Yesterday, a co-worker of mine and I had a lengthy discussion about this article posted on OSNews awhile back. My past writing about Linux has centered on general usability and sensible defaults, but his contention was that Linux is the Linux kernel and that anything beyond that is the responsibility of the distribution. The conversation took an interesting turn. Read on for more.
Post a Comment
Wow.
I find this to be one of the best articles written lately on OSNews. I think I agree completely. Choice is strength, but too much choice, is a weakness. Contradictions are everywhere in the Linux community. The most remarkable of course being the kernel itself. The center of the community that promotes choice, the linux kernel, lacks choice. There's a small group that decides what goes in, and what goes out. Just as Microsoft does it. Just as Apple does it.
That ain't choice. We can watch the choice being made, but that doesn't change a thing, It still ain't choice.
I really have nothing to add.
When it does not really matter if you run gnome or kde because with the apps you use drag and drop works, notification works and sound works, and so on, then how can choice be bad?
I'll agree on the fact that defaults matter, and in fact I use Fedora but think that their custom default kinda sucks: the gnome project default layout is a lot better, and takes away less screen real estate.
You are both right, in a sense: linux is only a kernel, and defaults matter.
So a "linux distro" defaults will be important.
Where is the problem if there are two (N) distros that look and behave very differently? It's a free market, you know?
OK if you belive it that strongly, make your own distro.
Let the market decide. It seems like everyone sets thier desktop up differently. Uses different editors etc. it's what works best for the way thier minds work.
If you think forcing every one to think and work the way a committee or a maintainer works is the way to go then build it.
There *are* people trying to replace the "monopoly" of Kernel. Debian/Hurd is one. Someone ported glibc to BSD. Much of Linux userland is built upon C and POSIX, not specific Kernel. So glibc matters.
There are lots of kernel patches. I'd say vanilla kernel and kernel-ck (http://members.optusnet.com.au/ckolivas/kernel/) are choices you can choose. -ck patchset is clearly targetted for system responsiveness and desktop use, which may conflict with vanilla Kernel's goals.
I agree completely.
The point is simply made that a distro should be simple but allow itself to be extended (if people want the kitchen sink they can have it but they've got to get it).
I just wish SuSE had had that focus when I first used it (now I've moved on).
if he dont want choise then go for linspire or similar distros...
Nicely said BUT there is one huge aspect the author missed.
"What is choice?"
Choice is what users perceive as choice and I strongly believe that most of the users don't see beneath the skin of the apps. They see the interface that the apps throws at them and nothing more... if it scratches their itch they will use that app if not... there is always THE CHOICE of using something else. Using sound programming design the interface can be something very easy to implement. The real hard part of an app is that what the users will never see AND WILL NEVER CARE ABOUT: the inner-works of the apps, the actual implementation of what that app is supposed to do. Why does KHTML exists? I'm talking about the widget not the Konqueror browser. Was Gecko bad? Does most of the users care if Konqueror uses KHTML or Gecko? How much code does Kopete and Gaim share? How much code does Abiword and KWord share? Designing or tweaking an GUI isn't a waste of time BUT perpetual reinvention of the wheel at implementation level IS. In my ideal world developers will be busy developing building blocks, ultra-smart widgets, for a true cross-platform toolkit and the interfaces we'll use will be "simple" high-level scripts aided eventually by some XML. The difference we perceive now between applications will be, in those days, perceived as we perceive now the difference between skins (in the apps that support skinning)
Certainly one of the worst articles ever published on this site and it's running against strong competition.
The problem the author describes is a non-problem. Take any of the userfriendly linux distributions and the choice on which editor to use will allready be made for you. If the fact that it is possible for you to choose an other editor if you are so inclined is unbareable, please stay away from any computer, any TV and try not to leave your room at all.
And his musings about standards are simply an inconsistent mess. What did the author want to tell us? We'll probably never know, as there were so many words from which to choose that he got all confused in writing this article.
But contrary to what the author seems to believe, standards are not about limiting choice, but about making choice possible. His example with MS Word is a case in point here, as the lack of a real open standard makes choice (in this case, choosing anything but MS Office) a lot harder then it should be or could be if there was an open standard.
If you want to remove choice, make your own distro with the choices removed. Make it so good that almost noone will think of using Gentoo, Debian, Slack, Red Hat SuSE or Mandrake for the desktop. That's the only way you'll succeed, otherwise it's a pipe dream. So yes, the way to remove choice is to wake Yet Another Distro.
Additionally, the writer seems to have little clue of how the Linux kernel works. Yes, it has a benevolent dictator that chooses for us, and decides some things don't belong there. Many patches will break other parts of the kernel, and they are not allowed in the main kernel. But we still have a lot of choice when compiling the kernel, between file systems, schedulers, firewalls etc. And if you don't like the default selection, you can patch in other things as well. Linus only chooses what is going into his tree. There are forks, and most distros use semi-forked kernels.
