Linux's place on the personal computer

Something has been bothering me for some time, but I have kept relatively silent on it. When I first "joined" (that may not be a correct term since I am just a user and not a contributor) I noticed there seemed to be the views:

1) That people should learn how to use a computer and not that a computer should become simple enough so as not to need a steep learning curve.
2) Linux is not for those not willing to learn.
3) Linux is not for the masses.

I guess I am wondering where all this is going? You have Windows which is the market leader, with MacOS making its biggest challenge in over a decade. Linux, the free alternative, is having a hard time making headway.

In mobile phones, iOS and the Android OS, went from nowhere to being the market leaders in less than three years. After seven years, I am using Linux again, and it is better than ever. I though it was a viable alternative seven years ago. We actually deployed Linux and OpenOffice.org in a friends internet cafe, but not for all the units. Even some Windows PC's were were running OpenOffice.org instead of MS Office.

Computers are now a necessity, for people who do things other than there life other than having to learn how one works. But there seems to be a resistance in the Linux community at attempts made to simplify or "dumb" down things.

Should a person have to be a mechanic to be qualified to drive a car, understand how a TV works before the use one... ultimately we use plenty of devices in order daily life without understanding how they really work. Why should a computer be different?

 

There are distros that cater to the users that do not what to learn how a computer works and just want to use them. There is a whole spectrum of distros that range from those that make it dead simple to those that are really challenging. Use the one that fits your particular need, taste, desire.

Nevertheless, a little word of advise is in order for those that want to make the switch to Linux a smooth one:

How to make the switch to Linux An alien’s viewpoint

 

But sometimes I feel like they are treated as Linux distro's that have betrayed the faith or something.

Unless it fragments itself to death.

 

[delete] as might make too many a nerd rage!

 

All three of those statements could be excerpts of my posts in this forum.

I disagree. Windows and MacOS have a big market share, Linux doesn't - that's right. But the point is, Linux (or better GNU) as a whole doesn't try to gain something like market share in the first place. The only exception to that rule might be Canonical (Red Hat and Suse don't count because they don't aim to put their professional Linux distros on the end user desktop market).
If you don't try something in the first place you can't have a hard time achieving it. GNU/Linux as a whole ecosystem is not commercial, so the measures we usually apply don't work. Basically it's all covered by #4 and #7 of [1].

Because this is a completely different market. Windows is strong on desktops because of its history. Most people today use Windows because they've always done that. Mobile devices is a whole different ecosystem, so Windows' biggest advantage on desktops doesn't work here.

In parts of the community that's right (and I consider myself to be part of that). The reason is #4 and #7 of [1] again. The strength of GNU/Linux is that nearly everybody who uses it is an expert somehow. Before Ubuntu there were nearly no Linux users who were to lazy to read manpages of do some research on their own. This is changing slowly, and I think this is bad. I expect people who ask for help to be interested in how things work up to a certain point.

Up to a basic level - yes, I think so. At least when this somebody is asking for voluntary, unpaid support. If you pay me accordingly you can send me the most ignorant users to solve their problems. But if they want it for free and expect me to invest my spare time on their problems then I expect them to help me to help them.

It isn't different. The expectations of how support should work are different. Nobody would expect a car mechanic to work for free but it's expected from computer supporters.

[1] Linux is NOT Windows

 

Canonical is trying to address that issue by making linux for the masses, but in my opinion fails at it miserably, resulting in buggy, bloated distros every 6 months. To that end I still use ubuntu 10.04, and probably will for quite awhile.
I for one am glad linux remains in relative obscurity, or rather just out-of-reach to the masses. I like using a niche product that isn't plagued with malware and viruses because it's so popular like windows and mac. I hope that doesn't sound smug as I don't intend it to be, but it's how I really feel.

 

I believe this is the first time I have completely agreed with debguy. Well put.

 

I have heard the same sentiment before, when I was at the Mandrake Forums, and a Red Hat forum.

I guess that it the part I do not understand, I would not have come back to Linux last December without all the help you gave me here. You really helped me out a lot. I did not hear "Google it" although I realized later on I could have "Google'd" most of it

Fair enough.

