“Even though I enjoy using my Linux boxes a great deal for the most part, there are still those little annoyances that are enough to drive a guy batty. After all, even something as elegant and as effective as the top Linux distributions can still do certain things to drive you nuts from time to time.” More here.
The Top 5 Things I Hate About Linux
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSAlert.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
So, the authors hates that people do and think stuff differently and don’t share his preferences? Tough. Oh, and hardware issues, boy, does he hate them.
This is why the Linux community makes him hate linux more than why he hates linux. I understand the connection though. A major part of using an OS is getting bogged down into its community.
There is a tendency for applications to get half written, but this is hardly a FOSS issue: This is a software development norm. About 5 minutes on download.com will show you how much software gets half written before the developer gets busy, gets distracted, or forgets.
The driver support is an issue, although I don’t blame linux for it. But it’s why I’m sitting on a Mac and using linux via parallels when I need it. I didn’t want to buy a half-working unix notebook, I wanted a wholly working unix notebook.
The hating on proprietary apps is definitely strong with the linuxers. But it’s not everyone and it’s not as bad as it was even 3 years ago.
Edited 2007-04-05 19:06 UTC
Of course not.
I mean I’ll probably be buying a couple copies of Nero Linux 3.0 to support a commercial company interested in Linux and get decent CD/DVD burner software thats possibly better than the FOSS alternatives, such as GnomeBaker or K3B.
I dont use Linux so much because its open source or free as in freedom. I use Linux because its free or cheap and just a nice Unix-like platform to have.
Perhaps the same thing can be said about OS X, bit I wont be buying an overpriced Mac that have potential serious defects.
http://www.appledefects.com/
I dare to tell you that if it was not for people like you there might of been way more open source drivers and/or documentations for writing such drivers in the hands of the (linux) kernel developers.
If you agree to use binary blobs the manufacturers have no incentive to change.
That is just my opinion the same way you have yours.
The issue he stated wasn’t about open source or proprietary drivers, it was the lack of ANY driver.
(no support at all from some manufacturers)
If you look at the Nvidia and Creative X-Fi drivers for Vista, I think “no driver support” is better than “bad driver support.”
I think more often than not people react to lack of Linux drivers for their _prefered_ brand/model of peripheral, because there’s almost always an alternative hardware that does have Linux support.
You obviously don’t work in pro-audio.
I love POSIX systems and abhor Windows and Mac OS but it’s almost impossible to capture multitrack 24/96 sessions with Linux. There are drivers for some now out of production units (like the RME Hammerfall series) but there are essentially zero acceptable options for professional sound recording on the go. Nothing rack mount works. Nothing works with a laptop. MLAN will probably never make it to the 2.6 kernel (you can patch a 2.4 kernel to get some functionality) and no one with a proprietary communication protocol is going to opensource it.
Hehe. If we’re forced to buy alternative hardware just to run Linux, doesn’t the whole advantage of cost evaporate? Might be why sometimes it’s costlier to switch to Linux in some cases.
Also, laptops outsell desktops nowadays. It’s a little harder to swap out hardware.
Also, laptops outsell desktops nowadays.
where did you read that ?
our company regularly places orders for 2500 desktops, and around 15 laptops.
our local shops have people coming in in their droves and leaving with box loads of components for their new desktop, and sell maybe 2 laptops a week.
even in our local pc world the desktops outsell laptops like 4 : 1
Let me clarify. In the United States and other economically similar countries laptop sales versus desktop sales is about 50/50. Check these links:
http://www.engadget.com/2005/06/04/laptops-outsell-desktops-for-the…
http://www.itwire.com.au/content/view/3572/0/
The first link was posted back in 2005, and I read a more recent story but can’t find it… will post it as soon as I find it. With the laptop market experiencing double-digit increases (22% in mature markets), desktops experiencing near double-digit decreases (8.6% in mature markets), and monthly laptop sales topping desktop sales multiple times, I really doubt if desktops outsell laptops 4:1 today.
ThanhLy said:
So my POINT was that laptops are a significant market, and that it’s more difficult to simply swap out hardware. So, maybe ThanLy’s advice is unrealistic for a very large portion of Linux investigators.
@gtada
I can see how laptops can be a sore spot for Linux support. But to be realistic, here’s what Linux laptop support looks like today:
– A small handful of laptop vendors selling units with Linux pre-installed.
– Numerous websites specific to Linux on laptops with a database of hardware compatibility, and tips for installing Linux and enabling certain features for specific brand/model laptops.
– Off the top of my head I believe there’s only 2 hardware issues: WLAN card and power management. The WLAN card is something that can be alternated as I originally suggested. Now that there’s an open source Broadcom driver I think this is less of a problem. Regarding power management I just don’t know enough about the current status to comment. Windows isn’t exactly perfect in this department neither. Remember the buggered USB driver that drains battery power?
<quote>Off the top of my head I believe there’s only 2 hardware issues: WLAN card and power management.</quote>
As a long time Linux user and laptop user, I would like to clarify that. There are actually quite a few driver issues with laptops:
<ul>
<li>Graphics chipsets (notably ATI and Nvidia) requiring binary drivers. Said drivers are often inferior to native Xorg drivers (e.g. ATI’s fglrx lacking AIGLX support, both ATI’s and Nvidia’s drivers not working well with software suspend, etc.)