And standards. Standards remove choice, sort of, but only to help interoperability between different implementations. So we have standards of pop3 and smtp to send and receive mail with thousands of different mail-clients, and we have the standard of the English language to write different opinions on OSNews. But as long as my opinion is different from Adam Scheinberg's, his dream of removing choice will be just that: a dream.
Well articulated! However, to discuss choice, we also need to discuss value. Basically, two things are relavent.
1. If I want to create an alternative (and thus create a situation where others must choose), I have to make sure that I add value. Otherwise, I am just doing the same again and that's bad as it confuses choices. Mind you, you can add value by combining features from separate pieces of code into one piece of software. This links in to the idea of innovation.
2. Following on to this, what does the person make a choice perceive as valuable. When choosing, it's obvious that you choose the piece of software that best fits your needs. That is a direct function of the "value" of the software. However, it's expected that there won't be a single piece of software (say a text editor) that does everything. But there should be 2 or 3 max that between them, allow you to do everything you need.
I keep 2 text editors on my Windows machine because they serve different needs. On the other hand, I've seen some Linux distros with ridiculous numbers of text editors (I counted over 6 once, and I am not sure if I got them all). Why?
Choice is golden, but value is where it's at!
My 2 cents.
I couldn't agree more. As a M$ Windoze user of some time I have tried repeatedly to "get into Linux" with various distros in my very limited spare time. I can understand how new users are confused by the hundreds of options for apps in every class under the sun, many of which are still pre v1.0 stage yet. I just want an easy, cohesive gui with easy access to the command line, a good central way of controlling system settings, easy driver updating, etc. It would be nice if it was a whole lot more intuitive than it is now, especially for non techies. I was a huge fan of BeOS, God rest it's soul, and while it was never finished the way it should have been, why can't linux adopt an underlying ease of use philosophy that is similar? I think that would really help. That's my 2 cents anyhow. 
If you don't like any product (in basically any industry) you can either
a) Not use it.
b) Make it better (if that's an option....& w/ GNU/Linux [yes LINUX] it is)
c) Bitch about it
Too bad so most people choose c) 
Everyone knows that, for some reason no one does it.
Probably because it's too much work, to do a consistent GNU/Linux OS as many people want some nasty things have to happen, being the main one taking the control of GNU/Linux from the community, meaning that kernel, toolkits, desktops environments, key applications they must be under control of a single company or entity that will choose how things will be.
Nasty stuff, but until there no consistent GNU/Linux OS will exist.
There are distributions who bundle everything and let the choice to the user, there are other who only provide a limitated choice of applications, at least by default.
So the users have already the choice to get or not choice..
I agree, the problem the author describes is a non-problem.
Open source is all about freedom. Freedom means choice. Linux without choice is fundamently just wrong. That doesn't mean distro's can't choose their defaults. Which is the solution. So what was the problem again?
I think it would be useful to ask where Linux users want choice, and where they don't. Personally, I want to choose applications, but I don't want to choose system components.
To make that clearer, places where there is choice in Linux, but I *don't* want to choose..
Sound server. (ArtsD,Jack,Gstreamer etc...)
Graphics subsystem and driver. (Directfb,Xfree,X.org,Vesa,Svgalib,DRI,Xvid,etc...)
Package Manager.
(apt-get,slapt-get,portage,rpm,yum,etc....)
Kernel version and modules.
(4k/8k stacks,which modules compiled,2.4 vs 2.6)
Everywhere else in Linux, I like choice, but in these areas, choice seems to lead to incompatability and duplication of effort. I think it would be more useful to address areas where choice leads to problems, rather than where it might overwhelm a newbie by a 'surfeit of riches'.
I just don't understand how the author can equate having choices to being a bad distro. People mainly use Linux because they want to customize (ie have choices). In the past few years, I have installed SuSe, Red Hat and Fedora releases. Each is quite usable for the normal person, and what I mean by normal is Joe Sixpack. Each distro connects the dots with applications working together and defaults screens quite readable. If you have a problem with reading the fonts with Linux, try using an Nvidia card with stock settings under Windows!!!
If the author doesn't want choices, he should really move back to Windows or Mac.
That sums up the article. Incredible ambiguity when it comes down to it. "Remove choice." Why? What choices are good? What are bad? "Sensible defaults." That goes without saying! And the successful distributions are/will-be the ones that have such sensible practices.
Choice is very good. Choice based on standards even better. Hence this is what you should be encouraging, not the removal of choice where you think it poses a problem.
Throughout the community there are efforts to create desktop standards to improve the end user experience. Look at the fd.o - http://www.freedesktop.org - website as an excellent example. They are creating, defining, refining, and implementing standards continually in a collaborative fashion where any and all are welcome to input. Things like HAL and D-BUS are emerging. Gnome has already adopted these, KDE will follow. Many applications work well in both Gnome and KDE by adhering to many of the standards laid down at fd.o and more will move toward such compatability.
No, choice is good. Very good. And the more the choice, the more the market and need for standards. And those standards will come because, at the end of the day, people rely upon the success of Linux as a desktop and the Linux desktop cannot be successful without standards that are widespread and adhered to.