That is where I get confused. I would assume a Linux user would want it to get more mainstream. On the one hand, I am happy using it, and why should I really care who else uses it. My last Linux desktop install was Red Hat Bluecurve, seven years later I have Fedora 15. It has not died for lack of users, so there is not reason it should at anytime in the foreseeable future.

But there where always be several Linx distro's... shouldn't those who support the more high level distributions, support the "dumb downed" ones for the rest of the world?

I guess there is this part of me that feels since I am getting something for free, and like it, I should do what I can to help those who built it.

 

you might be mistaking the idea of 'free' with FOSS.

 

@naticus

No need to be so polite. I do understand that "free" is not FOSS in the purely technical sense, but I think I don't understand really understand FOSS as a concept.

 

The point is that when you asked for help you gave others the impression that you were really interested in the backgrounds and that's what I was referring to. You might have needed a helping hand, but you were always walking on your own two feet, and under these circumstances I like giving support.

I want so much mainstream that enough developers are attracted to maintain and improve the quality of the software I'm using. Any more mainstream beyond that doesn't matter to me because it won't improve my Linux experience.

Exactly!

That's the case. Debian supports Ubuntu by giving it the base system. The same does Red Hat for Fedora and some others that evolved from Mandrake (Mandriva, Mageia, PCLOS). The same goes for Slackware and Gentoo.

That's a very good idea. Step one is to give help to people who are less experienced than you. Step to would be to write constructive bug reports. Step three would be to become a developer on your own.

 

scripts as well are also important, not so much on the implementation dev level, but a building block for 'post' installation. Crunchbang utilizes this as a forefront for those that cannot get much past 'the door'.

scripts are always welcome, but must be very-well implemented to the dev, and philosophy per distro.

 

Thanks, but you give me a bit too much credit. I can get it set up, can follow instructions, understand administrator privileges and the like... but I do not think I am walking on my own two feet. I rely on what others have written. Best I can do is search a bit before asking. No point having a person write a guide which you do not read and ask that same person for help in a forum.

I can do 1 and 2, number 3... that is way over my head.

I am completely lost.

I hope I did not offend anyone with this post. I just really do not understand. The Linux community is probably the most helpful on the web. I do not know why you all do it, but thanks again.

 

First of all, I don't really see the issue with Canonicals 6-month release cycle, especially considering the functional improvements in just the last year. You still have the option of a Long Term Support release, but there again, I hear LTS users complaining about 10.04.2 not having all the features of 11.04.

Hey, I'd rather have more hardware and driver support than the exclusivity of using a minority operating system. As it is, there isn't really a viable mass market for preloaded desktop Linux, mostly because the desktop user base isn't there. Sure, there are people who use Linux to breath new life into older Windows boxes and there are die hard developers, but we also need users of both categories to be substantially outnumbered by casual users, namely casual users willing to buy a notebook or desktop with Linux preloaded.

As far as security threats, Windows has been uniquely vulnerable because it is fundamentally different than any Unix derived operating system. Widespread adoption of Linux won't make current users any less safe, much in the same way that increasing marketshare hasn't sudden undermined the relative safety of OS X. Now that Linux dominates the server market, are we all suddenly less safe? No, not at all.

 

The problem is not the 6-month cycle, it's the absolutely fixed release model that forces them to release even half-baked software at the end of april or october. Bugs don't care for shedules.
I rather wait another 6 months or even a year for the next Debian stable release than having to worry about teething troubles.

That's the first valid point for a broader Linux user base I've ever heard. Thanks for that!

I don't think so. Linux is not inherently safer than a modern Windows after all I've heard. In fact, if I were able (which I'm not) to intrude a random Linux, even without root privileges I'd know how to do a lot more damage than on a current Windows. Of course part of the reason is that I have absolutely no idea how a modern Windows looks from the inside.
Also, in the german Debianforum there are threads from time to time dealing with hijacked Debian servers - mostly due to PEBKAC on the admin side.
I think due to its market share Android will be a good measure for the real safety of Linux. Some people expect it to be the "new Windows" in terms of being a target for malware of all sorts.

 

I found myself waiting a good 6 months for the Linux kernel to catch up with my hardware, but it finally did. Similarly, Broadcom only made their drivers available in the last year. All too often, the real "teething troubles" aren't related to the distro but bugs that can only be solved through kernel development or the release of previously proprietary drivers. For the inexperienced user, Linux is much, much more approachable than it was a year ago. I don't see any downside to that.