<li>Wireless chipsets, while somewhat of a problem, are actually less so than other things. Nearly all of them have some kind of project which supports drivers natively or with the use of firmware. For exammple Intel’s ipw2100, ipw2200, and ipw3945 projects, madwifi for Atheros chipsets, and ndiswrapper for many other cards)
<li>Special laptop buttons. There are bits and pieces of support for different buttons on different models, but it is touch-and-go. There are projects like tpb and kmilo for laptop buttons on Thinkpads.
<li>Integrated card readers. Support is iffy or non-existent for many of the integrated SD and multi-card readers. Some popular readers using Texas Instrument controllers have only very recently seen kernel support and the support is still iffy to non-existent with some models.
<li>Full power management support. As you mentioned. This has been a consistent problem and one that is really annoying for laptop users. Laptops are designed to suspend and go into hibernation. They have processors designed to go into different power modes in different situations. Support is there for some of it but its not as good as it should be in terms of just working out of the box. A lot of large distros seem to really ignore laptop users when in fact they should probably be focusing more on them and even creating special branches and packages for them. I don’t understand why distro’s don’t see laptops as core systems that deserve the full, mainstream attention of distribution maintainers.
<li>Fingerprint readers. Basically not supported by any distro. Having used a fingerprint reader on a Windows system and seen the full security benefits of having a single finger swipe authenticate for a whole series of important security functions including encrypted filesystem, logon authentication, etc. etc. I can say that it is a huge thing. Having worked in an industry that was security sensitive I also came to realize how important features like this are for protecting sensitive data.
<li>Hard drive shock protection. This is built in to many laptops and I think there is an open-source project to support it, but I don’t know of any distro where this comes configured and working “out of the box”.
<li>External monitor support. Something nice to have working on a laptop, especially when you want to plug in to a projector and give a presentation. Of course it can be configured in Xorg, but I know of no distro that has it set up out of the box and working.
</ul>
Each one of these issues may seem like something small enough to not worry a lot about on its own. But the laptop user experiences the sum total effect of lack of support all at once and this makes a strong immediate impression.
Advances are being made, but they are very incremental, and there is just the feeling that laptops are not considered really as important to many distros as they should be.
Perhaps not. While I respect that this is your opinion, I would venture to say the opposite.
Most companies that I’ve heard of releasing binary-only drivers (example: NVidia) are doing so because they feel that releasing their driver under open-source licensing would reveal too much of the IP of their product. If they were given a choice between either offering an open-source driver or not releasing their driver at all, they’d most likely say “screw that” and not release anything.
What this would impact more than anything are the 3rd party reverse-engineering open-source driver projects that have nothing to do with the manufacturer, as I’m sure there are many programmers out there who would rather work on their own projects than try to write a driver when they are a 2 minute binary-blob install away from having the functionality they want.
There are rumors that companies like NVIDIA are actually hestiant to release drivers not because they’re afraid of showing their IP, but because they’re afraid of competitors digging through their code looking for IP infringements.
I’d never really thought about it like that, but your probably right.
The way I see the whole “Closed Source” issue is let them make their own drivers, and let them seal it if it makes their Legal Department sleep at night, but at least we have the hardware working and the devs out there can at least poke around and make their own (possibly better) drivers in their own time.
Actually the one I had read was that there were some code for the Intel AGP bus within the nvidia drivers, and that if nVidia had open sourced them, Intel would sue.
That’s what happened with the nvidia-glx project that was created long ago http://utah-glx.sourceforge.net/
If Ubuntu incorporates a polished C-N-R experience, we won’t have to that whine again, “Linux programs are so hard to install”. I mean, one click, that’s as easy as it gets. I guess we could add retinal scanning, and if you stare at a program icon for longer than 3 seconds it “jumps” into you hard drive. I guess that would be easier.
If Ubuntu incorporates a polished C-N-R experience, we won’t have to that whine again, “Linux programs are so hard to install”. I mean, one click, that’s as easy as it gets.
Yeah, you are right, c-n-r is far easier to use than that Synaptic thing they have at the minute.
The developers of Synaptic are a joke, they must think we are all brainboxes.
If I want to install a program, I have to scroll down the list.
Click a box beside the the name of the program, then click Apply, the thing is then installed.
Whats the story with that ?
2 clicks and I am done ?
No way, like you, I want it done in 1 click.
BTW – I was not having a go at you or C-n-R, just having a go at the Windows trolls who always say installing on Linux is too hard with all that source code and the compiling and the debugging and the pain and the hitting and the hurt.
“The developers of Synaptic are a joke, they must think we are all brainboxes.
If I want to install a program, I have to scroll down the list.
Click a box beside the the name of the program, then click Apply, the thing is then installed.”
Whats the story with that ?
2 clicks and I am done ?
Except that it’s not just two clicks. You have to find the package (not application, mind you; package!) you want to install in a mess of nonsensical names. Even than, applications can consist of many packages, which, if you are lucky, depend on each other, but one might be at ‘d’, while the other at ‘l’.
Imagine this under Windows: “What do you want to install today? Microsoft Office, d3dramp.dll, tsdiscon.exe, wincmd.exe or ntdos804.sys?” Can you find which are useful? This is roughly how Synaptic looks like now.
