Why Linux Sucks & Why Linux Does Not Suck (Lunduke) [Archive]

bigbrovar

May 1st, 2009, 08:52 AM

Its FUD, why this is being proposed by the so called "linuix foundation"
I dont know...
Perhaps a microsoft buyout...

see the problem with linux? when we the users cant look at our selves in the face and tell ourselves what needs to be improved on with the linux desktops.. its not every linux problem that is hardware or proprietary software based. some are just obsession with bleeding edge and lack of standard way of doing things. i once talked about some issues i was having in the testimonial section and i was flamed down despite being very objective in my post. some people just cant stand objective criticism of Linux to them its tantamount to blasphemy. i completely agree with most of what was said in the videos and this are problems that lots of users face every day on the Linux desktop. non standardization of the audio framework,obsession with bleeding edge technologies that often times break compatibility and causes regressions, shipping buggy softwares with what is suppose to be a final release. scraping tested and trusted for bleeding edge technologies that are not ready for primetime and are less buggy than what they are replacing causing more problems than is solved. yeah i know the flames will start coming now. but whether you take it or not these things are the realities of the linux desktop today.

racerraul

May 1st, 2009, 08:53 AM

Great video... I was beginning to think I was the only one that saw things in that light.

I few weeks ago I started a thread about the fact that not enough people donate and gave an example to the potential problems some distributions face due to lack of funds. I mentioned how I felt if we all donated a single dollar how it would stimulate what we have access to.

Sure enough, some jackass chimed in about being worried about making someone rich richer and how it was better to just wait for some rich person to throw money at the projects and pray it was the one we were interested in...

Are you freaking kidding me?

This video explained why we don't have access to some great industry standard applications. Well there you have it... if you want Linux to be taken seriously as a viable business model, we need to put money into it, because money is what drives business.

The whole free software definition is misinterpreted. We are free to run the software as we want, we are free to modify the code to suit our purposes, we are free to assist each other and free to redistribute.

But that does little to compensate for the cost to make everything we use accessible to us... just the internet connection isn't free for us to go out there and get it.

I agree with everything said in that video... There are some definite positive opportunities in funding standardizations. Not everyone has to adopt them, but they would be beneficial for closed source companies to experiment with that business model. I think that Linux standardized distro even if it wasn't free but say had a price tag of $60 would be a bargain when you consider what the alternatives in Windows and Mac cost, specially if it offered access to industry standard applications for LESS! Not for a minute do I think that such a scenario would be detrimental to linux. I am willing to bet that even the smoochers would find it easy to borrow that little bit of money from a friend or family (and never pay them back) to have such access.

So ask yourself, do you have a buck to spare? Then donate it. It is the least you can do. Do with out a run to the candy bar, junk food machine, hell that crap costs more than donating a buck and its BAD FOR YOU!!!! :lolflag:

How is it that he put it?
Developers need to eat food too...

karellen

May 1st, 2009, 08:56 AM

see the problem with linux? when we the users cant look at our selves in the face and tell ourselves what needs to be improved on with the linux desktops.. its not every linux problem that is hardware or proprietary software based. some are just obsession with bleeding edge and lack of standard way of doing things. i once talked about some issues i was having in the testimonial section and i was flamed down despite being very objective in my post. some people just cant stand objective criticism of Linux to them its tantamount to blasphemy. i completely agree with most of what was said in the videos and this are problems that lots of users face every day on the Linux desktop. non standardization of the audio framework,obsession with bleeding edge technologies that often times break compatibility and causes regressions, shipping buggy softwares with what is suppose to be a final release. scraping tested and trusted for bleeding edge technologies that are not ready for primetime and are less buggy than what they are replacing causing more problems than is solved. yeah i know the flames will start coming now. but whether you take it or not these things are the realities of the linux desktop today.

totally agree with you. but for some people, Linux is not just an OS that, as any software, needs - and can be - improved. it's a credo, something to be defended from trolls/malevolent critics, almost as a religion. like all other religions, it's perfect and it holds the ultimate truth

racerraul

May 1st, 2009, 08:58 AM

I don't like everything this guy is saying. Standardization of packages would make it easier for linux as a whole but it would take away freedom.

no it wouldn't... things can remain as they are. Offering a standardization model can coexist. And if that happens to advance linux development at a faster rate it would be due to the people having been free to choose in using a standard model.

SunnyRabbiera

May 1st, 2009, 09:10 AM

karellen

May 1st, 2009, 09:25 AM

mikewhatever

May 1st, 2009, 09:44 AM

not it's not. Windows has only .exe's, but this doesn't mean that only one type of application per task is available. take browsers for example. or text editors, or media players. I believe you're mixing software standardization with platform standardization. from my point of view - and others might disagree as long as they want - having n! package managers, .debs, .rpms, tar.gzs and God knows what more, k! ways of packaging them all together (distros) is detrimental for the growth of Linux(as in market share). of course, for hobbyists, and I'm one of them (I've been using Linux since 2003, the way things are right now - fragmented - it's ok. but that's not the point. an OS for the masses (or should I say "human beings") != an OS for tech savvy persons
just my $0.02

What about .cab and .msi in Windows?

....i once talked about some issues i was having in the testimonial section and i was flamed down despite being very objective in my post. some people just cant stand objective criticism of Linux to them its tantamount to blasphemy.....

You were not flamed, in fact, many agreed. Don't you see a problem with self-proclaimed objectiveness? Any troll can say that, usually does, and every single word trolls type is objective criticism.

bigbrovar

May 1st, 2009, 09:54 AM

Standardization is just saying another way of locking people down, what do we have to do just use one UI, one toolkit, one browser, one audio server, one package manager, one type of installer?
I think linux needs to stay diverse, we dont need "standardization"
Want one UI, one toolkit, one browser, one audio server, one package manager, one type of installer?
Use Microsoft...

let me start by saying i love your repy this time. now we are talking issues :). Although i don't quite agree with you that standardization brings lock downs or limits our freedom. take a look at the *nix file-system structure which is standardized so we have /boot /etc /usr/ /opt/ etc you see the same structure on just about every distro.. also look at bash and the gnu core utils.. that is standard.. i could be on arch or redhat and run the same basic Linux commands and get the same result. how does this lock you down or limits freedom? what standards just means is that this is the way sound is handled for example .. this is how web-cam is handled. we all work to make this better and every app is built to work with that framework.. the very essences of free software is having standards for example when its comes to the GUI display we have the X server everything from gnome to kde sits on the xserver. and although the x server is not perfect but at least we know how to attack the problem and fix it. not so for sound.

packaging might not be that easy. not because we cant have standard packages for Linux for lack of a technical know how. but for mostly political reasons and the fact that most distros like to work in isolation without looking at what is doing on in the other camp.. i feel each distro could still maintain their own packages, debian/ubuntu keeps their debs for e.g but at the same time there is a standard packaging procedure which all packages most adhere to.. and also a unified packaging format which (might need and installer) will work on all linux platform. and the standard for building this format is then published to reduce breakages. if you go to firefox site or openoffice or pidgin you would find out that depending on your distro .. installing the windows or mac version is just easier. the linux format comes in a .tar.bz format. which in most cases (except pidgin) can be installed by just dumping the file in /opt and running it from there. why not just have a standard package format which does this. this doesnt mean that we still wont have debs for the same program just makes it easier to make binaries for 3rd party applications so that people can just install them no matter the distro. the source codes will still be there for people you want to compile their software. which as u know there is a standard to compiling which works across all distros so why not just have same for installing binaries ..

bigbrovar

May 1st, 2009, 10:00 AM

You were not flamed, in fact, many agreed. Don't you see a problem with self-proclaimed objectiveness? Any troll can say that, usually does, and every single word trolls type is objective criticism.
well you are right about that. but what i meant was that for every issue i raised i attached a bug report that showed that it was in fact a software problem rather than a hardware problem as most people were saying i was having. and to proof the regression problem that is often caused by obsession with bleeding edge apps and shiping buggy applications and libraries with the final release of a distro. even though i agree it doesnt take away your point about self claimed objectivity which can be subjective :)

JohnFH

May 1st, 2009, 11:04 AM

not it's not. Windows has only .exe's, but this doesn't mean that only one type of application per task is available. take browsers for example. or text editors, or media players. I believe you're mixing software standardization with platform standardization. from my point of view - and others might disagree as long as they want - having n! package managers, .debs, .rpms, tar.gzs and God knows what more, k! ways of packaging them all together (distros) is detrimental for the growth of Linux(as in market share). of course, for hobbyists, and I'm one of them (I've been using Linux since 2003, the way things are right now - fragmented - it's ok. but that's not the point. an OS for the masses (or should I say "human beings") != an OS for tech savvy persons
just my $0.02

Totally agree!

What about .cab and .msi in Windows?
and what's your point? .cab and .msi are not different ways of doing the same thing, they have different purposes.

tadcan

May 1st, 2009, 11:04 AM

Its FUD, why this is being proposed by the so called "linuix foundation"
I dont know...
Perhaps a microsoft buyout...