No, choice is good. Very good. And the more the choice, the more the market and need for standards. And those standards will come because, at the end of the day, people rely upon the success of Linux as a desktop and the Linux desktop cannot be successful without standards that are widespread and adhered to.
If choice is so darn good, than why hasn't open-source software penetrated the desktop market? I've been hearing the "it will happen" arguments for over 5 years now-- still, not much has changed. On a desktop, choice isn't good. It's a limiting factor. That, put together with the total lack of integration between various parts of a desktop-oriented Linux distribution is basically what's keeping Desktoplinux from succeeding at it's goal.
The best *nix-based desktop-OS (OS X)(**for the average user!!**) lacks the limiting factors of desktoplinux: lack of integration and overflow of choice. And lo and behold, there are no depenency problems, there is no need for complicated dependcy-resolving package managers, drag-and-drop is perfect, etc. etc. all those features a desktop OS can't do without.
For a desktop-OS, too much choice is bad.
I too think it would be a good idea if there were a formal Linux Standards Base of some kind. I think one of the bigger strengths of OpenBSD/FreeBSD is that it has a base system that is well tested and is consistent with every install.
As far as all the comments go about such a system removing your choice, I really don't see how that is so. And I also think such comments are completely ludicrous and that if the day comes when the major distros do conform to a Linux Standards Base of some kind that 90% of the people that are complaining now will see the inherit value and will become advocates of the "new" way. (The other 10% are hopeless, zealot rebels. But the beauty of Open Source is that there will always be a "rebel" distro for the toothless, unbathed, uneducated rednecks to get behind.) Everyone wins.
The whole article seems to be based on the faulty assumption that standards are indirectly proportional to choice. In fact in most instances I think it is the reverse. Except that OSS gives you the option of even opting out of the standard and limiting your choice.
He's not saying remove choice, he's saying that choice can be a good thing and a bad thing at the same time.
To give an analogy look at the language Perl.
It's been said that it's good points are also it's bad points.
Total flexibility is a good thing but it also gives the ability to write utterly incomprehensible code which is *not* a good thing.
If Linux is ever to get any market share on the consumer's desktop it'll have to be a very different distro than what's available now.
Nice article, shame that it's rather pointless. There already is a choice between distributions with sensible defaults and lots of stuff. And some of the distributions with too much choice still install "1 of each", or at least have a system-wide default (Mozilla for browsings seems common, OpenOffice for word processing, etc).
For example, I use a Mandrake with lots of bells and whistles, my girlfriend a bare Xandros with little "choice" but not as confusing as what I would install with Mandrake by default. That's choice, you know :-)
did it ever occur to you that standardization is something we choose?
I mean, you don't have to follow standards. You can do whatever you want but the most of use out here wants many things in common. Among these we want simplicity and comfortness.
How to achieve this may differ, but I think the most of us agree on that standardization is a good way to get that Just Works (tm) that we want.
Linux - in a general sense - needs a good starting point that doesn't call upon you to make a zillion decisions.
For a start, the question someone new to linux won't be what e-mail client to use. Go to any usenet newsgroup concerning linux installation, and you'll find the most frequently asked question is not about Evo or Balsa, it's about where to start: the choice for a distribution.
Though i will agree that this makes Linux difficult for new users, the existance of many distros is fundamental for the development of Linux.
The most important choice a distributor has to make seems to be the one between KDE or Gnome. But that is only because the other choices have already been made, not by the distributors, but by reality.
For example, most Linux distros i know use bash as a default shell. Not because someone told the distro's they have to, but because it apparantly is the most sensible one to choose. But should it be compulsory now? Should someone forbid distros to use or even ship other shells? Hmm.
Consider X. Had anyone asked a year or two ago, which distribution of X was the standard for Linux, almost everyone would have answered 'XFree86'. So, if there would have been a 'law' about which X to ship, it would most certainly have said: XFree86. I think most distros are glad such a law didn't exist.
And look what happened. Most distros switched, and they all switched the same way: Xorg. It still was a choice, but apparantly, some choices are trivial. And for the most important underlying technologies, the choice becomes trivial some day.
And consider this: who would even start programming a new mail client, if there's no chance at alle somebody will ever use it, because that would mean choice?
But what makes choice really good for Linux is the fact that if any distro thinks it can do better by doing something different, it can do so. That distro will act as a testbed for all others And if it is an improvement, then, as long as it's open source, any other distro can choose to copy it. And thus, Linux gets better and better. This mechanism is the strength of open source in general. And whithout choice that just won't work.
Call me a religious zealot if you like.
did it ever occur to you that standardization is something we choose?
Yeah, ok. How many implementations of securing e-mail are there now? I can think of SenderID, Domain Keys, and SPF without doing any research. Each supporter is trying to push their "standard." Standards are a long fought battle. What will the next version of the internet run? IPv6? IPv9? There are choices all around, and they all have MAJOR ramifications. Your precious standards too. Even the open source ones.
Some people here don't seem to have properly comprehended what I wrote - the article is about how choice is a double-edged sword. Not how we must sacrifice all choice. But I suppose if nothing else, my point is proven - take any group of Linux supports and suggest the removal of something, and in that zen-like way, they became samuari intent on proving you wrong.