Don’t misunderstand me. I am not bashing it; I still think that software installation is very convenient on Linux. But not too user friendly. If you want to remove a library noone uses anymore, you should be able to do that, of course. But you don’t want to deal with all those excess packages when you want to install blobwars.
I think what Synaptic, Adept, etc. need is a separate “application” view. In there, the user could select the applications he wants to (un)install, and the system takes care of all the rest. CNR looks nice, but (as far as I can see it) it distances the application from the packages, which I do not think is right; if the user ever has to deal with packages explicitly, he won’t know what they are. But we will see how it turns out.
I think what Synaptic, Adept, etc. need is a separate “application” view. In there, the user could select the applications he wants to (un)install, and the system takes care of all the rest.
Gnome-app-install
http://www.flickr.com/photos/jsgotangco/107106644/
Gnome-app-install
Is it a separate application? Yes.
Does it list the packages associated with the applications? No.
Basically, it’s everything I said I don’t like about CNR
I think what Synaptic, Adept, etc. need is a separate “application” view
emmm, they do, it is called “Add/Remove Programs” and it is usually found in the main Applications menu. Well… for Ubuntu based distros.
And it lists them as;
CD Player
Firefox Web Browser
KMPlayer media player
OpenOffice suite.
and so on and so forth.
Now, I could really care less who uses proprietary software and who doesn’t. But what I dislike is people who are relatively new to GNU/Linux and bring their garbage along with them.
But thanks to Mr. Raymond’s nonsense neutered ideology their garbage sticks, and it stinks. This is accomplished by removing any of the philosophy of free-software and relying on alleged technical superiority. Now, anyone with a clue realizes that more often than not, proprietary software will win — at least in the end-user general desktop sense. So there is no reason not to use proprietary software if you subscribe to this.
Which is fine, but please, why not do it somewhere else? And the arrogance of people who just got here. I think I finally know how the poor Redneck feels when a bunch of “liberals” move into his town and drive up property prices.
i miss the days when GNU/Linux was relatively obscure enough that you didn’t have to listen to what everyone and their dog thought, and I miss the days when tripe like the article posted didn’t carry much weight. And I really wish all the “open source” people would just go use BSD, OSX, or Windows.
“””
i miss the days when GNU/Linux was relatively obscure enough that you didn’t have to listen to what everyone and their dog thought, and I miss the days when tripe like the article posted didn’t carry much weight. And I really wish all the “open source” people would just go use BSD, OSX, or Windows.
“””
Here’s an idea for you. Finish up the Hurd and you’ll have your own obscure OS that you can call GNU/OS or whatever you want to call it and which will hardly have anything written about it in the press. Just like you like it. It’s not like Hurd has not been in development for 17 years… or a year longer than Linux.
Then again, since you feel that philosophy is more important than functionality, why don’t you just “go and use” GNU/OS right now instead of sponging off our Open Source Linux kernel.
Of course I’m not really serious about that, and I only say all this to highlight how petty the views expressed in *your* post seem.
-Steve
Edited 2007-04-05 20:09
“New people” are starting to use Linux, because it is more user friendly.
There is an ambivalent message that comes from Linux community. Once they say that people should switch to Linux, and next time they say that people that switched to Linux should share Linux community values. It is a contradiction.
People who use particular product, an operating system for example, have very different values. They choose the product because they find it useful. If Linux community wants faithfull users only, then they should not encourage other people to switch, but limit their invitations to ones that share their values.
In fact, I remeber the days when Linux was used by people with enough IT skills only. There was not much politics then. Either you could install and use it or not. Most of them couldn’t. Nowdays every proud KDE or GNOME user claims to be more faithfull “to the cause” and more fundamentalist then the one next to her/him.
@trenchsol:
“Nowdays every proud KDE or GNOME user claims to be more faithfull “to the cause” and more fundamentalist then the one next to her/him.”
This is true. I remember when the community actually used to be friendly. Hell the community even won the ‘Best Technical Support’ award from infoworld magazine in 98.
Now its about calling people noobs and generally pissing off as many newcomers as possible. When there are no newcomers to harass, these people generally argue about the virtues of the package manager of ‘current distro in favor’ vs rpm/deb etc.
this is a crappy article. sorry to say this, but it’s simply a timewaster IMHO…
this is a crappy article. sorry to say this, but it’s simply a timewaster IMHO…
So was your post. I want my 30 seconds back.
Tomcat. You spend most of your time bashing Linux, GNU software, GPL and GNU/Linux users in general.
Why don’t you post under this story? http://osnews.com/story.php/17640/30-Days-with-Vista
That one ought to fit you
IMHO the linux is a biggest timewaster
Yes I agree, Linux is wasting my time. I get to do work under Linux, I dont get time to mess around when the systems are down !!!!
1) They have a poor understanding of history.
2) They have a poor understanding of social dynamics.
3) They have no appreciation for politics, largely because of (1) and (2).
There are far too many techies that don’t have a good understanding of the history of the computing industry, the political forces that made it look the way it does today, and the social forces that encourage people to write software for free. The author’s three last points basically are perfectly illustrative of these issues.