Linux Foundation gets videos from other sites and puts them in one place. That video was made from the guys at http://www.jupiterbroadcasting.com/

oomingmak

May 1st, 2009, 12:15 PM

Everything he said makes perfect sense, but you will have about 1000 people claiming blasphemy.
That is so true.

Some people would rather attack those who point out Linux's shortcomings than actually accept the criticism and see it as an opportunity to improve matters.

t0p

May 1st, 2009, 12:42 PM

Of course Linux apps need income to develop. I've sent money to a couple of devs. Not much, but I haven't got much. But I try...

Thing is, standardization/centralization/lets-make-Linux-the-same-as-the-rest-ization is not what it's all about. Linux isn't a single OS created by a single corporation. It's a whole bunch of OSes, interrelated and partially interchangeable but nevertheless a heterogeneous group of apps and utilities. That's what makes it so great to me.

Look at Windows - there are a bunch of different devs making different apps etc, but they all have to bow to the dictates of Microsoft and their products are exactly that... products. Windows is for sale, even if many apps are free-of-charge. Maybe that's what some of you want for Linux, but I certainly don't share that vision.

Maybe I'm a selfish geek who would prefer Linux to remain "my" little "secret". Well, I'm not... but I certainly don't want Linux to destroy itself by trying to become Windows.

JoshuaRL

May 1st, 2009, 12:56 PM

i feel each distro could still maintain their own packages, debian/ubuntu keeps their debs for e.g but at the same time there is a standard packaging procedure which all packages most adhere to..

Well, I know that you mention it later, but that's what source code is about. If an application is important and tested/trusted enough, then it gets packaged my maintainers. That way you get the pluses of repos and proper package management. Otherwise, you kinda take your own system into your hands. For a developer, release in source code and include all the bits you need for most distros. Leave an open line of communication and users will inform you of what is needed/broken. This works pretty well.

why not just have a standard package format which does this. this doesnt mean that we still wont have debs for the same program just makes it easier to make binaries for 3rd party applications so that people can just install them no matter the distro. the source codes will still be there for people you want to compile their software. which as u know there is a standard to compiling which works across all distros so why not just have same for installing binaries ..

Because .debs and .rpms will always be second best. Your applications aren't updated unless you do so manually. If there's no version in the repos, and you don't feel like packaging and submitting your favorite application to Ubuntu (or whatever distro), then you do what's second best. Or third best, whatever. Having a .deb will never equate getting the same application from the repos. That's why GDebi warns about doing that exact thing.

If you're suggesting a separate repo and packaging system for ALL Linux distros, that's not very likely, nor very worthwhile IMHO. It's like this: Debian-based distros are designed from the ground up for dpkg and APT. The security model is based on it; all of the updaters and other applications are based around it too. Same goes for Red Hat-based distros. I've tried running YUM on an Ubuntu system. It didn't work well at all. If you include a new system, there's nothing to say a pre-compiled and packaged application will work correctly on all systems. But that's the beauty of source code. It will. There's a set way of doing it for all distributions, and if it's needed it works great.

whoop

May 1st, 2009, 02:24 PM

I agree with allot of his arguments. I do however NOT agree with most of his solutions. I don't even really agree with his main goal.

I don't feel for this whole linux desktop must succeed or the year of the linux desktop stuff. It's doing just fine as it is for me.

I for one am not looking forward to commercial linux, or even mainstream commercial software for linux.

If we all made this happen what stopping "them" (conspiracy theory's here ;) ) from adding DRM, registration codes, phone home functions, closed source bugs, online stores, spyware, malware aaaarrgghh.

I know you can say you have a choice, but for once I like it that I don't (for the moment).

Bölvaður

May 1st, 2009, 02:43 PM

I disagree with some things like having 1 package management system.
It wouldn't be possible even, as most distros would just ignore the standard. So in the end we'd end up with exactly what we already have except we have 1 package system that dominates (.deb is that today as for the popularity of those distros) and the rest uses what ever they want.
The non standardized distros will probably adopt the best package manager more often than not, so a package manager like pacman would rule. So basically we'd have a standard (stable) and "bleeding edge"(unstable).

That would mean like he says in the video, we'd need to stick with the standardized filesystem hierarchy we got already but need to define it better. I have sometimes scratched my head when Im trying to find some things on different distros. For a standardized package protocol it is vital.

Also I dislike my config files being stored directly under ~/ but me being annoyed and asking the config files to file up and jump into ~/.config/ is a matter for another tread.

And sound is a problem. Pulse is gaining momentum to become a standard already so this may not be a big issue if things continue as they are going at the moment.

There already are standard each distro must comply to. Adding a standard package management system for few distros isn't going to kill off linux. There will be the same diversity and will stop developers saying that they dont know which distro to design for... give them less excuses not to port their software to linux.

The big issue is DEVELOPERS NEED FOOD. And that when it comes to certain type of software it needs a core team working on it constantly. That's why we should donate every year or every other year like $20 to our top 3 beloved projects.

Mr. Picklesworth

May 1st, 2009, 02:44 PM

I really don't think we need to standardize packaging formats so much as packaging tools. We need something like cmake for packaging and deploying Linux applicaitons. Give it files, versions, a changelog and dependencies; it sorts out how to build a package for X distro, popping up a nice interface when it needs to and remembering settings for the next time.

Then the user can use the package manager which has been deemed appropriate for his distro.

Sure, standardizing the format itself would be nice, but that's a pipe dream so far whereas making the tools easy could happen immediately.

geoken

May 1st, 2009, 03:02 PM

How does standardizing a package manager/format restrict freedom?

For example, if an app developer now only has to spend time making one package how has your freedom been impacted? Worst case scenario you extract the package manually and install from source.

At it's core, a package is just a compressed archive of source files with some type of manifest that includes instructions which can be used by the package manager. So even if your OS didn't want to support that package format you can still use the files.

SunnyRabbiera

May 1st, 2009, 03:31 PM

How does standardizing a package manager/format restrict freedom?

Well lets give an example of why not having a one size fits all approach is a good thing:
Say all the package managers were merged into one, called the Linux install manager and it uses .lim packages to work.
In order to limit choice and maintain a standard there might be a need of a non GPL licence, after all its GPL that allows there to be multiple linux distros.
You are not allowed to change what kind of packages you get and dont get, you are not allowed to maintain the OS at your own pace, instead you are forced to upgrade packages, no choice in whsat is upgraded and when.
Then this .lim system needs 10 different dependencies for not just a few apps but all of them, there is no control over how many packages are needed and downloading such dependencies can take hours.
Nowadays both .deb and .rpm have both got very good with dependencies, dependency hell is a lot less common then it used to be as the two package managers have been allowed to progress.
I think in the near future both would only need 2 or 3 dependencies to function as right now dependencies have come down from up to 12 down to maybe 5, with the major DE's and open office excluded as they require more packages but even in those cases its gotten a lot lower.
But with this new package manager4 you cant change or modify the installer, no choice, no freedom and no luck when there is a major issue.
At least when something is wrong with apt it can be fixed, same as other package managers.
But what if this new one cant be fixed, it would make linux clunky and just too much like MS:
restrictive

Screwdriver0815

May 1st, 2009, 03:36 PM

I only partly agree with the guy in the video because:

1. drivers and stuff
of course its a mess with the audio and this has to be sorted out but speaking so generally as he does ("always when there is a Kernelupdate the video driver is broken") is wrong because it is not generally and necessarily so. When you install the drivers from the repo and NOT from the Manufacturers homepage its fine. This guy has more knowledge about Linux than me so for me this is really like "I search for something where Linux sucks because I did not find enough"

2. funding of projects:
he is always talking about applications like photoshop and that nearly everybody needs it but as you can not get it in Linux nobody wants to switch over from windows. To be honest, I can not hear this anymore.
What I have learned about open source is: when you write a program, you can either give it away without charging money, or you can sell it (take money for it). The only thing you have to do generally is to provide the source code to make it free. So when there is a market for (for example) photoshop, then Adobe could sell it. No matter if it is open source or not. But there is no market. Because all these people who "need photoshop" actually do not need it.
He says developers need food. Thats right and I totally agree with that. And this is also a thing which I do not understand: when you want to write a program because you need it, in the open source world you are free to put it into the internet. You also can charge for it. So when your application is a real killer, lots of people would buy it.
Why does nobody charge for the programs? As far as I understood open source, charging is legal!
Personally I would pay for all programs I use, because they are great and I think I am not the only one on this planet.

Red Hat and Novell charge for their Enterprise versions. Thats right and they earn money with it which also right.

So the big question for me is: why do developers give away their stuff without charging money and then complain about that they can not afford their food?

Of course, when you make the source code available like it is the rule of the GPL, people could compile it on their own, but personally I would not because its way too much effort for me. So I'd rather pay for it. Also in this case I think I am not the only on this planet who would do it like this.

And regarding the repo's and click n run stores:
its easy like nothing to do such things. Canonical has it already for PowerDVD and other apps. just go shopping and when you pay, you get a repo-link which you put into your system and you get the software and the updates. Whats wrong with that? This can be done with all apps, no matter if they are closed source or open source.