How can choice ever be bad? Hmmm just kidding... It's true that a lot of people simply gets overwhelmed if they have to choose between multiple applications that perform the same task. The reason for that is simply information overload and if they actually wanted to spend the time to learn the advantages and disadvantages of a similar applications they would eventually settle on the one application they feel most comfortable with.
The elimination of choice before a user has had a chance to make up his mind is a bad thing imho. I also recognise that there are a large group of users e.g like corporate users that only would use a few specific applications and where the choice of similar applications in the distribution are not that important.
The main point is that choice is a cornerstone of freedom and I don't want to see a situation where people start to tell application developer X to stop wasting their time on application X because application Y has more features but only works in desktop Z, but who cares?
The author sure does not seem to grasp the concept of open standards or I don't. Standards don't just appear out of nowhere and writing an application that uses a good standard certainly is not detrimental to innovation.
Wow, good example!
You chose something that as of today has NO standard like "securing emails" (what's that supposed to mean, btw)...
Try to look at tcp/ip, http, HTML, XHTML, CSS, fd.o standards for interoperability for window managers, drag and drop, message passing, etc...
I can switch from metacity to sawfish 'cause there are standards.
patience grasshopper, as its a slow crawl. if someone dont want linux on the desktop then you cant force him. if you force him then he will just fight you even harder. people will understand in time
Standards are important. What would the world be like if we didn't standardize lights on a car? What would the world be like if we didn't decide that green means go and red means stop? What would our civilization be if we didn't decide what's a G and what's a T? Standards are what organizes the world around you.
I, for one, am very happy that I don't choose the standards around me. Yes, I can choose what car I drive, but essentially, all cars are the same. On all leverls of the universe, standardization organizes everything. No standards equals chaos.
And that's what the desktoplinux community is like at the moment. Pure chaos. One group saying A, another one B, and another one X, etc. etc. Chaos cannot exist. Chaos tends to organize itself. Just as the force of diffusion among molecules; molecules want to spread evenly accross the "space" they "live". so, even uncounscies things organize themselves.
The Linux cummunity refuses to do so. Or, at least, a group of people within the Linux community prevents it. No way Linux will ever succeed on the desktop without organizing itself.
But, I could find a quote that represents what I think about the subject:
"Everybody's a mad scientist, and life is their lab. We're all trying to experiment to find a way to live, to solve problems, to fend off madness and CHAOS." - David Cronenberg
It came from here:
http://www.nonstopenglish.com/reading/quotations/k_Chaos.asp
*Sigh*
I haven't seen one comment suggesting that standards are not a good thing, so what is your point?
And about linux and standards, you don't seem to have any experience with linux or you'd know that your claim that linux and especially desktop linux doesn't follow standards is simply absurd.
You don't even have to know anything about linux to know that, you just had to read the comments here.
And where do I state that I'm replying to any of the above posts?
What linux is, and no matter how many articles the media prints calling for the community to steal its own freedom the community will still hold onto its freedom.
And within that freedom people will individually do as they please, packaging over 250 distributions at any given time, half of which are barely not Debian.
If you don't like it, well too bad. It's like your neighbours barking dog, you'll have to learn to live with it.
this is tantamount to the same philosophies espoused by communist and socialist goverment. in their attempts to limit choice and prevent confusing the ill informed masses ... they inevitable get it wrong, and worse, stifle innovation.
the melting pots HAS to be a mess .. it has to be dynamic .. it has to be constant flux of old and new, good and bad, sane and insane, mainstream and fringe. and out of this pot will emerge the best ideas for that time ... and if the pot is maintained, these idea will be replenished with new ideas. very few innovations are timeless.
and if these is some confusion ... so be it.
the onus is on the masses to get educated. education is the key. not suppresion. democracy too is worthless if the individuals are too ignorant and lazy to elect an appropriate government (as we have seen recenlty in a notable case, talking of the president)
Any successful product class has choice. Even when you buy bricks, you can choose colors, size, and so on. They have to satisfy standards to be called bricks but you still get choice.
You get choice for fridges, car, hifi, airliners, riffles, cars, TV equipment. And this visionneer wants linux to be the exception ? Some of this stuff is complex, some is simple but choice is everywhere. There is even choice for DRINKING WATER !
Even if Linux had a fork, it wouldn't matter as long as they implement the same standards and interfaces. You have different web servers, different routers, different switches and strangely enough, the internet still works.
Every linux uers will tell you that once you've played with a couple of linux distros, you can cope with most others. So what the problem ? Most people using computers today got ever trained (if they even got trained that is) on win 3.11 or win 95.
Choice is GOOD. You don't have to run ALL existing distros out there anyways. And they all use the same components but put the emphasis on various areas. The industry has even come up with terms like LAMP. There are many implementations of it didn't you know ?
There are a few countries where choice is, by principle, the exception rather than the rule. And they are crippled : North Korea is the biggest example. So leave us alone. I am happy with my distro but delighthed that with what I've learned, I can try so many other distros now ! And I'll still be using kmail, mozilla, open office, webmin and so on.