If you fail to appreciate the politics of the GPL, you fail to understand what the open source movement* and its success. It is the GPL, restrictions and all, that have made open source successful, by providing incentives for developers to create Free Software. People who write open source code do so to a great degree to contribute to a community of developers and users. The GPL protects that community by ensuring it can’t be taken advantage of. In economic terms, the GPL solves the freeloader problem for what is basically a particularly pure public good (being completely non-rival), by turning it into a group good**. This is an advantage that works both ways. Individuals are more likely to contribute, because corporations won’t use the software without contributing as well. Corporations are more likely to contribute, because their competitors won’t just fold their contributions into their proprietary products. That’s why companies spend hundreds of millions of dollars on GPL’ed code, and not on BSD licensed code. The GPL might be distasteful to the real hippy idealists (the ones who talk about perfectly “pure” free code), but its a particularly effective license with some very shrewd logic behind it.
More generally, politics has had an enormous hand in shaping the software industry as it exists. The story of the software industry is not one of technologies, but one of people, companies, marketing, and business plans. If you don’t see how the GPL plays a huge role in these regards, you’re completely missing out on what’s actually going on.
*) Of course, there was open source before the GPL, but an integrated, completely open system with an active, worldwide developer base? No way.
**) Economic theory asserts that in general, a free market system is efficient, in the sense that it will spend resources producing exactly those things that its people value the most. It also notes a few types of goods that are exceptions to this general rule. Public goods, goods which are available to everyone regardless of whether they payed for it (like clean water), are underproduced by society because there is little incentive for people to pay for it since they can benefit from it even if they don’t. This is called the “freeloader problem”. Consider if something like national defense was payed-for by a subscription rather than a mandatory tax. A society might be willing to purchase $100bn of defense to get the level of safety they wanted, but would actually spend far less than that (and get a lot less defense) because of the freeloader problem (you have to defend those who didn’t subscribe). Open software is essentially a public good, and suffers from the same problem. The GPL circumvents this by turning it into a group good (like a country club). It’s still non-rival (you can’t use it up), but by denying use to those that don’t pay, you eliminate the free-loader problem.
Edited 2007-04-05 20:06
…but by denying use to those that don’t pay, you eliminate the free-loader problem.
That’s a fascinating analysis. I recognize and thoroughly appreciate the efficiency and excellence of the free market and separately I thoroughly appreciate free software. But I don’t thoroughly understand why free software works and why the free market embraces it. Comments like yours shed a little light.
You mention appreciate the history, and yet I wonder.
>>> If you fail to appreciate the politics of the GPL, you fail to understand what the open source movement* and its success.
Open Source != GPL. I know many many people would be extremely annoyed at your comments. If you look at the successful projects you will notice that most are non-GPL. Look at the Apache, Mozilla, Perl, Python, Ruby, etc projects and they are non-GPL. Linux is the exception and it has exceptions to allow binary drivers, etc.
>>> Individuals are more likely to contribute, because corporations won’t use the software without contributing as well.
The benefit of the GPL, or the so-called benefit is not what drives people to invest in Open Source. (http://www.onlamp.com/pub/a/onlamp/2005/06/30/esr_interview.html)
>>> That’s why companies spend hundreds of millions of dollars on GPL’ed code, and not on BSD licensed code.
BS! Look at where the monies are flowing and it is in projects like Eclipse, Apache, Python, Mozilla, Perl, etc, etc. And these code bases are not GPL’s, but mostly BSD type. Again Linux is the exception. Look at the Open Office suite that is explicitly GPL and it is languishing. It is not a direct competitor to Microsoft Office, nor will it ever be. On the other hand Mozilla with its easier licensing terms is a competitor. Recently I read in a German magazine that Mozilla has grabbed about 30% market share.
>>>Economic theory asserts that in general, a free market system is efficient, in the sense that it will spend resources producing exactly those things that its people value the most.
I disagree with this because people have come to understand that there are behavioral aspects to a free market system. People are herding animals and will do things even if it does not make logical sense. The idea that markets are efficient is a purely neo-classical economic idea that asserts everything in economics can be modeled using natural science. This is not true.
How can you argue that the GPL circumvents the free-loader problem? There is a clause that GPL software can charge a distribution fee, but all it takes is a single person who is willing to distribute said software for free and the system breaks down. Open Source kills the traditional software business model, no ands, ifs or buts. You can argue whether it is good or bad, but there is no arguing that charging for software is not possible.
You might want to argue on how Redhat makes money. Well, Redhat in its decade of business makes peanuts. Redhat makes 375 million per year. Compare that to other businesses that sell binary only applications, or companies that use open source (eg Google, Yahoo) and you will see the billions.
I like Open Source because it puts me in control of my software.
I think you’re exaggerating a bit here. First of all, there are other GPLed projects that have been quite successful, among them the Gnome and KDE desktops, and many, many KDE apps. Do these projects make a lot of money? Probably not, but in the FOSS world success is not measured simply by financial results. Also, didn’t Sun recently state that they would release the next version of OpenSolaris under the GPL?
As far as OpenOffice goes, I completely disagree with you that it is “languishing”…sure, it’s still got quite a lot of ground to cover in order to catch up to MS Office’s quasi-monopoly, but so does *every* other office suite out there.