3. standardisation:
there must have been a reason for .deb and .rpm. Why did they do it like this in the past? Of course it would be better to have one standard but now its too late for it because there are such a lot of different programs in .deb and .rpm that this would be a huge effort to put all of them into one single fileformat.

SunnyRabbiera

May 1st, 2009, 04:13 PM

there must have been a reason for .deb and .rpm. Why did they do it like this in the past? Of course it would be better to have one standard but now its too late for it because there are such a lot of different programs in .deb and .rpm that this would be a huge effort to put all of them into one single fileformat.

Well I do know that the first format installer for linux was .tar.gz, it was definitely on slackware for a while.
The split most likely happened when debian first came out, when .deb was still fresh as well as apt, the reasoning for .deb was probably because so much was available in .tar.gz from software installers, to source code, to themes and everything in between, heck its still like that.
But at some point came redhat who saw the early issues of .deb and developed .rpm.
The two took very separate roads from there on in, but these days the two are similar enough to develop separately but they have come a long way since the days of dependency hell.

Icehuck

May 1st, 2009, 04:56 PM

If you're suggesting a separate repo and packaging system for ALL Linux distros, that's not very likely, nor very worthwhile IMHO. It's like this: Debian-based distros are designed from the ground up for dpkg and APT. The security model is based on it; all of the updaters and other applications are based around it too. Same goes for Red Hat-based distros. I've tried running YUM on an Ubuntu system. It didn't work well at all. If you include a new system, there's nothing to say a pre-compiled and packaged application will work correctly on all systems. But that's the beauty of source code. It will. There's a set way of doing it for all distributions, and if it's needed it works great.

So you think that having download pages such as this one (http://www.virtualbox.org/wiki/Linux_Downloads) is fine? It's a flippin joke to have that many different installers for the same software.

Polygon

May 1st, 2009, 05:01 PM

So you think that having download pages such as this one (http://www.virtualbox.org/wiki/Linux_Downloads) is fine? It's a flippin joke to have that many different installers for the same software.

is it too much to expect that the user downloading something has at least a shred of intelligence and knows what distro and architecture they are running?

Icehuck

May 1st, 2009, 05:03 PM

is it too much to expect that the user downloading something has at least a shred of intelligence and knows what distro and architecture they are running?

Why do developers have to waste time and create that many packages just to get their software out there? They could just give source, but then they wouldn't many people using their software.

SunnyRabbiera

May 1st, 2009, 05:09 PM

So you think that having download pages such as this one (http://www.virtualbox.org/wiki/Linux_Downloads) is fine? It's a flippin joke to have that many different installers for the same software.

But Virtualbox is different in each system, like Opera.
QT3 and 4 play a big part on why there is so many variants of Vbox

Icehuck

May 1st, 2009, 05:11 PM

But Virtualbox is different in each system, like Opera.
QT3 and 4 play a big part on why there is so many variants of Vbox

Isn't that because of a lack of standards? Wasting more developer time...

Giant Speck

May 1st, 2009, 05:14 PM

Isn't that because of a lack of standards? Wasting more developer time...

And developers gotta have time to eat! :p

Icehuck

May 1st, 2009, 05:19 PM

and developers gotta have time to eat! :p

bingo!

SunnyRabbiera

May 1st, 2009, 05:21 PM

Isn't that because of a lack of standards? Wasting more developer time...

Well look at what Vbox installs on, OpenSuse 10.3, Mandriva 2007.1, Ubuntu Dapper and Gutsy, most of those have QT versions below QT4.
QT4 is still new, and its also still very unpopular by many.
QT4 is still very new, older distros that still have support have packages for them.
Also Q4 had some revisions.
Opera is very similar, it too offers a flexible QT as QT4 is not yet fully there yet...

toupeiro

May 1st, 2009, 05:25 PM

I agree with SunnyRabierra. Linux needs to stay diverse in the contexts this video, and other posts are referring to. I will point out that linux has adopted many standards which have been standards for decades:

runlevels
standard naming conventions
standard compilers
standard ACL's

I say these are standards because you see them across variants of Linux and UNIX. Compared to windows, which has no standards, but sets "standards" with EVERY release of their OS which they rarely abide by. I've seen c:\windows, c:\WINNT c:\WINDOWS, I've seen the registry absolutely mutilated beyond recognition of its original intent from Microsofts own apps and patches, I've seen ACL and security changes, both in ideology and technicality, with just about every windows release. So how can you exactly say there are "standards" for linux's only real widespread competition? Where is your baseline, and is it really a baseline you want to hold linux to? Not me...

Linux has standards imposed where standards need to be, which allows for everything else to mold as technology demands it to mold. IMO, this is the only reason linux is even in any sort of the state it is in today from a visibility standpoint. Sure linux has problems. Software WILL ALWAYS have problems. I'm not above saying linux has, and always will have, plenty which needs fixing. Why the linux desktop sucks is an opinion. To me, this doesn't hold any water. I'm perfectly capable of forming my own informed opinion. Simply put, if it sucked, I certainly wouldn't have been using it the last 3-4 years as my primary OS. I do not consider myself a linux apologetic, I just feel confident that for my uses of a desktop, linux does exceedingly better than anything else out there. Considering also that I put food on the table and a roof over my head based on how I support and implement linux, UNIX, and some limited Windows still (not near as much as I used to), I must be accurate on some level, or I probably wouldn't be doing what I do for a living.

Icehuck

May 1st, 2009, 05:27 PM

Well look at what Vbox installs on, OpenSuse 10.3, Mandriva 2007.1, Ubuntu Dapper and Gutsy, most of those have QT versions below QT4.
QT4 is still new, and its also still very unpopular by many.
QT4 is still very new, older distros that still have support have packages for them.
Also Q4 had some revisions.
Opera is very similar, it too offers a flexible QT as QT4 is not yet fully there yet...

This is why commercial products(photoshop or AutoCad) are not available for Linux. There is way to much going on and trying to support every version would be too much of a headache.

toupeiro

May 1st, 2009, 05:37 PM

This is why commercial products(photoshop or AutoCad) are not available for Linux. There is way to much going on and trying to support every version would be too much of a headache.

This is different from AERO and UAC how? Compared to packaging differences and library versions, I'd say microsoft gives vendors much larger headaches than linux would.

Phasmus

May 1st, 2009, 05:41 PM

.
...
In order to limit choice and maintain a standard there might be a need of a non GPL licence, after all its GPL that allows there to be multiple linux distros.
You are not allowed to change what kind of packages you get and dont get, you are not allowed to maintain the OS at your own pace, instead you are forced to upgrade packages, no choice in whsat is upgraded and when.
Then this .lim system needs 10 different dependencies for not just a few apps but all of them, there is no control over how many packages are needed and downloading such dependencies can take hours.
...
But with this new package manager4 you cant change or modify the installer, no choice, no freedom and no luck when there is a major issue.
At least when something is wrong with apt it can be fixed, same as other package managers.

...just out of curiosity, do you ever get irked at MS for their failure to adhere to web standards? How they run off and do their own thing, and everyone else either has to follow, or is unable to access a bunch of content? Would the internet be better (freer?) if there were half a dozen different, incompatible versions of HTML and every browser and web developer could decide which they wanted to support/implement?

Packaging standards would address the same problem in about the same way as web standards. Someone (theoretically a lot of fairly bright someones representing the major distros) writes a spec. All the distros that want to play along then support that standard with their own package management software (I would guess while maintaining legacy support for whatever they used before too).

Nobody cracks a whip or enforces a license. The idea is, if you want to reap the benefits of the work everyone else is doing under this standard, you follow the same standard. If you want to run off and do your own thing you can, but your devs will have to work harder and your users won't have as easy access to all the stuff out there that does follow the standard. Similarly, if a single web site decided to implement and use their own ZTML (Zowie Text Markup Language) instead of HTML they're free to do it, but they're working harder and they've restricted their audience.

If it takes a hypothetical closed-package-license that no distro would ever touch, a package system worse than anything we have now and a cabal of freedom-hating, jack-booted package enforcers to make this sound like a bad idea, it probably isn't a bad idea.

days_of_ruin

May 1st, 2009, 05:43 PM

ugh... i hate when people call it "uh-bun-too"

:-\"

Once you start pronouncing ubuntu one way its really hard to change.

toupeiro

May 1st, 2009, 05:58 PM

...just out of curiosity, do you ever get irked at MS for their failure to adhere to web standards? How they run off and do their own thing, and everyone else either has to follow, or is unable to access a bunch of content? Would the internet be better (freer?) if there were half a dozen different, incompatible versions of HTML and every browser and web developer could decide which they wanted to support/implement?

Packaging standards would address the same problem in about the same way as web standards. Someone (theoretically a lot of fairly bright someones representing the major distros) writes a spec. All the distros that want to play along then support that standard with their own package management software (I would guess while maintaining legacy support for whatever they used before too).