I also don't think that Ubuntu's releases are especially rushed. It's important to set goals within an organization and Canonical has done just that. Now, it's obvious that Ubuntu builds on Debian, and that's a good thing. On the same level, Ubuntu might be built on the work of others, but the result is uniquely accessible to novice users.

Wow? I can't believe that? Really? I've discovered the first ever "valid point for a broader Linux user base?" Should the Linux user base be confined to only hardcore developers? That's a bit like opening a comedy club that's only open to established comedians, but excludes aspiring comedians as well as spectators. I don't see how that's sustainable.

Ah, yes, the theory that closed source software provides the ultimate in security. I think that the geniuses at Microsoft have debunked that theory quite nicely.

Still surprisingly rare, considering the absolute dominance of Linux in the server market. It's not that there is any lack of valuable targets running Linux, it's just that the vulnerability isn't there. Why would spammers gather huge "botnets" of Windows PCs when it would be far more logical to gain control of much small number of Linux email servers? It all comes down to the inherent vulnerability of Windows.

I'm still not sure that Android is "the new Windows," in any possible sense, positive or negative.

 

I gathered the opposite experience. Most of the teething troubles I've encountered were due to poor configurations caused by insufficient testing prior to a release by the developers of a distro. That goes for Ubuntu, OpenSuse and Mandriva. On the other hand I've never had kernel-related problems.

I don't see the point you're trying to make. That statement has always been valid in the past 20 years.

Did you notice that in the last 3 years every Ubuntu release (except for 10.10) was towards the end of the month? I get the impression that the developers say "We need more time!" but the marketing says "No way guys, the month is over. We have to release NOW!"
Setting goals is fine, but achieving them is a whole different story.

I said it's the first good point I've heard. Neither did I say it's the only one, nor that it's new. All the other "reasons" I've heard like "With a bigger market share Linux can kick MS's butt." are just nonsense.

I never said that the Linux user base should be confined to hardcore developers. I said that a community-driven project like GNU/Linux doesn't profit from people who don't contribute to that community.

And again, I never said that and didn't mean to imply that. But the conclusion that Open Source is inherently safer than closed source is wrong too.
Since you seem to have quite some knowledge about Linux and OS in general I can't shake the feeling that you're intentionally getting me wrong.

The potential for PEBKAC is simply bigger in the sector of average desktop users than in the one of server admins. Therefore desktops are an easier target.

 

Still here, listening and learning.

 

1) Sadly this is the treand OS' are heading. Look at windows, osx, and ubuntu. Sad but true and why I use debian over any of those.
2) This is why its not taking any serious market shares. People like being ignorant and dumb. Another sad reality, but people just want to plug it in power it on and only require an IQ of 50 to be used. Linux requires at least an IQ of 70 to be worked properly.
3) True. Linux is for the hackers/modders, computer science types and those who seek knowledge and enjoy learning. (thats not to say all other OS users are not)

Linux has its place. On the above list of users computers and servers. It will never be main stream as 99% of the worlds computers come shipped with windows or OSX preinstalled. Why would the lamen go out of his/her way to install a new and seemingly difficult OS when they for the most part are computer illiterate? Sad but true.

The flip side is that linux still caters to those of us that like all levels of computer use from compiling binaries to watching youtube videos. Linux does this perfectly and Im glad. If linux were mainstream it would end up like OSX or windows and that would NOT be good. I love linux for what it is, does and suits me perfectly.

 

The Dell E6410 i915 bug did exist and was solved by more recent version of kernel - namely 2.6.38. This precisely why I can't use Debian, which appears to still be using 2.6.32.

Or in other words, you want GNU/Linux use to be limited to a small community of developers, the only people who "contribute" to any FOSS community?

 

Yes, but why does any Linux distro have to be "seemingly difficult" to install? Ubuntu is dead easy, and has gotten a lot easier in the last year. I've installed Ubuntu in the most minimal install from the command line, which is easy enough, for me at least, and I've also used the graphic installer, and quite frankly, there's nothing wrong with that. Easy is good.