You seem to be making a connection between Firefox’s success and its “easier” licensing, but I’d be curious to hear you explain how the two are related. Correlation does not imply causation, and yet your statement is based on the fallacious assumption that it does. The truth is that Firefox could very well have had the same amount of success if it had been released under the GPL.
You cannot make such a statement without a more complete data set. Right now it seems as if you’ve only picked examples that support your theory…so unless you can provide some empirical evidence to support your assertion, I’ll tend to go with Rayiner on this one.
Rayiner said:
christianhgross said:
archiesteel said:
I don’t see a lot of empirical evidence from Rayiner either. I’d say archiesteel is guilty of the same thing he’s accusing christianhgross of (or at least guilty of playing favorites). –shrugs– I’d say that neither rayiner or christianhgross has presented a lot of empirical evidence and wouldn’t accuse one of having less than the other.
I do agree with archie that OOo isn’t what I’d call “languishing” at all. It’s a huge undertaking, so I expect that it’s not going to be as fast-moving as some smaller projects. Maybe some have unrealistic expectations, but I’m happy with it and applaud the devs.
Comparing Linux to BSD-licensed projects like apache is a stretch. Let’s compare Linux to a BSD-licensed operating system.
I don’t dispute this. The reason I wrote that I tended to go with Rayiner on this one is that he’s been on this web site for quite a long time, and through writing many insightful and informative posts hehas demonstrated a good amount of knowledge on such matters. Therefore, I tend to trust him more on this. If you want to see this as being “guilty” of anything, then be my guest. Frankly, I really don’t care.
Arch, all I’m saying is if you’re going to lecture christianhgross about presenting “empirical evidence”, you could stand to take your own advice.
Seriously, what’s the point of empirical evidence if you’re just going to take Rayiner’s word for it? How does misinformation start? 
I’m not “lecturing him”…he made a statement which I find dubious, and I’m asking for more details to back him up.
As I said, I know Rayiner from having read his posts here for years, and I know he doesn’t make unsubstantiated claims. His claim also has a basis of truth, i.e. most projects on Sourceforge are GPLed. I don’t know christiangross, and I’d like to have some info about why he makes his claim, that’s all.
While you are correct that open source != GPL (which I noted in my post), the open source movement is most definitely driven by the GPL. David Wheeler’s analysis of the RedHat distribution several years ago showed that fully 50% of the distribution was under the GPL. The second largest license was the MIT license with 8%, followed by the LGPL with 8%, followed by the MPL with 7%. The GPL numbers since then have probably grown, not shrunk. Since that time, GNOME and KDE, which are both LGPL’ed, have gotten a lot bigger. Mozilla has been triple-licensed under the GPL/LGPL. OpenOffice has been LGPL’ed. Qt has been GPL’ed. Java is being GPL’ed.
There are certainly high-profile non-GPL’ed projects out there. However, if you look at sheer volume of code, sheer volume of developers, and the sheer amount of money being spent, the GPL far outstrips any other single license in popularity, and perhaps a good number of licenses put together. The effort spent on Linux alone probably dwarfs the effort spent on all the projects you mentioned combined. Then factor in the effort spent on GNOME, KDE, GCC, etc, and there’s no contest.
If you look at copy-left licenses more broadly, the prevalence is even more striking. The Mozilla MPL is a copy-left license, like the GPL. So is the Solaris CDDL license and the Eclipse EPL license, though they are weaker in their terms than the GPL.
How can you argue that the GPL circumvents the free-loader problem?
The GPL doesn’t eliminate the free-loader problem, because people can use the product without contributing to it, but it does a far better job of getting around it than non-copyleft licenses do. The GPL eliminates the most onerous freeloaders: those that benefit directly from a program by using parts of its code but not contributing them back.
I like Open Source because it puts me in control of my software.
Developers like open source licenses that let them keep control of their software. That’s why so many developers use the GPL.
Edited 2007-04-06 02:56
I haven’t read the article but saying there’s something wrong with Linux is being a troll because is perfect and it’s better than *any* other OS. There’s no way to enhance Linux and there will never be. It’s just perfect.
I assume Mr Norris (who by the way, I hear CAN calulate the square root of -1), those probably aren’t horns on your head. You probably were just bitten by mosquitos on both sides of you head and large bumps were produced. Am I right?
The square root of -1 is “i”, the imaginary number.
That said, I do think you’re right. The OP was himself trolling by making a highly sarcastic remark.
don’t mess with him.
not only can Chuck Norris divide by zero, his calendar goes straight from March 31st to April 2nd…
no one fools Chuck Norris.
I can divide by zero too. I just can’t explain the result in any sensible way because it is outside the range of definable numbers (quite a trick in itself – took me a whole night and a very strong coffee to figure it out).
I could divide by zero if I wanted to, but it would destroy the universe and that doesn’t suit me at the moment.
Do not wind him up, he will roundhouse kick you into oblivion
In parts, he doesn’t get it.
Unlike more flexible open source licenses that would have given the Linux kernel the same protection and freedom to grow, the GPL has been a thorn in my side for years.
The only reason why Linux became popular is because Linus chose a license that held it all together. There are no options to stick binary blobs on to the kernel, and therefore more work goes into it rather than on closed extensions that drag down the overall quality. It also encourages people to contribute, because everyone is doing so on a level playing field.