Nobody cracks a whip or enforces a license. The idea is, if you want to reap the benefits of the work everyone else is doing under this standard, you follow the same standard. If you want to run off and do your own thing you can, but your devs will have to work harder and your users won't have as easy access to all the stuff out there that does follow the standard. Similarly, if a single web site decided to implement and use their own ZTML (Zowie Text Markup Language) instead of HTML they're free to do it, but they're working harder and they've restricted their audience.

If it takes a hypothetical closed-package-license that no distro would ever touch, a package system worse than anything we have now and a cabal of freedom-hating, jack-booted package enforcers to make this sound like a bad idea, it probably isn't a bad idea.

But it doesn't take a closed license to do this. Microsoft didn't do what they did to improve code, as history will show given how horrible activeX and other MS markup languages were from a security standpoint. They did it to control a market from getting away from them, not by superior standards, but by unethical business.

the world of protocol and hardware greatly survives by standards. Committee's like IEEE, which I had the pleasure to be a member of before bad economic times for my employer caused me to let my membership go, are instrumental in developing standards which make it possible for technologies to be massively compatible, and secure. I loved being a part of that organisation, because I truly love standards, but at the levels where they make sense. If I wanted to build a IEEE802.3 capable device myself, all I really have to do to make it work is follow the 802 standards which are available freely to anyone who wants them (http://ieee802.org/) This, I believe, would have been the natural path for software had Microsoft not been at the helm at these crucial times in Personal computing.

geoken

May 1st, 2009, 06:20 PM

Well lets give an example of why not having a one size fits all approach is a good thing:
Say all the package managers were merged into one, called the Linux install manager and it uses .lim packages to work.
In order to limit choice and maintain a standard there might be a need of a non GPL licence, after all its GPL that allows there to be multiple linux distros.
You are not allowed to change what kind of packages you get and dont get, you are not allowed to maintain the OS at your own pace, instead you are forced to upgrade packages, no choice in whsat is upgraded and when.
Then this .lim system needs 10 different dependencies for not just a few apps but all of them, there is no control over how many packages are needed and downloading such dependencies can take hours.
Nowadays both .deb and .rpm have both got very good with dependencies, dependency hell is a lot less common then it used to be as the two package managers have been allowed to progress.
I think in the near future both would only need 2 or 3 dependencies to function as right now dependencies have come down from up to 12 down to maybe 5, with the major DE's and open office excluded as they require more packages but even in those cases its gotten a lot lower.
But with this new package manager4 you cant change or modify the installer, no choice, no freedom and no luck when there is a major issue.
At least when something is wrong with apt it can be fixed, same as other package managers.
But what if this new one cant be fixed, it would make linux clunky and just too much like MS:
restrictive

I don't even know where to begin with this.

First off, your entire premise is based off a completely unfounded assumption. Why would the package manager not be GPL'd? By your logic there is no such thing as a GPL'd web browser because maintaining a standard is incompatable with the GPL according to you.

Secondly, who's even telling you to use a package manager? Just extract the package with any archive extractor and install from source. And why would the package manager force updates? That doesn't even make sense.

karellen

May 1st, 2009, 06:31 PM

This is different from AERO and UAC how? Compared to packaging differences and library versions, I'd say microsoft gives vendors much larger headaches than linux would.

Windows has backwards compatibility. I can run applications written for XP in Windows 7, 7-8 years after the release of XP. Show me a single Linux distro that does that ;)

toupeiro

May 1st, 2009, 06:36 PM

Windows has backwards compatibility. I can run applications written for XP in Windows 7, 7-8 years after the release of XP. Show me a single Linux distro that does that ;)

hah ok, I'll do you better than that. I took a tar file from a vintage 1995 SGI Indigo UNIX machine and extracted that tar file onto a RHEL5.1 64-bit machine and typed: make ; sudo make install, and guess what. a 14 year old UNIX application was working on Linux. Thats called cross platform, backwards compatibility. Something Windows has NEVER done.

And, your XP applications working in 7 are a 100% hit and miss scenario. Face it, you got lucky, and thats all, because Microsoft makes vendors go through logo compliance (http://www.microsoft.com/whdc/winlogo/default.mspx) with 3rd party software for every single version of windows. If it doesn't pass that compliance, it is officially unsupported.

Windows 7 software logo compliance toolkit. (http://www.microsoft.com/downloads/details.aspx?displaylang=en&FamilyID=27028822-b172-4cec-91a3-26b610a4da79)

karellen

May 1st, 2009, 06:43 PM

hah ok, I'll do you better than that. I took a tar file from a vintage 1995 SGI Indigo UNIX machine and extracted that tar file onto a RHEL5.1 64-bit machine and typed: make ; sudo make install, and guess what. a 14 year old UNIX application was working on Linux. Thats called cross platform, backwards compatibility. Something Windows has NEVER done.

And, your XP applications working in 7 are a 100% hit and miss scenario. Face it, you got lucky, and thats all, because Microsoft makes vendors go through logo compliance with 3rd party software for every single version of windows. If it doesn't pass that compliance, it is officially unsupported.

well, if we compare apples with oranges, you're right. I wasn't talking about compiling stuff(something that an average user doesn't know to do it anyway), but about binary packages. say .deb, say .rpm ;)

Namtabmai

May 1st, 2009, 06:43 PM

Windows has backwards compatibility. I can run applications written for XP in Windows 7, 7-8 years after the release of XP. Show me a single Linux distro that does that ;)

Only when it suits them, pre-XP stuff can be hard getting to run on XP and I doubt Windows 7 will be any better.

toupeiro

May 1st, 2009, 06:47 PM

well, if we compare apples with oranges, you're right. I wasn't talking about compiling stuff(something that an average user doesn't know to do it anyway), but about binary packages. say .deb, say .rpm ;)

pft. Ok, binary packages then. MSI/MSP/MST, EXE, how about WyseScript, Installshield, NetOpSys, Nullsoft, FEAD,. Windows has 2-3 times the amount of package formats than Linux, and absolutely nothing like alien which converts them...

Apples an oranges is not exactly what my comparason was, but by making a point of binary packages, its like comparing candied apples to apples. One may taste sweeter and to some, look better on the surface. However, eat enough of them and watch what happens to your health. Same thing applies to your OS.

xArv3nx

May 1st, 2009, 06:47 PM

This is different from AERO and UAC how? Compared to packaging differences and library versions, I'd say microsoft gives vendors much larger headaches than linux would.
Not really.

I mean, sure, it MIGHT (according to you, which I disagree with) be more work on Windows Vista.. but, then again, I'd think you would make more money on the Windows Vista platform due to the amount of users it has. Therefore, the amount of work you put into actually packaging it for Windows Vista (which methinks is virtually none), you'd still get good payoff on it.

Let's face it, the Linux community is full of free/GPL-licensed fruitcakes. If you have a proprietary product, it'll def. sell better on Microsoft Windows.

wsonar

May 1st, 2009, 06:50 PM

Haven't watched the vid

there are some lyrics by a group called The Cure

the song is jumping someone else's train

"everyones happy where finally all the same because everyones jumping everybody else's train"

sure we would like all are computers to be able to do anything any computer can do and do it the best way that's only logical

we have a few options

windows
Mac
Linux
BSD
make our own linux distro

I like the fact that I use Ubuntu or another linux variant most people haven't even seen heard of or know anything about(that's just me)and I want my proprietary apps two ( we as humans are very demanding) and eventually the squeaky wheel gets greased

I'm the guy that wants to know how to do everything on anything
not everybody is like this

linux gives choice you can go with and easy distro like ubuntu or mint(haven't even used mint but heard how easy it is)

or something more challenging like a slack, or bsd

this post doesn't really have a point except trying to find the fine line between not wanting to be mainstream and having all the luxuries that would come with it

Therion

May 1st, 2009, 06:58 PM

"Marketting" ... ???

Very professional, dude.

toupeiro

May 1st, 2009, 06:58 PM

Not really.

I mean, sure, it MIGHT (according to you, which I disagree with) be more work on Windows Vista.. but, then again, I'd think you would make more money on the Windows Vista platform due to the amount of users it has. Therefore, the amount of work you put into actually packaging it for Windows Vista (which methinks is virtually none), you'd still get good payoff on it.

Let's face it, the Linux community is full of free/GPL-licensed fruitcakes. If you have a proprietary product, it'll def. sell better on Microsoft Windows.

While I agree with you that you could make more money easily by putting your product on windows (for obvious reasons of visibility of that platform) I disagree linux doesn't host major sales of proprietary software because of its GPL software base. Why don't you check out companies like Halliburton/Landmark, Dynamic Graphics Inc (dgi), or paradigm (pdgm.com), or motorola's canopy prizm products, and see for yourself just how much money can, and is being made with proprietary software on linux. These are just a few of many global companies with software offerings that primarily cater to Linux today.

days_of_ruin

May 1st, 2009, 07:10 PM

Is there a way to download this?