He doesn't really make any revelations here and I think he's wrong.
The real problem isn't really choice, it's that the defaults for the two major WM/DEs aren't balanced and easy for most users. Gnome makes some easy things irritating to do, while it provides things like network monitors and other fluff that KDE doesn't have. KDE on the other hand let's me tweak everything much easier and doesn't make me want to rip it's heart out and steak it. So basically, if linux had everything integrated and fulfilled most users needs on a consistent basis, you'd have much fewer people talking about choice.
Choice is needed when you don't have really good apps. Good apps should appeal to the vast majority of people. While niche apps satisfy those with some small subset of functionality that is very important to them.
In short this author missed the point.
Take Ubuntu for example:
It would have never been possible without the choice which free software provides... for developers.
It takes away almost all choice from the user... and that's its biggest strength.
Because it's completely standards compliant and free, it will always be your choice if you use it or not... everyone should sing and dance.
You may call that a paradox, but I call that common sense. 
Everyone is going fine in the free software world as far as I can tell.
[The problem the author describes is a non-problem. Take any of the userfriendly linux distributions and the choice on which editor to use will allready be made for you. If the fact that it is possible for you to choose an other editor if you are so inclined is unbareable, please stay away from any computer, any TV and try not to leave your room at all.]
i could not agree more. it is indeed a non problem, this author tries to adress...
When will brain-damaged idiots like Adam get it through their heads that what is refered to as "Linux" is [...] created and driven by and for people who pretty much want nothing to do with the Adam Scheinberg's of the world.
I can't imagine how you could come to that conclusion honestly, unless you're simply trying to bait me, which I mostly believe you are, but I'll bite this, even though you are a blatant troll.
That would be a very bad idea for Linux companies to not try to engage people like me. Know why? I'm the IT Manager for a pretty large company, I administer hundreds of machines, and we've recently begun transitioning to Linux. We're even investigating a mass thin-client rollout using RHEL4 and the work from LTSP. I'm EXACTLY the target people Linux companies should aim for. I represent money, and you represent an exclusive, arrogant attitude that frankly, has help Linux back more than anything else.
You probably have a single unpatched install of Fedora running. Or maybe you're more the bleeding-edge, "emerge --update world" every week Gentoo type. Either way, it's clear from your response you have no clue how commercial Linux will succeed, and with whom it will succeed. Answer me this, people who happily flame me in an all too familiar chorus, how have you contributed to Linux? Have you written code? Have you ever purchased Linux? If so, was it more than a single box? Have you ever taught someone who uses Windows how to use Linux? I doubt it. If you'd ever worked with a real user, you wouldn't be so clueless.
All it means, this freedom to choose, is that we as individuals, groups, or even businesses have the ability to copy, modify, and distribute open-sourced software by selling it or giving it away. As long, as in the tradition of academia, we all have the freedom to access the source code.
It has absolutely nothing to do with religion, with politics, or any other "ics" or "isms". It is just code, much to the dismay of "redmondites", that is shared and developed by free-thinking individuals around the world.
The author is completely confused about what is a standard and how it is different from having Gnome, KDE and all the rest of the so-called "desktop managers". I don't think the author is a developer.
Standards are for developers, not end users. They actually allow developers to give a choice to the user, by creating different applications that stay compatible because they follow the standard of their purpose.
There are numerous e-mail clients around the world, even in the Windows world (yes, even there you have a lot of choices), what makes this possible is that they all implement standards, such as SMTP and POP3. What does SMTP and what does POP3 is of no concern for the end user, it is only for the application developper.
As for the problem of choice, I think the author's main difficulty is to believe there can be more than one way to use a computer. Besides, if he installs a RedHat distribution some day, he will observe that he hardly have to make choices during the process.
The magic of the FOSS community is that people can find the best way, for themselves, to use their computers. Some people will stick to the defaults of the distribution CD that's on their desks, while some other, even when they use Windows, will change everything from the mail client, network browser to the window-manager equivalent.
The thing is, Microsoft has no interest in people who like to use other softwares, so they will push people into the defaults (IE, OE, WMP). On the other hand, RedHat and other Linux distribution vendors have no interest in pushing you into one direction or another. Sure, they will make one easy way, but that's mainly for support purposes.
Quentin Garnier.
I just submitted a contribution to this thread only to find that it had been moderated down. When I checked the moderated down comments (there were six comments then) it seemed to me that most of them contained valid points about the article. I don't think that calling an article stupid in the context of raising real issues concerning the article is sufficient grounds to moderate down.
My own comment may have been a little off-topic but fell well within the boad area of discussion raised by this article. Read the first six moderated down comments and see if you agree with me.
"Choice" is not bad. It's a mantra for Linux and UNIX systems because they are intended to be adaptable to the needs of those using them. In essence, "choice" is merely customizability.