The people who just cannot understand this never cease to amaze me.
As with the GPL licensing politics, Linux users are bred early on to believe that proprietary software is a dangerous thing.
Not really. It’s just that proprietary software is very difficult to install on a Linux distribution, so people don’t bother and stick to what came with it and what’s in the repositories – open source stuff.
The only reason why Linux became popular is because Linus chose a license that held it all together.
No, the reason why Linux became popular was that it offered significant value in the server room in displacing systems that cost thousands more. Its popularity isn’t due to the license.
Not really. It’s just that proprietary software is very difficult to install on a Linux distribution, so people don’t bother and stick to what came with it and what’s in the repositories – open source stuff.
Nonsense. Proprietary software is no more difficult to install on a Linux distro than open software. For example, my commercial editor (SlickEdit) installs and runs just fine on Linux, thank you.
No, the reason why Linux became popular was that it offered significant value in the server room in displacing systems that cost thousands more. Its popularity isn’t due to the license.
How did Linux reach the point where it could offer that kind of value in the server room? It’s because the GPL’s protections encouraged a level of participation in the system’s development that no competing open system could match, even those that started with vastly more mindshare (BSD).
There was a study that showed that the majority of Linux’s contributors were now paid developers. This state of affairs would very like not have come to pass if Linux was BSD licensed. There is no way a company like IBM would spend millions of dollars on Linux development if HP could just fold the improvements into HP-UX without paying for them.
How did Linux reach the point where it could offer that kind of value in the server room? It’s because the GPL’s protections encouraged a level of participation in the system’s development that no competing open system could match, even those that started with vastly more mindshare (BSD).
Nonsense. Linux reached the point it reached because IBM and other companies dumped a ton of money into the effort, in order to co-opt Sun.
…piss poor examples. Not to mention being somewhat misinformed.
I vote that from now on, we enforce a no publishing your musings about Linux for public consumption ban unless:
1) You’ve using Linux as your primary OS for no less than 5 years.
2) You’ve taken at least 20 credit hours of classes relating to Journalism.
How about actually addressing his points rather than complaining about “piss poor examples”? You’re wasting server space with these kinds of posts.
Are you the author by any chance?
Since you are a most ardent Linux-hater you might want to post under a different article
… this crusade of yours could be quite unhealthy for you (mentally that is) .
Give in to your hate. It only makes me stronger… BWAHAHAHAHAHAHAHAHA!
Most people, like in these comments, don’t understand opensource at all, the author of madpenguin too.
The Top 5 Things how to make it to OSAlert or digg would be appropriate for this topic.
I’m not very surprised to see this topic at OSAlert, Eugenia has written similar topics at her very own blog. Eugenia, why not post these at OSAlert, why such a lame approach to support your very own thinking of opensource? Is this the FUD-year? The FUD of opensource, telling people x points why this and that sucks? Who lame. It’s very easy with opensource, be a part of it or leave. There isn’t something to demand, you can give something – it isn’t just poking around in someones elses code.
But I’m sure we’ll see tons of BE*stuff and tirades toward opensource/Linux etc. – it’s the way to go nowadays. And exactly thereforce opensource sucks, because there are people who don’t understand it at all!
1. sound architecture
only one OSS channel + ALSA, what if You have 2 OSS apps that ALSA OSS emulation does not helps? On FreeBSD I have as many as I want multiple virtual channel with kernel mixing, even for old SB16 with One Hardware channel.
2. mess in kernel
to get ATA and DMA support on Linux [for example for nForce2] You have to add Generic ATA + ATA Support + nForce2 ATA Support + HardDisk ATA + HardDisk DMA + DMA Support [+ maybe some more] and for some chipsets Generic ATA BROKES DMA support [for example Intel 945]. FreeBSD: device ata and everithing is done. KISS.
3. /proc
echo “asd” > /proc/some/hidden/magical/file.
4. rpms
anybody is still using it?
5. linux partitioning
for Linux You have to use standart partitions, if You want to have many separate filesystems You have to use extender partition. In FreeBSD You can use SINGLE Primary partition and have 7 partitions INSIDE that SINGLE promary partition [SWAP + 6 FS].
ALSA also does software mixing very well. I don’t see your point in complaining about crufty OSS and crufty old apps that use it.
In recent kernels you have libata and it works perfectly with most mainstream ata controllers.
[*] ATA Device support + [*] Your controller. Not a whole lot more.
What?
If you don’t like RPM, use Debian- or Gentoo-based distro. That’s the very point in having multiple distros – everyone gets what he likes most.
LVM ( http://en.wikipedia.org/wiki/Logical_Volume_Manager_%28Linux~*~… ) has been around for a very long time.
To conclude, I don’t know what year you are in, but it’s 2007 and most things you are complaining about are no longer actual/relevant.
Edited 2007-04-06 01:32
No. I believe this line of thinking stems from a lack of understanding of the free software purist position. We believe the proprietary software is dangerous because it is a stumbling block to society as a whole.
When the user chooses to be subject to the terms of proprietary software, the user chooses to give up their right to control their own computer. Since the vendor is controlling the software and not the user, the user needs the vendor’s permission to control the software; the user is no longer free to control their computer. Moving to a functionally superior program is still more of the same thing; the user remains subject to one vendor.