Therion

May 1st, 2009, 07:20 PM

Is there a way to download this?
Yup: https://addons.mozilla.org/en-US/firefox/addon/3006

Npl

May 1st, 2009, 07:57 PM

"Marketting" ... ???

Very professional, dude.Thats how arguments are made, your awesome dude :-k

Not watched the whole presentation yet, but hes spot on with x.org (DIEDIEDIE), lack of quality in the huge majority of unfunded OS-Projects and the audio-mess.

geoken

May 1st, 2009, 08:04 PM

While I agree with you that you could make more money easily by putting your product on windows (for obvious reasons of visibility of that platform) I disagree linux doesn't host major sales of proprietary software because of its GPL software base. Why don't you check out companies like Halliburton/Landmark, Dynamic Graphics Inc (dgi), or paradigm (pdgm.com), or motorola's canopy prizm products, and see for yourself just how much money can, and is being made with proprietary software on linux. These are just a few of many global companies with software offerings that primarily cater to Linux today.

Those aren't really good examples. The video is talking about small developers not being able to make money with apps like video editors and entry level digital audio workstations. You're trying to counter this with corporations selling massive hardware mainframes (and the accompanying software) which is not even remotely comparable.

The presenter starts off by saying embeded Linux and Linux on the server (which would include specific use workstations) is awesome. It's Linux on the desktop he's arguing about.

Twitch6000

May 1st, 2009, 08:08 PM

That video makes many great points and I think if anyone here truly cares about Linux then you will take some advice from the video.

I know I will be donating some money to my favorite projects now.

Bölvaður

May 1st, 2009, 08:21 PM

I know I will be donating some money to my favorite projects now.
O:)

Standardized packages are not the same as a standardized package management system. I thought I wouldn't have to say this but... ok let's say we got a standard named "Data emerging binary" or .deb which is a package like we know them.
Then we have programs like "Add stuff / Remove stuff", "Synoptic Package System", "app-get" and "apptitude". Each of them could handle the packages differently and some be GPL or not... we dont care.

As long as every distro uses the same packages you only have to make 1 package for all distros using that standard.
But there are many programs that handle those packages which you can pick and choose from.

SunnyRabbiera

May 1st, 2009, 09:04 PM

This is why commercial products(photoshop or AutoCad) are not available for Linux. There is way to much going on and trying to support every version would be too much of a headache.

Nonsense, they could easily code a version of autocad or photoshop, but its not the formats its the money we can give them.

Screwdriver0815

May 1st, 2009, 09:36 PM

Nonsense, they could easily code a version of autocad or photoshop, but its not the formats its the money we can give them.

all the applications like Photoshop, CAD, and so on have a professional background. I mean, no average user would install a CAD system on his home desktop only to play around with it.

In the professional field aka companies windows is the rule. Maybe except those who do a lot of graphic stuff - they use Mac's.
What I want to say is, that there is no market for photoshop or CAD systems at the moment. Which does not mean that there are no people who want these apps to be ported to Linux - but they are way too rare to make the software companies think about a business case.

The other side is simulation software: at work we have to do a lot of simulations and what do we use for this? Red Hat Enterprise Linux. Why? Because you can get the best and most reliable simulation software only for Linux. And Linux is the OS where you can rely on that it doesn't crash when it has to perform Tasks which last weeks - not hours, weeks.

What I want to say: if there is a business case, the companies will do applications for Linux. But there is no business case.
The right moment will come, when more big companies have switched over from windows to Linux also on their desktops and start demanding CAD software and all the other stuff.
But this takes some time.

In the meantime the open source world could sort out this issue with the payment of the developers. Because this in my eyes is the most important thing. But I don't have any clue how to solve this beside charging money for the apps which are made. I think, donations are not the way like it should be as it is not reliable enough for the developers.

So, what I would like to know is: why is it grown like it is now, that most of the apps are free of charge? Why do commercial distro's like Mandriva have problems, making money while Red Hat and Novell have a solid funding?

chris200x9

May 1st, 2009, 09:49 PM

I disagree with the universal package manager, I mean we have that...SOURCE CODE...and don't give me "it's too hard" 99% of the time it's [code] ./configure make sudo make install[\code] next "what about non-open source programs?" I guess no one else has figured they are the same binary...I run arch I've been known to get binaries out of debs if I can find something in the repos or aur...

geoken

May 1st, 2009, 10:00 PM

I disagree with the universal package manager, I mean we have that...SOURCE CODE...and don't give me "it's too hard" 99% of the time it's [code] ./configure make sudo make install[\code]

Until you run into hard to find dependencies. You'll probably feel differently when you're scouring mailing list archives to find a dependency of a dependency of a dependency.

Paqman

May 1st, 2009, 10:07 PM

Good video, and I find myself agreeing with most of what he says. The fact that there's no financial incentive for companies to fund the development of desktop apps is indeed a problem.

I also find myself agreeing with the state of packaging. We do need to standardise. I also find it a bit ridiculous that one of the responses from his audience was "But packaging is what differentiates the distros!". So what? We don't need to have lots of distros. In part the reason we have so many is because of the fractured package landscape. If settling on a package standard made certain distros obsolete then that's still a win for Linux as a whole.

I do wonder if the fact that Ubuntu has become a bit of a juggernaut amongst distros will lead to .deb as the de facto standard. Personally I couldn't care less if Ubuntu switched to .rpm, but we need to pick one standard and stick to it.

As for sound, eugh! We need to get all the sound system devs into one room and have them fight to the death or something. It's a mess.

Paqman

May 1st, 2009, 10:09 PM

Yup: https://addons.mozilla.org/en-US/firefox/addon/3006

Not actually necessary. Flash videos can be found in your /tmp. Just let it buffer all the way to the end of the file and you can copy it to anywhere you like.

zenithdave

May 1st, 2009, 10:53 PM

I can only speak from a narrow band of interest but he is spot on about about many audio systems.

Watching that inspired me to check out the Ardour site :guitar:

Will we have a fund to pay a top Developer ? i would contribute.

geoken

May 2nd, 2009, 12:08 AM

I also find myself agreeing with the state of packaging. We do need to standardise. I also find it a bit ridiculous that one of the responses from his audience was "But packaging is what differentiates the distros!". So what? We don't need to have lots of distros. In part the reason we have so many is because of the fractured package landscape. If settling on a package standard made certain distros obsolete then that's still a win for Linux as a whole.

Agreed, if the only thing differentiated a distro is it's package system (which in the end is delivering the same apps) then it probably shouldn't exist anyway.

I think a lot of people cling dogmatically and illogically to the choice == freedom argument. They would rather have an arbitrary choice between multiple low quality apps/frameworks then be able to do some of the things that would be possible if a single, quality framework/app existed. Either that, or they're under the impression that there are an infinite amount of developers and this resource isn't strained by a dozen parallel apps.

visionaire

May 2nd, 2009, 12:09 AM

Paqman

May 2nd, 2009, 12:25 AM

I think a lot of people cling dogmatically and illogically to the choice == freedom argument. They would rather have an arbitrary choice between multiple low quality apps/frameworks then be able to do some of the things that would be possible if a single, quality framework/app existed. Either that, or they're under the impression that there are an infinite amount of developers and this resource isn't strained by a dozen parallel apps.

Absolutely. The main beef the guy raised in the video was that there's a lot of duplication of effort going on. Every dev that's packaging is one that isn't working on an app.

So yes, broadly speaking choice is good. But if you've got so much choice that it's affecting quality, then it's only logical to rationalise.

dragos240

May 2nd, 2009, 12:28 AM

I couldn't watch any more than 30 seconds of it. TL: DW,
can someone explain the points he made.

geoken

May 2nd, 2009, 12:35 AM

I like diversity, i like the fact that i can choose distros, DE, themes and so on, but like somebody said above, i would like that audio and video work fine, even if there are diversity en these fields also...

I'm using Hardy because, it's the only one where i cant watch .rmvb videos without suck up my CPU like ans insolent vampire lol

But i VERY LIKE DIVERSITY, i don't want other windows

PD: English is not my native language, so, sorry for the semantic madness =)

I'd argue that if the people working on different DE's collaborated on something you'd still have the same choice, the only difference would be that said choice would be gotten via internal options rather than competing products which are 90% similar.

Lets make a hypothetical situation. For the sake of this argument we're going to ignore the different languages involved and any possible technical hurdles (since that isn't relevant to the question). Imagine Gnome adopted Kwin compositing. Gnome still looked and acted exactly like it always did, we just use kwin composting (instead of the current defacto compiz). Lets also assume the compiz devs move to kwin now that compiz is functionally dead. With the compiz plugin makers now making kwin composting plugins, what specific end results of your previous ability to 'choose' will you know be missing out on.

michaeldt

May 2nd, 2009, 12:52 AM

I think pretty much everything in his video could be solved with what seemed to be the primary goal of his presentation. Funding. Market forces will bring about change whenever money is involved. If there were enough commercial applications for Linux, then all it would take is for those developers to say, "We're only making .debs" and suddenly any linux distro which wants to have some market share will need to support that. Right now, there's no incentive.