Why not encourage Linux go mainstream? "If linux were mainsteam it would end up like OSX?" Well, that would be a good thing. Apple took BSD as a base and made it an increasingly mainstream success, and at the same time BSD itself is increasingly marginalized. I don't see anything wrong with making Linux easier to use. Why "compile binaries" when you can use the brilliant "Ubuntu Software Center?"

 

OSX progression has taken what was good to a progressively worse OS as time goes buy. OSX is similar to linux, but really is just a gross mutation of the source.

Mainstream ruins OS'. Linux is great because it still caters to those who prefer it, rather then everyone and their brother...

 

I don't see why both modes of Linux cannot be mutual. i mean if you want your linux system built the way you want use Debian, Slackware, Arch and build it from the ground up. If you want a premade system choose one of the plethora of Linux version out their -- *buntu, crunchbang!, Fedora 15.

Why does it matter which path someone takes in choosing their preferred system and the way they want to undertake it?

To each his own.

 

You can install ubuntu, crunchy, fedora, suse, etc from ground up too. Do a net install, and install one by one, you get the system built the way you want.

Those "easy" distros can be customized too. The "hard" distros on the other hand don't have "easy" modes.

 

Just install the backports kernel and you're done.

No, I don't. But I think these should be the only ones to make decisions.
If some non-developer (e.g. me) likes to use some flavour of Linux I see no reason why he shouldn't do that. But somebody who doesn't contribute to a FOSS project shouldn't complain about this or that missing feature. If you want a feature to be included, do it on your own, pay somebody to do it for you or kindly ask the developers to do you the favour (which differs from complaining). If none of those options works for you because you neither have the skills nor the money nor the benevolent counterpart just leave it be.

 

That's just a logical fallacy. If you're right you're right, if you're wrong you're wrong. Whether a coder or user.

For example, about 5 or 6 years ago I got into an argument with the chief hacker for the PostgreSQL project because the database was taking dates like 07/13/09 (american date style) and if your date style was european, it would flip it around to 13/07/09 so that it would fit. It was also doing things like taking ISO dates like 2009-13-07 and instead of throwing an error was turning it into 2009-07-13 so it would fit. We agreed that the ISO data mangling was absolutely bad behvaiour, but he disagreed with me on the other format mangling being bad. My point was that it's not the databases job to guess what I meant and mangle the data to fit. Either it fits as is, or it throws an exception and you find out your data is mucked up and fix it. I eventually won that rather polite argument (the only one I can recall winning against him) and the change was made. If users don't report bugs or ask for new features, then often a project gets old and stale. Chief developers make decisions in a vacuum void of user experience and needs.

The users are just as important as the coders. They may not be privileged to make demands, but requests are wholly welcomed and in fact needed to keep a project vital and relevent.

 

Couldnt agree more.

 

Nope. Installation would have been extremely difficult with an older kernel, especially using a graphic installer, since the bug was with the Intel graphics. Some people had to use an external monitor to do an installation.

So, yes, it does matter that the Debian community doesn't stay current with kernel development.

So you're claiming that non-developers just don't matter? This is a recipe for decline and irrelevance in any community.

 

I really like that we have so many distros to choose from - we can stick with a distro of our choice. But it is when the distro you loved so much makes a huge decision that you may not like - you go hunting for a distro of your choice again.

I honestly will stick with a distro which provides fairly new version of the packages, a powerful package manager, good support for CLI based configuration tools and a good supportive community.

Unfortunately the PC market consist of a large number of people who just install and forget - they only want their system to reliably work to use their favourite apps, games and such stuff - they dont use it to try out new and exciting stuff.

 

In servers I'm usually far more conservative with regards to installing newer distros and stick to mostly two or three year old versions like RHEL5 or Ubuntu 8.04. But recently while building a db server we needed a kernel that ran well with 48 cores, and those older distros are not very good at utilizing > 16 or so. So I had to [ut Ubuntu 10.04 on them. But man, the performance difference was HUGE so it was worth using something newer. And as much as I've liked RHEL for dbs in the past, I was NOT putting my production db on Fedora Core.

 

Non computer "types" can use certain linux distros. You don't have to know how to build an Airplane to fly one.

Like someone said, Canonical is doing us all a favor by popularizing it. More hardware support = awesome.