to the work of Linux Standards Base (LSB) (which seeks to define a few standards, in terms of common API, file system hierarchy, even behavior of certain programs), but with a bigger charter -- to provide in every compliant Linux distribution an install option, called Linux Base, that will install the same set of basic packages to get a user up and running in a similar setup across distributions. Use of this install option would give new Linux users something to stand on while they explore Linux, the various available distros, and their features. Further, it would simplify the answer to the question of "which distro?" from a Linux newbie to "Install any of those, choose Linux Base, then look around."
For the distributors, most of whom have subscribed to LSB standards, it would be a simple extension to agree on a basic package set.
Notice the similarity to the Windows installation model. Install the OS first and then choose what other software to install, or in the case of Linux, choose other packages from the distro, or an entirely different distro.
The choice would still be there, just as it is now, but a common, basic install option would eliminate the need for potential users to have to ask too many questions before they can even get started.
Vinod Vijayarajan works on BlackFin-based CE systems in Emuzed, India. He is a Linux enthusiast who believes in raising the bar every time the OS beats the one he set last.
Note: Comments are owned by the poster. We are not responsible for their content.
Precisely.
So what YOU should be saying to them is "Hey, it's not a big issue. It really doesn't matter WHICH linux distro you choose. They are really all the same. They all come with the same stuff. They all work the same. You can mix and match you know. It's really very simple.<nobr> <wbr></nobr>.. and, hey, if you get really unlucky, throw that one away and choose a different one<nobr> <wbr></nobr>.. or -here- take this one, it's good - I use it."
You know, education of this sort will help scared wanna-be-newbies.<nobr> <wbr></nobr>.. and if THAT doesn't calm them down, then SCREW THEM! Yeah, that's right - that type are more hassle than they're worth, and the reality is they will never switch.
<nobr> <wbr></nobr>.. And as a final point, Linux is not about 'beating' Microsoft. If anything, Linux is about UNIX on a PC, Open Source/Free Software, choice. That way, we are building on what's good, rather than competing with what's crap.
(sig) IANOD - I am not a dog.
People who don't want to do even the most basic research deserve what they end up with. They are solely responsible for their own satisfaction.
Why don't we hear that there are too many cars, houses, radio stations, TV shows, pants styles, greeting cards, etc. to choose from?
Instead of having a Linux standard, how about just having a Linux Distro Consumer Advice Sheet (or something). A committee could still sit and judge, it could still have 'clout' because of how it evaluates and reports, it would have expertise in Joe Wannabe and marketting and<nobr> <wbr></nobr>.. um<nobr> <wbr></nobr>.. er<nobr> <wbr></nobr>.. (new idea, not quite formed yet)<nobr> <wbr></nobr>.. there it is.
I think a single sheet (or short web page) with big friendly writing and short friendly words would be so helpful. It could list, summarise and compare the 'easiest' or 'safest' distros for newbies, and at the top, just list two or three 'for refugees from the dark side'. Yes it would be biased, and us geeks would hate it - but Joe Sexpic and his girlfriend Jill Sexpot might really appreciate it.
Because they don't care. And the few that do care know where to find this information. But they haven't solved the problem by reducing the number of models of televisions sold.
Choice is only choice if the user understands something about the choice they are making.
The user doesn't need to understand the choice if he doesn't care. Whether emacs comes with the foobar extension isn't important if he doesn't need the foobar extension. People don't analyze the towing capacities of pickup trucks if they aren't doing any towing.
Anything else isn't merely useless (guessing) but actually -bad- because it embarrasses the poor user who doesn't understand the funny man at the counter and feels inadequate.
The job of the salesman is to explain the different options to the customer in a way that he can understand. It's the same for people trying to encourage others to use Linux.
Explaining and evaluating the choices is the solution, not reducing them.
might make sense IFF vendors can agree WTF these phrases should mean.
Market-speak has been around for much, much longer than operating systems have. "Retsin", "Fahrvegnugen", and "Flouristat" come to mind. They are "ingredients" or "features" designed by some marketdroid to confuse the consumer, not enlighten him.
When people goto the store to buy a battery, they want it standardized, not some unique thing that won't work with their equipment.
Batteries? WTF. Does your distro go flat after a few weeks? Do you plug your distro into some kind of socket in your computer. And if you get the wrong distro it blows the fuse? Are you crazy.
Where I come from, we have many brands of batteries, each with its own selection of 'heavy duty' or 'cheap as crap', and many different shapes and voltages. I know what shape I need, I guess at the brand and I take a punt at 'heavy duty' or 'cheap as crap'. I don't argue with everyone about how there should be some "Standardised" battery. "<nobr> <wbr></nobr>.. why are there AAA and D sizes? someone should just<nobr> <wbr></nobr>.. "
<nobr> <wbr></nobr>.. why is there KDE vs Gnome, someone should<nobr> <wbr></nobr>.. blah blah.
Have you ever tried Gnome or KDE? I have. Have you ever actually gotten to the point of installing (or even using) any of these millions of different distros you are talking about? I have. And you know what, on some of them KDE and Gnome even end up having the same menus and the same applications. Is that what you are talking about when you say 'someone should make a version that combines them'?