Quality can also improve when there is financial backing. Look at what Canonical has achieved with Ubuntu. Apply the same to the various Linux projects and you suddenly have quality software rushing forward. Of course, bringing commercialisation and standardisation into Linux can harm the community. We don't want commercial developers to force distros to use a package format just because they refuse to use anything else. What we need is the desire to switch because what we're switching too makes sense and improves on what we already have. Similarly, is it fair for anyone to make profit from a distro, when in doing so they are including software produced by developers trying to scrape toegether a living, without passing on any of that success?

I don't really know what the solution is here but developers do need to get paid. The software I use for raw image processing is commercial and I pay for that. It's developed for Windows, Mac and Linux. Yet when I load up Ubuntu, I haven't paid for any of that software. People have, for a lot of it, developed it for little or no reward. Yet if Canonical charged, I dunno, £10 for each release (well not charged, but made provision for donations with the OS still being 100% free for those not wanting/able to donate) and then ensured that funding was used to support the various projects, I'd be more than happy to pay that. On the other hand, if I went and made individual donations for ALL of the software I use, well, I just couldn't afford to.

I also think that if commercial developers and distros worked together to streamline the process of finding, buying and installing software, then there could be potential income for distros through the actual process of distributing and testing software. In other words, if Adobe release Photoshop for Linux, then Canonical would help them package it for Ubuntu, provide the means to sell the software inside the OS as well as provide the distribution channel via Add/Remove or something similar and in return, Canonical gets a portion of the sale.

But funding aside, when it comes to usability, I do sometimes wonder at a lot of the duplication around. It seems that whatever application you search for, there's tons of free applications which do the same thing. I'm sure they have their reasons, but if an application is well liked, well used and well supported, instead of creating something from scratch, wouldn't it make more sense to find ways of assisting in the development of existing projects instead of trying to duplicate them?

All that aside though, I do think Linux has come a long way and the only reason these are current issues is because Linux is on the cusp of a major event in it's history. It's at the point where it has the potential to be a major player in the OS market and create a real upset. The problem is finding out how to make that next big step and in what direction to go.

arashiko28

May 2nd, 2009, 01:06 AM

Ubuntu founder, Mark Shuttleworth created a fund with no less than 10 million dollars, out of his own pocket, he said it would be an emergency funding in case of God forbids, Cannonical withdraws it's support from Ubuntu.

Now, our, duty is to enlarge that funding. I think that even a buck is too low, sure we have some very young users, but even an 8 years old, handles more than 10 bucks in 6 months, how about it?

You, kindly, every 6 months, while downloading or asking for your free version donate 10 dollars, or whatever you want. Buy a shirt if you're the kind that doesn't like to give money away, buy a pen, the store has some pretty cool things, I'm saving myself to get one of those backpacks.

For me, Linux is one of the greatest OS's ever, and Ubuntu the most user-friendly, perfect for new users, Linux introduction, and popular distro. I don't dual boot anymore, I only use Ubuntu, and it's true that is a work in progress, but as world situation is right now, It won't stay free for much longer. We are the ones who can make it different, and are just complaining and throwing the towel after the first bump and saying enough, I'm going back.

Thanks to all of the people who takes it's time to just answer forums questions, I have now a perfectly running OS, an stable version of 9.04 64-bit with no flaws. And now, I think I have experience enough to do the same, I wander through the forum looking for some question I know the answer, and that's one of many ways for contributing.

English language has that sort of let's call it "mix up" one word, written on the same way and used on the same sentence has several meanings.

"Ubuntu is and will remain free of charge", that's the promise, are we willing to see one great OS crash and burn for lack of fundings? It's not that they're broke NOW, but in a couple of years things may change.

You're free to do whatever yo want with it, but remember that it took money to create it, working hours, several testings, it is nice to have the final product, so let's keep it going.

1 dollar every time you download a new version, but then, you might get all cocky and rant again because you feel you paid for it...:(

visionaire

May 2nd, 2009, 03:24 AM

I'd argue that if the people working on different DE's collaborated on something you'd still have the same choice, the only difference would be that said choice would be gotten via internal options rather than competing products which are 90% similar.

Lets make a hypothetical situation. For the sake of this argument we're going to ignore the different languages involved and any possible technical hurdles (since that isn't relevant to the question). Imagine Gnome adopted Kwin compositing. Gnome still looked and acted exactly like it always did, we just use kwin composting (instead of the current defacto compiz). Lets also assume the compiz devs move to kwin now that compiz is functionally dead. With the compiz plugin makers now making kwin composting plugins, what specific end results of your previous ability to 'choose' will you know be missing out on.

Yes, you make a valid point, but i'm not very techy in matters of the internal programming of the DE's but, maybe it'll be too hard to implement things like that, i dunno really

But i completely understand your vision =)

Zorael

May 2nd, 2009, 10:06 AM

I'd argue that if the people working on different DE's collaborated on something you'd still have the same choice, the only difference would be that said choice would be gotten via internal options rather than competing products which are 90% similar.

Lets make a hypothetical situation. For the sake of this argument we're going to ignore the different languages involved and any possible technical hurdles (since that isn't relevant to the question). Imagine Gnome adopted Kwin compositing. Gnome still looked and acted exactly like it always did, we just use kwin composting (instead of the current defacto compiz). Lets also assume the compiz devs move to kwin now that compiz is functionally dead. With the compiz plugin makers now making kwin composting plugins, what specific end results of your previous ability to 'choose' will you know be missing out on.

Yeah, but then some (evil GNOMEs!) *will* complain that there are too many options and fork.

k2t0f12d

May 2nd, 2009, 11:45 AM

As well meaning as the host in the video may be, his presentation is just mast#$%^&*@n. There are irreconcilable technological problems that they don't even talk about concerning the unified package management format. DE's are programmed differently in different languages. The entire premise attempts to cram the bizarre into the cathedral, and fortunately for us the bizarre cannot be controlled that way. At the very least, he is realistic enough to mention at least once that the different distributions will never come together and get behind one of the things he thought might make GNU+Linux more attractive (I've forgotten which that was).

The thing that GNU+Linux has that most other systems don't is unity in diversity. It is from that term we arrive at the institution of _university_. Free software may not do everything, or in the exact same way, but what we do is done in a evolutionary and democratic process. What the whole of our species needs to realize most is that we need power and control to be horizontally integrated in _all_ areas, not vertically, and not only in computer software development.

I could have simply shut it off (but didn't) as soon as he begged the audience to ignore the freedom issue of free software. Its fine if his presentation didn't have anything to with that, but if it were the case, he need only omit talking about it at all himself.

Lastly, the money thing. Speaking for myself, I don't lend any credit to a speaker who repeats over and over that money has to be paid for the recipient to eat. We all know that in the sick and twisted state of society under the monetary system that most people will need to get money one way or another in order to live comfortably. That's fine, and although I think the world would be a better place with a resource rather then money driven economy, I'm happy to contribute code and cash to the projects I can get behind and see a positive outcome in the end. Its just that I don't see that in what this guy is selling.

toupeiro

May 2nd, 2009, 04:06 PM

Those aren't really good examples. The video is talking about small developers not being able to make money with apps like video editors and entry level digital audio workstations. You're trying to counter this with corporations selling massive hardware mainframes (and the accompanying software) which is not even remotely comparable.

The presenter starts off by saying embeded Linux and Linux on the server (which would include specific use workstations) is awesome. It's Linux on the desktop he's arguing about.

I was following the overall digression of this post and comments made about proprietary software v/s OSS, in which case those examples are on par with AudoDESK (HUGE company) and Microsoft. Not sure where you are coming up with selling massive hardware mainframes. I happen to have supported all of these vendors products on commodity hardware. I'd say that makes it VERY comparable, expecially since these desktops ALL run linux...

I've also made full admission that it would be much easier for emerging companies and developers to develop on the windows platform to turn a profit, but lets face it: If you were offering something that actually needed to use hardware efficiently, Microsoft didn't (and arguably still doesn't) have a way to really do this until they brought 64-bit desktops out. The 64-bit linux desktop offerings are much more mature than Windows. So, even emerging companies stand to gain much more out of their product by offering it on linux depending on the discipline of the application. If you're writing greeting card software, put it on windows, you'll make more money...

kamaboko

May 5th, 2009, 05:27 AM

http://www.youtube.com/watch?v=YoYL4R3Te2s

Tipped OuT

May 5th, 2009, 05:35 AM

44:00 minutes? Too long, I phaze out after 2 miuntes. :P

kamaboko

May 5th, 2009, 05:40 AM

True, but it really does bring up a lot of failings that have to be talked about and solved. It's worth the time.

wiznillyp

May 5th, 2009, 05:44 AM

Video player doesn't allow seeking. Must be one of the reasons... ;)That is the first thing that came to my mind.

wiznillyp

May 5th, 2009, 06:04 AM

44:00 minutes? Too long, I phaze out after 2 miuntes. :P...or 160 characters

Browser_ice

May 18th, 2009, 09:44 PM

At first, when I saw the title "Why Linux suc (http://lunduke.com/?p=429)ks!", I was pissed off thinking its another video to trash Linux in general.But then when watching the video, I realized that he is not actually saying that Linux sucks and that people should completely drop it. Instead he is addressing some of the most common Linux problems inciting people to think that way and giving solutions. Some of those solutions do make sense.