When the user chooses to be subject to the terms of proprietary software, the user chooses to give up their right to share the software to friends and neighbours. RMS considers this to be a form of censorship as society does not have the liberty to share information in the form of software. Society loses because a vendor desires a monopoly on the movement and sharing of a piece of knowledge.
We believe the irresponsible ones are the vendors that subjugate users to themselves. Software is a collection of instructions that work together to achieve a purpose. If the software’s functioning is not correct for a particular purpose, the user should have the right to get it changed so it works better for themselves, and yet, vendors feel the need to maintain a monopoly on the knowledge by not allowing users the freedom to improve the software’s function.
X is what is wrong.
X is dead slow. Even remote it is slower and harder to set up than Remote Desktop.
X is ancient.
X is a big, huge mess.
X lacks mouse support. Sure, I can move the cursor around but to use my mouse properly I have to search forums for some cryptic lines to manually put in xorg.conf. Still, only some applications (Firefox) works with my back and forward buttons. And this has happened with every distribution and is still happening with Ubuntu 7.04.
X lacks real configuration tools. I install a dist and it always sets the resolution to the highest possible (insanely high that is) and it refreshes at a not so comfortable 60 Hz. To change this I have to run a command line configuration program that writes a new xorg.conf. It asks ME about my video card and my monitor instead of just detecting what is plugged in and ask the monitor about it’s capabilities. If I later want to use a higher default resolution I have to edit xorg.conf again. It is as if Windows 95 never happened.
Well, X is ancient and so on, but your experiences are weird. I remember the times, when turning whole X up and running was a week challenge a using first versions of KDE on top of it was was not a piece of cake.
Today, most distribution lets you choose resolution and frequency during setup/install and in most cases, they work with it after. There are distros with less support, but I thought that Ubuntu is capable of such thing for years, I used it just once for couple of hours.
Regullary I use Debian, I have a Matrox P-series card which causes a bit more pain to starting whole graphics up. Yes, I am experienced user using distro for people like me and start whole X from fresh install in hour or two, 3D sucks with Matrox in many cases, but that’s deffinitely not X’s fault.
I also install linux on many computers around me to my friends or customers, I use Mandriva and I must say, that without 3D, there is no problem at all with ATi or nVidia if you set it right, many displays get discovered succesfully, so the resolutin/frequency boundary is easily managable. Intel graphics cards rocks in linux support, but they aren’t easy to get
And one more thing: Windows is system for money, you pay for it from time to time and it somehow works, not for me, but for many people. Yes, detection of mices was the biggest pro of windows for very long period of time – something I would not sacrifgice my nerves for.
Linux (and other open souce) is for knowledge, once you get the point, there is no need to spend long hours learning anything new, just some minutes from time to time. The knowledge is portable, you can use it not only on your computer, but also any other. The unix knowledge is universal and that is the biggest pro of unix-like systems. I was able to move part of Ultrix to another drive on running system just after about 20 minutes of getting known how DEC treated drives 15 years ago. Try to use your XP knowledge to do something similar to NT4 – move Program Files of better directory to drive D: – and remember, that’s just 5 years.
made a doublepost, sorry…
Edited 2007-04-07 13:34
Your article has been the most poor reasoned mish-mash of unrelated matters since mash potatos with gravy…
I’m always on this website reading news and such but I always couldn’t be bothered registering to post comments, but your article is so bad it made me sign up just so I can express how much i think you suck… Your parents should try again…
That’s why the server (Mad Penguin) runs on FreeBSD.
The title of the article describes accurately what it is about – the top five things that the writer hates about Linux. It does exactly what it says on the tin! And I agree with those points, you’ve pretty much covered my top five too.
A constructive reply should be “no here’s my top five” or “this is why you are wrong” – not “you suck” and with no detailed explanation why…
Maybe it’s impolite, but may I ask, did we have this discussion some time earlier?
The article mentions some points that have been proved to be not a problem caused by Linux or to be in Linux’s responsibility.
Let me first say that I’ll use the term “Linux” for the family of OSes based on the Linux kernel (distributions), just to make clear I’m not talking about the OS’s kernel.
Let me then refer to the article, just to repeat:
1. Lack of Driver Support for Select Hardware Products. […] But as Linux adoption continues to grow, this will have to change as people like me will top spending money on hardware that will not work as we need for it to.
Hardware support is up to the hardware manufacturers and vendors. As long as they continue using established standards (e. g. direct access for mass storage media in mobile phones and digital cameras -da- vs. strange command sets -ugen-), their hardware is supported out of the box. The Linux specifications are free, they are free to use them. If they “invent” something new (in most cases, they’re making easy things complicated), they are responsible to offer either drivers for Linux (which they develop theirselves) or to offer specifications so Linux developers can implement the drivers.
2. Half-Done, Poorly Supported Open Source Projects. […] If you are going to start the project, then do everything you can to do it right if you are maintaining it.
This is a point a can agree to. While in BSD world, nearly everything part of the OS has an own manpage, in Lunix applications don’t have. To extend this, if you install KDE on BSD, you cannot “man ksomething”, because the are no manpages, you need to look up something in /usr/local/share/doc or via WWW. Documentation is essential, especially for developers.