The fact is there is no real difference between any of them. And because you don't seem to realise this, it makes me think that you have never really actually used linux. Joe Shitzu was able to migrate from Windows 98 to Windows XP. So he is able to handle either Gnome or KDE or even twm, you know - if you configured it to work like MS Windows. And he can definitely handle RedHat or Fedora or Mandrake or Suse or Knoppix or Slackware (but probably not Hurd). If he asks you which one to use just pluck one out of thin air - they are all equivalent.
If you had your way, the 'Officially State Sanctioned Standard Linux for Users' would just look and feel like Windows and run the same applications. But here's a dilemma for you - Which version of Windows? 98? XP? Longhorn?
Gee, it's all too confusing for me. Think I'll stick with DOS (That's AmigaDOS)
Wow, Synonymous, you have had an idea that no-one ever thought of before. I wonder why all those other distributions that sell hundreds of thousands of copies a year never thought of that. "Just take the good out of each and put it into one." Wow. RedHat would be much more popular if it stopped putting all that bad stuff in. And same with Suse. And come to think of it, Mandrake would be so much better if it just stuck with the good stuff. Gee, even Gentoo would be better if it was just standardised a little bit. Or Debian, now Debian would definitely benefit from losing all that crazy bad stuff - it would be so much more stable and y'know.
Well, looks like you are all set up to get a band of like-minded people together and change the world. You are lucky you are doing it in an Open Source environment, cause you would have buckley's (that means 'no hope') of doing it under MicroSoft.
PS. I appologise most sincerely if I am being a little harsh on you and your wonderul idea. It's just that this kind of thinking is ill-informed. I find it a bit hard to explain to you (see some of my other posts) but somehow you come across like someone who has a bit of knowledge but no experience. Come on Sysnonymous, stop complaining and just jump in. The water's fine.
The problem here in the Linux world is not so much a question of choice as it is a matter of communication. I have converted a number of casual computer users over to Linux. One thing that stands out my memory is this: When then person that is new to Linux starts to get overwelmed by the range of available choices in a Linux distro, I ease their apprehension by giving them one little tip: THE VAST MAJORITY OF CHOICES IN LINUX ARE MATTERS OF TASTE.
To a new user, the presence of so much choice can be interpreted as the obligation to choose. But what basis does a new user have on which to make such choices? The answer is that they have none at all. The deeper insight is that most choices need not be made on a rational, logical foundation. Choose Gnome because you think the icons are pretty, choose Konqueror because you think the name is cute. All are good choices.
One thing to remember is that even though most people are not trained economists, they do have an economic sense in their gut. People are aware of the concept of opportunity cost (even if they don't know it by that name). In the proprietary software world, choosing a Mac means NOT choosing a Windows box. Choosing WordPerfect Office means NOT choosing Microsoft Office. Every choice that is made comes with consequences in the form of choices that can no longer be made (without incurring additional cost). Their are very, very, few analogs to this circumstance in the Free Software world.
Free Software is like a buffet, you can consume as much (or as little) in quantity and in variety as you want and when the check comes, the charges are still only $0.00.
That is a gem!
I think I have just discovered the perfect sig.
THE VAST MAJORITY OF CHOICES IN LINUX ARE MATTERS OF TASTE
<nobr> <wbr></nobr>...let's compare two users, Abe and Bret, who have purchased hardware and seek to install an OS. Abe buys a copy of Windows and proceeds to do a simple installation.
Which Windows? XP Home? XP Professional? 2000 Professional? Server 2000? Server 2003 (pick your flavor - Enterprise, Small Business, or Web Edition)?
When it finishes, he has a working setup, just like every other Windows user in the world. His out-of-box (OOB) experience with the OS includes the most basic tools he needs, and lets him install additional software for his specific needs. These he chooses after due research or based on peer input.
I will assume that the machine didn't hang during the install, and that the hardware Abe chose was all recognized properly. To save a few dollars, Abe went with XP Home, but when he goes off to college, finds that he can't join the campus Active Directory network. Oops. Time for another trip to the store.
Before he can begin, however, he must decide which distribution. Fedora? Debian? Mandrake? Gentoo? Xandros? Bret is, understandably, in a quandary. He spends considerable time scrounging around for information -- browsing forums, asking for opinions, visiting distro Web sites.
Interesting. Abe's activity is the sensible-sounding "due research" with "peer input," but Bret is wasting his time "scrounging around." And why should Bret be in a quandary? All of the distributions mentioned, with the possible exception of Gentoo, should be simple to install and offer the functionality he seeks. If it doesn't work out for some reason, a different distribution is only a download or a few dollars away. His first choice isn't a "'til death do us part" decision.
Unlike Windows, it is not enough just to choose Linux. One must choose which Linux. An average user, when confronted with more than a thousand distributions (or even just the 10 most popular) and asked to make a choice at the very outset will, in all probability, take the easier route and stick with Windows...
See above - you do indeed have to choose which Windows. Of course, most systems come pre-installed, so this choice is made for you. Bret can do the same thing if he buys from walmart.com, Monarch, Pogo Linux, etc. And you still haven't explained why Bret needs to select his ideal distribution match on the first try.
Vance
Short Answer
Posted by: ThoreauHD on November 06, 2004 02:48 AM#