Where "choice" is bad is when it's made a distraction. While a data center might have specific wants that drive their selection and configuration, the same may not be said of desktop users. Indeed, desktop users come in many varieties. An engineer may want to customize his desktop environment quite specifically and use a number of different editors for different jobs. However, a person with little knowledge of computers who wants to surf the web and read e-mail but do little else has quite different needs.
The fact of the matter is that everyone wants "choice", but the less sophisticated consumer wants the choice to be made for them rather than make it themselves. From the standpoint of Linux, there are a wide array of distributions that run the gamut from "for idiots" to "for PhD computer scientists". Not that the "for idiots" distribution (like Linspire) cannot run all the same software as the system used by the computer scientist, it's just that it simply has made some semi-informed choices about what the average "idiot" wants and leaves it up to the user to figure out what to do when they want "more".
So, certainly Slakware is a bad choice for my mother. However, I know for a fact that Linspire suits her just fine (this is a woman that not only cannot program her VCR but actually shopped for VCRs without the ability to be programmed because she was afraid that she might press the wrong button and get the machine "stuck"). Same OS, different configuration, the "choice" being which configuration to go with.
For what it is worth, at home we use Mandrake Linux. I configured my wife's KDE desktop to look and behave much like the Windows systems she uses at work (though she's finding some very pleasant differences in the UI) and she's quite happy. My environment, on the same machine, is fantastically different. On her side, she has one menu item or icon for each activity, I have a list... Again, it's nice to have the choice...
It hits the nail on the head. And apparently a lot of people too! Ouch!
Like Nicholas Blachford, I'm not sure those who rant here have really read the article.
Strategy of the article: Create a false dichotomy between choice and standards in very general terms and then argue that the solution is to remove choice by leaving the poor user to fend for himself by compiling what he needs.
No, that is not the solution. The reason why people pick a distribution is because the distribution is supposed to make certain choices on your behalf. So if you set up Mandrake's discovery edition which is what you would give a Linux novice, you only have a best-of-breed application. As he learns more, he then realizes that there are other apps he might want to try, but instead of going through the horrible pain of compiling software, the new user users the software in Mandrakeclub's mirrors or in contrib/plf to install software easily. More importantly, this is software that has been packaged that is known to work for his system.
This flexible approach allows the new user an easy path into Linux and allows me to select the packages that I need for my daily work, packages which would not be installed by default by the distribution. That way everyone wins.
I swear that this site is the epitome of controversy for controversy's sake. I have never seen a web site so focused on meaningless polemics and on dividing the free software community by raising meaningless issues of contention that are not there upon closer examination.
Do you realize you're agreeing with the author? Yet you think he misses the point.
Thanks Ralph - I think I shall repost a Bowdlerised version of my original post and see what happens:
In Praise of Linus
After reading the flawed arguments in this article, so ably refuted by hanky, it made me realize who importance Linus' contribution to the methodology of software development which is just as if not more important as his development of the original Linux kernel.
Linus was the first person to actively and consciously apply Evolutionary Theory to software. Modern Evolutionary Theory deals with the change in a replicator (DNA in biological systems) that is variable and is under selective pressure. Dawkins was able to extend this idea into the field of human culture with the concept of memes as a replicator. Linus recognized that computer code was an evolutionary replicator (a form of extra-corporeal meme, possibly the third replicator) and that he was shepherding "a herd of cats" in the evolutionary development of the Linux kernel. Evolutionary development like this is based on freedom and choice not on the dictatorial decision of some "chief architect" who knows best as with MS and proprietary software.
The anti-evolutionary control freak approach in part extends into the free and open source movement. For example the some software is developed under a much more controlled environment environment where "software design" is considered to be paramount and whose protagonists denounce the anarchy of Linux development and also deny that software development parallels biological evolution.
Nice article. I agree that most distros should have a default but I disagree on not needing choice.
Why shouldn't I be able to customize my user experience as I see fit. Why should I be locked into using a product when there is something out there that I believe is better. For the novice leave them with the defaults. As they get experienced then they too will love what I have come to love.
Besides choice means competition and competition leads to a better product.
Peace out:)
Please don't moderate this down, it's insight into this issue. Adam, your project called flip (flipsource.org) is another CMS. There are many CMSs out there like you say in http://www.flipsource.org/faq.php?section=project#3 - but there are also flat file versions too.
Tell us why you have this strong vocal opinion against choice but choose to express your freedom in providing another choice in content management systems? Does the problem only happen when the software is packaged together in the form of a distribution? What is your real issue here?
I think the abundance choice is wonderful in most arenas....except in the case of marketing of a product that already has too much saturation. Don't get me wrong, giving customers a choice of colors in a Gamecube or offering 130 different consumer options on an Audi A8 is absolute bliss.
But I think the point of the article was not geared towards IT professionals like we are (or may not be, don't know your background) but more towards the general populous.
Zealots want the whole world to run some version of the Linux kernel for some strange reason. Linux is not going to take over the world, a good Linux marketing and distribution company will take over the world. They'll give people direction in the Linux world, they'll make it easier for people to use Linux just as MS made it easier to use the pc.
I use Linux but sheesh, the world needs direction. Many distributions now moderate how much software gets included in a default configuration which I think is great but there much more to do. Then again, I don't want to be limited to just using Linux....