But to release anything as a “tarball only” type of thing and then extend support as far as a ReadMe file is not acceptable and quickly becomes an ammunition round for Windows Developers.
This depends on the complexity of the application and for whom it is designed to use. While something like OpenOffice comes with an installer and online help, pdfnup comes with just a manpage and some docs, which is enough because the people who deicde to use it are smart enough to know how to access these ressources.
3. The Politics of the GPL License.
I would, under no circumstandes, deny the right of existance of the GPL. Allthough I feel more comforable with licenses like the BSDL, GPL is fine for most purposes. As you can see from SuSE / Novell, you can even make money with the GPL.
4. Hatred Towards Proprietary Applications. […] Linux users are bred early on to believe that proprietary software is a dangerous thing.
There are niche markets where only proprietary solutions exist, that’s true. And there is proprietary software you can trust, that’s true, too. But I do not feel good with software running on my system in essential or security related parts where I cannot control what it does…
Yes, both [OGG or Theora] are superior formats in almost every way, as is often the case with many open source solutions, but the fact remains that until we can pressure fair adoption, there is no reason to punish the rest of us who are simply looking for a cost effective alternative to Windows.
It’s because the worst solution wins. In Germany we say: “Das Schlechteste setzt sich durch.” (The worst one prevails), and that’s true nearly everything. You can even imply: If the worst one prevails, then the solution that’s used outside of the mainstream, is definitely the better one. The more people use something, the worse it must be…
5. Animosity Towards the Inclusion of Ubuntu Feisty’s Hardware Detection Utility and Linspire’s CNR. I have read a number of posts from people who feel strongly that Feisty is walking on the dark side by “encouraging” the use of proprietary video drivers and/or CODECs. Give me a break! What the developers are finally doing is providing functionality that beginners have been screaming for since Dapper.
I think this is true and even neccessary. If Linux wants to get more usage share (and oh joy market share), it first needs to to everything “Windows” does, meaning, playing strange video and audio formats, reading obscure files and supporting obsoleted hardware. Because if it does not, there’s a quick complaining: “But my ‘XP’ does it!”, and the Linux CDs enter the waste bin.
Maybe I did a mistake understanding the article, so I would see the critics encapsulated in the article do not “offend” Linux, they do criticise hardware vendors etc. in fact. In my opinion, the author makes real interesting points, but compares apples with olives sometimes.
Blogged this earlier…
In a nutshell, Hartley just wants the best of both worlds — taking advantage of the free (as in beer: “simply looking for a cost effective alternative to Windows”) while remaining prisoner to hardware companies that refuse to support more than one or two operating systems. I make my hardware choices on the basis of my operating system. If it won’t work in Linux, I’m not buying it. Neither should he. That’s the only way companies will ever open their hardware to other OSes.
This is where I have my strongest disagreement with Hartley’s list, especially as it relates to open versus closed source: it’s mostly a hardware problem, not a software problem. Nobody’s stopping him from using proprietary code (apps or drivers); some distros (Mint) include both. Hardware manufacturers and retailers make money from the sale of their work. Linux developers generally don’t — at least not directly (which goes to his misguided point about projects that don’t get “completed” or developed). I’m not going to blame the guys whose work I’m using for free (Linux developers) for the chief mistake made by the guys who do make money (from hardware sales) of having hitched their wagons to one or two operating systems.
here is no reason to punish the rest of us who are simply looking for a cost effective alternative to Windows.
Maybe linux will get better reviews in these types of articles when people see linux as more than just a “free windows.” I use linux for many reasons, the price is just an added bonus
1. GPL nonsense.
2. GPL nazis.
3. GPL restrictions.
4. GPL lovers.
5. And finally- GPL.
GPL is worst license ever created- it destroyed collaboration between open source *BSD and GNU/Linux communities.
No, GPL is just another (and somehow incompatible with others) frame of cooperation. When someone is doing some work (write the code), it’s only his/her right to choose how he/she wants his work used – so he/she may choose BSD, GPL, closed source, use any other existing license or even create new one.
Some people choose to make BSD, they made their license accomodating their needs. After some years, there came BSD v2, a bit less restrictive, alowing inclusion of BSD code into GPL source base in the firs place.
Then there came people willing to make software, which is as unusable for direct selling as possible and there came GPL – in a simple way: it enables the cooperatin between developers gurantiing that nobody will charge anyone directly for their work. You can sell support, you can sell your installer or any other particular piece of sotware, you may sell the package with manuals, but you can’t sell just porgram writen by somebody else. Look at BSD/OS or SSH in many proprietary system, see ip stack in early versions of windows – the BSD licensed code is beeing sold for hard money with minimal effort around. Well BSD/OS is down dor some years, but it’s not so long time from OpenBSD’s financial problems, when it came out that openssh is directly used in solaris (under sun’s name and maybe some changed lines), but OpenBSD project, which made openssh and maintains it got nearly nothing from Sun. If you make BSD licesed code, you have to cope with that many people (and/or companies) will earn their money using your product, they may make some usefull improvements into your code, but there is no force that will show them to you.
Sorry for simplicity, those licenses are a bit more complicated (not even talking about GPL3).
It’s about a choice.
Edited 2007-04-07 13:51