So he is not trashing Linux in his video. He is giving solutions to get rid of that thinking "Linux sucks because ...".

What do you think about what he says ?

Should the Ubuntu Dev look at it to improve Ubuntu better ?

*** LINK ADDED ***
sorry about that

dragos240

May 18th, 2009, 09:46 PM

Link?

Giant Speck

May 18th, 2009, 09:47 PM

Link?

Remember that video a while back that you said was way too long, so you didn't watch it?

It's that one.

dragos240

May 18th, 2009, 09:48 PM

Ohh yes. Why desktop linux sucks. Right, I'm going to watch that.

Browser_ice

May 18th, 2009, 09:52 PM

Ohh yes. Why desktop linux sucks. Right, I'm going to watch that.

You are taking a conclusion based on the title only. That is what I had done originally but if you watch the video, you will see it is NOT ABOUT TRASHING LINUX but rather how to improve it to STOP PEOPLE THINKING THAT WAY.

So do not take a conclusion based on the title alone. I know it is not a good choice for a video title and he admits it. Do watch it.

swoll1980

May 18th, 2009, 09:53 PM

I watched that video. He pretty much just says the same things the rest of us do.

Browser_ice

May 18th, 2009, 10:44 PM

HUGH ????

Why is it that when I click on my OWN thread I started 30 min ago, I wind up in this one ??????

What happened to my thread ?

[added comments]
I guess some mighty powered admin decided to move my thread in here !

aysiu

May 18th, 2009, 11:43 PM

madjr

April 25th, 2010, 06:37 PM

interesting presentation by developers and users trying to tackle/fix major linux problems:

http://www.youtube.com/watch?v=YoYL4R3Te2s

thumbs up

note: dont pay attention to the "suck" in the title, is not a rant :)

PurposeOfReason

April 25th, 2010, 07:39 PM

That was last years linuxfest; I was at this years yesterday. Was at that talk too. This years was way better IMO.

NCLI

April 25th, 2010, 07:53 PM

Was the new one recorded too? I'd like to see more from this guy.

PurposeOfReason

April 25th, 2010, 07:55 PM

Was the new one recorded too? I'd like to see more from this guy.
It was. He said he'd have it up in a few days. His site, http://lunduke.com/

A lot of the stuff was the same, but I still think it's worth a watch when he puts it up.

madjr

April 25th, 2010, 08:02 PM

It was. He said he'd have it up in a few days. His site, http://lunduke.com/

A lot of the stuff was the same, but I still think it's worth a watch when he puts it up.

awesomeness

where you in the back or front? :)

Anything you specially liked?

are we progressing in some of the areas he mentioned?

PurposeOfReason

April 25th, 2010, 08:15 PM

awesomeness

where you in the back or front? :)

Anything you specially liked?

are we progressing in some of the areas he mentioned?
Was in the back, standing for an hour. Room was packed this time. He made quite a few good points on that we do, whether we like it or not, have to pay for some software if we want to compete with apple and microsoft. You want CS5 quality software, well that costs $400 a pop. Of course this roused everyone into saying we've done just fine without large money, but the linux kernel is paid for. So is Gnome and to a lesser extent, KDE. He made the same audio argument, said video got much further. Made fun of X.Org for always breaking and said they shouldn't release things as quick as they do. Oh, and he made a shot at RMS so that was great. :P

sdowney717

April 25th, 2010, 08:40 PM

You work at a job and you get money for what you do. I see no reason to not pay people to code and produce great software. Money is a powerful force, a motivator, the grease that gets things to happen and this is the way it should be really.

Странник

April 25th, 2010, 09:05 PM

Very cool video. thanks

Giant Speck

April 25th, 2010, 09:22 PM

A lot of the stuff was the same, but I still think it's worth a watch when he puts it up.

That's kind of sad, don't you think? A whole year later and the same problems are being discussed?

ikt

April 25th, 2010, 10:53 PM

That's kind of sad, don't you think? A whole year later and the same problems are being discussed?

There was an excellent response to this 'issue' by Elvis:
http://forums.overclockers.com.au/showpost.php?p=10362142&postcount=20

The author of this video has done nothing but look at his own immediate needs and assumed that they represent 100% of software users. Likewise he's assumed that Linux "needs to beat Windows" (when in all reality, most Linux users don't give a toss what Windows is doing). He, like so many, has fallen into the common trap of assuming his little microcosm represents the world at large.

Yes, I'm all for positive and swift changes to "Linux" (including "Desktop Linux"). No, his answers aren't sensible, nor even plausible.

arnab_das

April 25th, 2010, 11:00 PM

Chris Pirillo on Ubuntu:

http://www.youtube.com/watch?v=5vOUyK8KgdU

very straightforward comments, harsh, hard to accept, but true.

Excedio

April 26th, 2010, 12:07 AM

Chris Pirillo on Ubuntu:

http://www.youtube.com/watch?v=5vOUyK8KgdU

very straightforward comments, harsh, hard to accept, but true.

So I was watching this video and well...awesome advertisement. Check it out...

madjr

April 26th, 2010, 12:24 AM

So I was watching this video and well...awesome advertisement. Check it out...

wow nice

but dont click on Ad guys or they will lose money, type it in instead if you want to visit

madjr

April 26th, 2010, 12:35 AM

There was an excellent response to this 'issue' by Elvis:
http://forums.overclockers.com.au/showpost.php?p=10362142&postcount=20

The author of this video has done nothing but look at his own immediate needs and assumed that they represent 100% of software users. Likewise he's assumed that Linux "needs to beat Windows" (when in all reality, most Linux users don't give a toss what Windows is doing). He, like so many, has fallen into the common trap of assuming his little microcosm represents the world at large.

Yes, I'm all for positive and swift changes to "Linux" (including "Desktop Linux"). No, his answers aren't sensible, nor even plausible.

i dont think that response was excellent at all

Ubuntu is aimed at converting non tech users (mostly from windows)

So our aim is to "just work" (tm)

if "elvis" likes things the way they were (past tense) and likes to compile and fix everything by hand then he should stick with a distro like Gentoo or slackware, arch, etc.

tons of distros for him where you can spend all day tweaking/fixing and installing from scratch

richs-lxh

April 26th, 2010, 12:39 AM

Why desktop Linux Sucks, And What We Can Do About It

Firstly the Linux community needs to stop being so separatist and try to come together a bit more. Share knowledge and join together as a whole.
Split up a 1% entire desktop computer market share between all the distros, and that leaves a pretty insignificant impact for each distro. Even Ubuntu with its large user base. Windows and Mac users laugh at the little Linux hobbyists bickering at eachother.

Secondly, distros need to squash bugs, and keep them squashed before moving on to even more pretty guis and features. None of it is any good if you upgrade to your newer version and your old hardware that used to work perfectly for the last two releases is now broken. Solid base, with universal standards for all distros, then start adding the pretty features.

No good painting a block of flats a pretty colour and adding the latest windows and doors if the foundations are mud. Sooner or later, things will start to fall apart.

Start supporting Hardware manufacturers who support Linux. Boycott the ones who don't.

Buy games from companies who offer Linux versions as well. It may inspire other companies to start creating Linux native games if they think we'll buy them. Gaming is a huge drawback for Linux. Gaming is a huge worldwide market. If Linux could attract gamers, there would be a lot more publicity.

Just a few ideas.

chucky chuckaluck

April 26th, 2010, 12:46 AM

for free (as in, "i did not pay for it"), it doesn't really suck. most normal people (none of you) could be fine with it. for using a browser, playing some crappy cd and emailing someone no one else likes, most os's will do (except haiku and minix which really do suck, and anythingbsd if you want to watch youtube videos, which most normal people want to do).

phrostbyte

April 26th, 2010, 12:57 AM

Linux (and especially Ubuntu) is a great OS for my desktop as it is. :)

arnab_das

April 26th, 2010, 01:28 AM

So I was watching this video and well...awesome advertisement. Check it out...

most people use ad blockers. hence i didnt even see it :)

MaxIBoy

April 26th, 2010, 02:37 AM

Old thread, but a lot of that video is still relevant today. There are some solutions to the problems he raised.

I don't remember totally, but I think he didn't mention Autopackage; Autopackage isn't perfect by a long shot, but it is very good. For those who don't know, it's a way to build installers where the user experience is similar to windows MSIs, but it does dependency checking. This is done by looking for actual files instead of seeing if a package is installed, thus it is portable across many distros. The package itself is a bash script generated by Autopackage's developer tools. Autopackage isn't any good for distributing libraries or core system software, and it's not supposed to replace a conventional package manager, but for most applications it's great.