I don't usually come to this site, but what entertainment! The author writes an article about, as he wirtes in a comment, the double edged sword of choice (Adam: I found it clear that you were making this point). Then all but about three comment miss the "double". Half the readers think he's criticizing side A of the sword and the other half think he's criticizing side B of the sword. I guess it was the "choice" of how/if you want to read the article.
I think a lot of people replying missed the point. Choice - yes, and lots of it - between distros, but not WITHIN a distro. Your distro should have direction, purpose, distinction. Providing 5 text editors does not provide any of these.
I made the same argument earlier this year in an OSNews article about Slackware, and I stand by my statements. Pare down the choices installed by default and provide an easy way to install them later. You could get most distros down to 1 CD that way. That saves bandwidth, enhances usability, and a host of other good reasons.
I don't advocate taking away choices, just streamline them.
Joseph,
I cannot understand why there are a few people here just DYING to take me down over something I am NOT arguing. You said, "you have this strong vocal opinion against choice." Eh...what?! I did not say that I am AGAINST choice. I said that in any implementation of something, you have to limit choice. Because too many choices, or uninformed choice, is bad. Read up on Flip. If you like it, feel free to use it. If not, move on. Hell, write me and tell me it sucks for all I care. But you won't get confused, because you won't find it if you're not looking for it. Doesn't change the fact: too much choice can be bad.
Flip is my own thing for my own entertainment, and I don't pretend that it's more. But if you visit this page (http://www.flipsource.org/viewcomments.php?cID=1095194672), you'll see that exactly as I stated above - too much choice obscured the goal, and the entire project was scrapped in favor of something more directed.
Here are some examples: Photoshop is absolutely intimidating to non-technical users. You might say "Don't use Photoshop." But in real life, that translates into customers, into SALES. There's a possibly benefit to a less cluterred interface, possibly with fewer options. Less intimidation might mean more users. That's the less "choice" I'm referring to.
If something works for you, fine. But please don't pretend that I have some wacky agenda to limit everyone's choice. That's just silly, and it isn't what I wrote.
The author of this article makes it clear that he knows there are both good and bad aspects in having choice. Yet he also makes clear his own position: "If you ask me, I'll tell you straight out: I'm all for removal of choice from Linux."
Some people here don't seem to have properly comprehended what I wrote - the article is about how choice is a double-edged sword.
Your inability to make yourself clear is your problem. Don't try to make it ours; such attitude is arrogant, insulting and condescending.
Not how we must sacrifice all choice.
Oh, so is that why you have sentences like "If you ask me, I'll tell you straight out: I'm all for removal of choice from Linux." scattered all over your article? I'm sorry, Adam, but your sudden backpedalling is not only unconvincing, but once again, arrogant, insulting and condescending.
But I suppose if nothing else, my point is proven - take any group of Linux supports and suggest the removal of something, and in that zen-like way, they became samuari intent on proving you wrong.
Wrong, Adam. All it proves is that if you phrase your article like a flame-bait, the flames will inevitably erupt. Had you written a truly well-thought out piece of journalistic work devoid of flame-baits like the one I quoted above, you wouldn't have gotten so much negative feedback.
Really, stop blaming your readers for your own journalistic incompetence. And, while you're at it, quit hiding behing the "I'm modding down every comment that someone reports that contains a personal reference to me" tosh. Your article contains a number of ad hominem attacks on people who disagree with your position, so it is rather hypocritical of you to demand of others something you yourself cannot do.
Now, go ahead, prove my point and mod me down.
In the linux community, it's argued that choice is a good thing.
And that standards are a good thing.
How can you have it both ways?
These two beliefs reflect and underlying belief in the architecture of software. It's the kind of architecture you see in network stacks that makes the internet work as well as it does. It's the kind of architecture that lets you take the same code for linux apps and build them on BSD.
With the way the internet is designed, you can be connected whether you're on a FDDI connection, an ethernet connection, a wireless connection or any other kind of connection you care to think up. All these different technologies converge at the IP layer - that's how things get around the internet. Above the IP layer, you can talk TCP, you can talk UDP, or you can talk any other way you decided to make your app communicate (as long as what's on the other end can understand it, of course). The standards for networks and internetworking have led to a lot of choice and availability.
POSIX standards, C language standards, HTML and CSS capability and rendering standards - they help PROVIDE choice, not take it away.
Look at C language standards - because they guarantee certain things about how the code is compiled, you can write language bindings so that you can write parts of applications in one language, parts in another, and have them work together. Take away that standard and see how much choice you're left with.
Now, go ahead, prove my point and mod me down
I would in a heartbeat. You're adding nothing constructive to this thread. If this thread even closely resembled a discussion and not a flamewar, no one would be reporting the posts as offensive. You've just spent a few minutes typing a lengthy or formatted comment that violates the osnews terms you agreed to by clicking "Submit Comment." Need help finding that?
http://www.osnews.com/rules.php
2. No personal attacks on story authors, other commenters or news editors of this web site.
Want to talk about choice? Do it. Want to give me a hard time? Write to my e-mail account, available above to ALL, and do it. I'm not afraid to defend my points or myself, but I'm not going to get personal in a forum. Unlike you.
I think the problem was how you wrote what you wrote. I guess if one considers the title "Paradox of Choice" they could understand you're arguing both sides. The above example in your reply is a good one. Taking it into consideration you mean less exposed choices instead of less software choices. I agree with less exposed choices. But I heavily disagree there should be less software choice. As was pointed out, this is somewhat of a non-issue.. you can choose a distribution, window manager, desktop environment, terminal, or even editor with less exposed choices. This is much like the differences between similar software.. GNOME vs KDE, Windows vs MacOS, Wordperfect vs Microsoft Word and so on.
There are all kinds of choices to be made, no matter if it is OSS or non-OSS. Just remember there are different levels of choice and thus there really is no/very little of a paradox. You're comparing apples with oranges when referring to software choice and exposed choices in software.
The thing the trolls seem to be missing here is that what this author is about is not to remove Gnome from the surface of the Earth in favor of KDE. Nor is he about giving you a gaa-gaa-goo-goo baby interface to your favorite distro. He's talking about the BASIC package, you know the smallest set of things one needs to do one's job? Userland Linux is after that idea, and yes, many other people are doing it right now, so we're still going to see some splintering of default choices for a while.
But being able to install KDE on a Gnome-based distro is totally different from having to choose between the two. Look at it this way: on Mac OS X, the GUI application to use to manipulate the filesystem is the Finder. And that's it: _The_ Finder, period. Yet, when was the last time Apple lobbied Cocoatech to cease and desist from doing PathFinder, an application that replaces your Finder? When was the last time they screwed their filesystem or their frameworks to break compatibility? And when was the last time you installed Mac OS X and you had to select between 6 flavours of the Finder?
The point is twofold: first the KDE- and Gnome- teams should stop rooting for domination and commit themselves to interoperability, to the point that their applications interoperate in STANDARD ways with each other. Second, non-general-purpose Linux distributions should commit to a specific, cohesive set of applications. General-purpose systems like Deb, Slack, Gentoo, Fedora should be seen also as PLATFORMS. And again this is what UserLinux got right: instead of trying to cripple and reduce the near-total coverage of the Debian system to make it a simpler distro, just use the actual Debian system as the fertile ground on which to grow nice little User tomatoes. When a distro becomes as large and as all-encompassing as Debian, it gets nearer to becoming such an object as a standard, not a user product.
BTW I'm using the UserLinux example out of ignorance for the other equivalent projects.
Standards vs choice are abstract terms which are indeed somewhat contradicting. I don't think the author meant to say 'choise is bad' though. He wanted to show an inconsistency.
However, when you've drawn you can use that inconsistency again to draw a different point also, the author forgets "choice BASED on standards".
I've always felt as though I don't want to spend my time trying to decide which word processor to use. I don't want to flip flop between Kontact and Evolution, Gaim and Kopete, Abiword, KWord, Open Office.org Writer, etc.
The point is, that when you chose for a DE, you don't have to chose because there are related applications for that DE and that toolkit.
Ofcourse, then some people say: But i don't want to run a "QT application under GNOME because [...]" it doesn't look nice, or costs more memory. Here the "improving interoperability" applies. 2 nice solutions are in the works: making QT look like GTK2 and vice versa (e.g. RedHat did this) or porting the application over to the other DE (e.g. Gecko to Konqueror, GIMP to KDE).
And a design from the grounds and agreements are important like e.g. GNOME HIG does. Where techies decide for users to chose good options most will like, whereas still leaving he oppurtunity to modify the default. The default has to be good for most people. Because sometimes people don't want to make a choice. Instead they rely on the standard or default, trusting that's good enough. A form of laziness. However if they don't like the default they want to have the ability to chose. Distribution's diversity also is a strength in this aspect: they can try to reach a nieche, or try to create a unique environment.
Note the difference between "freedom of choice" ad "freedom to make a choice"...
lame comments. A distribution should have a limited software set that furthers the well-defined direction of the distribution. Choices in a distribution shouldn't be 'everything we could find a freshmeat' or immediate at time of install, but should all be available afterwords for any changes you want to make. It's just more sane that way.
Exactly my point! Providing choice based on standards. Great description and examples btw.
"I don't advocate taking away choices, just streamline them."
Agreed, but it needs to be done by using a progressive, fruitful way. Do you have suggestions?
I was thinking about 2:
1) UI experts analyzing which application is better (features are more than not obvious already to anyone reading. Though, add that if you prefer.)
2) Constructive feedback from users. Take part of the Free software community!
3) A platform where people (e.g. those from group 1 and 2) come together to develop the discussion.
The author doesn't know what he's talking about apparently. Standars are the only way to make choice available, communication between people and programs are the biggest necessity today, and that's what standars are for, if you follow the authors conclusion to the end, we must all use the same OS, with the same apps, and the ones that chose other choices, shouldn't be able to communicate in any way with "standard" users.