Excedio

April 26th, 2010, 03:18 AM

most people use ad blockers. hence i didnt even see it :)

That is very true...but now you know what you missed. :-)

madjr

April 26th, 2010, 05:19 AM

madjr

April 26th, 2010, 05:56 AM

Chris Pirillo on Ubuntu:

http://www.youtube.com/watch?v=5vOUyK8KgdU

very straightforward comments, harsh, hard to accept, but true.

weird

i didnt it feel it was harsh at all.

he was way more harsh in the past, when linux was harder to use.

but at least half his viewers are linux users, so he tries to keep things fair

Mr. Picklesworth

April 26th, 2010, 06:08 AM

I like the thread merging. Found an old post of mine, and upon reading it I am stricken by a great sense of progress!

I was talking about packaging formats and cmake. There's a recent project called BuildJ (http://live.gnome.org/BuilDj) which is really promising. It isn't a new build system, or a new packaging system or any of that, but a format for defining a build environment, then the build system is configured automatically. (And it happens to have a working reference implementation for my favourite build system).

Messing with Autotools from scratch may well be open-ended, but it's insanely complicated and there is a lot of boilerplate nobody really needs to touch in a normal case. (Besides, would anyone rather write in SH than Python with some JSON files? No? Good). BuildJ provides one very opinionated, simple to use system for one purpose. Extra functionality is implemented in plugins, so as not to torture you. It's there, but you aren't submerged in it with no escape.

I think it's a real breath of fresh air, and a big step forwards; Buildj could make it as easy to set up a build environment "the right way" as it is to do it in an IDE, which is one place where the competition has us completely beaten at the moment.

If that happens, and sees some real adoption, I think the disparate build and distribution tools will start to feel a lot less hostile.

purgatori

April 26th, 2010, 11:49 AM

He makes some good points, but I didn't watch the whole thing. I don't really care about 'the desktop', I care about 'the workstation.'

Shakz

April 28th, 2010, 05:56 PM

Why Desktop Linux (Still) Sucks. And What We Can Do To Fix It.

Watch here
http://www.jupiterbroadcasting.com/?p=1886

This was a presentation given by Bryan (duh) and posted on the Linux Action Show. I am admittedly a fan but all that aside I like the presentation a lot.

It kinda woke me up. I am only on a month of using Ubuntu but still felt it was important to contribute to my favorite software so far.....so I ordered a Lucid shirt yesterday and got a beenie too! Granted Ubuntu deserves more than that from me and in the future it will prolly get it. I have not reversed my close min max in Lucid as I plan to be a long term user and I know Ill get used to it.

Also made a $10 donation to Gnome Do and ordered a Gnome T-shirt from the affiliate on the Gnome website (gnome is supposed to get a portion of the sale.) As after trying 10 or so of the top dirstros no matter what flavor it is...I am a gnome man.

Thoughts on the presentation?

aysiu

April 28th, 2010, 06:02 PM

Is there a transcript so people don't have to watch the video?

If the presentation doesn't say desktop Linux needs to be marketed as a physical product people can try out in a physical store, then I don't agree with the video. Android didn't become successful by saying, "Hey, try to download this disk image and flash it on to your iPhone or Blackberry."

The main reasons Android became successful: it carried with it the Google brand name, it was available on multiple physical products people could purchase, and Verizon pumped a ton of money into its Droid advertising campaign.

If the primary means of migrating to Ubuntu (or some other desktop Linux) involves finding an .iso, downloading it, checking the download's integrity, burning it to CD or USB, checking your BIOS to boot from CD or USB, repartitioning your hard drive, installing, and then troubleshooting hardware compatibility, desktop Linux will always fail. Always. Same goes for Dell's offering untested Ubuntu options it doesn't advertise or price competitively or offer in many countries... and then recommending Windows 7 on the very page it offers Ubuntu on.

Doesn't matter if it has video games. Doesn't matter if CS5 gets ported to it. Doesn't matter if the interface looks pretty. Doesn't matter if we have one package type (.deb or .rpm, etc.). As long as most people are expected to install and configure the system themselves, it will still remain an obscure OS for nerds, geeks, and power users... or the friends and family members of those nerds, geeks, and power users.

NCLI

April 28th, 2010, 06:04 PM

Yep, great video(Watched it yesterday). The first one inspired me to become a Gnome Philanthropist, this one made me donate 100 USD to GIMP(Or at least I will next month). He highlights a lot of important issues, but also gives an idea of how much progress we have actually achieved.

Linye

April 28th, 2010, 06:44 PM

as long as most people are expected to install and configure the system themselves, it will still remain an obscure os for nerds, geeks, and power users... Or the friends and family members of those nerds, geeks, and power users.

qftt

LowSky

April 28th, 2010, 06:53 PM

I think Ubuntu Netbook Remix would look great on a tablet...

Dragonbite

April 28th, 2010, 07:06 PM

qftt
"qftt"?

wojox

April 28th, 2010, 07:07 PM

"qftt"?

Quoted for the truth.

gnomeuser

April 28th, 2010, 07:20 PM

I strongly disagree with his proposal that we trail on drivers.

What we need is automated performance and functionality regression testing. Like Chromium does, if we get slower it either needs to be justified or fixed (e.g. becoming a few procent slower at the expense of not tearing, locking up or crashing).

Packaging is a problem but something like Conary really makes packaging easy and it brings DVCS concepts into the packaging world. PackageKit solves the issue of having the same frontend commands. Yes we should work towards standardization but so far there hasn't been a big push towards this end. I think this is a project that should be undertaken as a FreeDesktop.org or Linux Foundation project. MeeGo might be a good base for such an effort.

A similar issue is that we don't have a platform, our best bet is likely going to be converging with MeeGo on something. This would give developers a solid API to code against across multiple device deployments including bindings to multiple languages.

Audio editing, we suck, agreed. A better platform such as MeeGo and bindings for more productive languages than C would help. Life is to short for C.

Games. Steam and Source is coming, I understand this will bring a number of great games to Linux and a market to buy them from. There is also a growing market of indie game companies which support Linux. Drivers are getting better. I think this isn't going to be a problem for much longer. With Linux being a platform for mobile, embedded (think TVs here) and desktops you as a developer get with MeeGo at least one API to code your game for all of these. I think this will be an important step towards making game development profitable for Linux.

Donations, sure. I am a friend of GNOME, I support the EFF. I paid for the Yo! Frankie game in advance to fund it. I think this can be done but the experience I had running the Nouveau pledge drive the real problem is infrastructure. How to get money from the users to a central place where the project can manage their funds. This is really hard currently and it is the SOLE reason why the pledge failed. People are willing to give generously but there is no good way to handle it.

I also think here that there is a lot of money he is ignoring in support contracts and development contracts for distributions such as Ubuntu. Additionally there are areas where we do need to buy support, media codecs, dvd players. Things that are impossible for distributions to ship or to fund development of for legal reasons. I bought the Fluendos products and they are awesome, I also know that a large part of the profit from them goes to fund GStreamer development.

aysiu

April 28th, 2010, 07:26 PM

I think this can be done but the experience I had running the Nouveau pledge drive the real problem is infrastructure. How to get money from the users to a central place where the project can manage their funds. This is really hard currently and it is the SOLE reason why the pledge failed. People are willing to give generously but there is no good way to handle it. Reminds me of this story from a couple years ago:
Is Money Useless to Open Source Projects? (http://www.codinghorror.com/blog/2008/07/is-money-useless-to-open-source-projects.html)

lykwydchykyn

April 28th, 2010, 07:34 PM

Doesn't matter if it has video games. Doesn't matter if CS5 gets ported to it. Doesn't matter if the interface looks pretty. Doesn't matter if we have one package type (.deb or .rpm, etc.). As long as most people are expected to install and configure the system themselves, it will still remain an obscure OS for nerds, geeks, and power users... or the friends and family members of those nerds, geeks, and power users.

I agree with you, but as I'm sure you know there is a chicken-or-egg problem here. Some of those things you mention are the stumbling blocks that prevent OEMs from pre-installing Linux distros or finding a niche for them when they do pre-install.

aysiu

April 28th, 2010, 07:35 PM

If there are people like me too lazy to watch the video, you can find the OpenOffice Impress presentation here:
http://lunduke.com/wp-content/uploads/2010/04/linux-sucks-2010.odp

After looking through all the slides, I have to disagree still. Improving the product quality isn't even half the solution. If the product is an engine you replace in a car, then it's a product you sell to car manufacturers and enthusiasts, not everyday drivers. Likewise, if the product you sell is an alternative operating system, then it's a product you sell to OEMs, not everyday computer users.

And marketing matters, big time. Look how many units of the iPad Apple has sold just because it's cool. It doesn't matter there's no webcam, no 1080p video, no Flash plugin, no USB ports. It looks cool! So people get it. And it's from Apple. So people get it. I even want to get one (well, not the first generation, of course). If Apple has taught us anything from the success of the iPod, iPhone, and iPad, it's that features and specifications are secondary concerns to most people.

People want a whole product that functions well at least in very basic tasks. That's the masses.

They don't want a half-product they have to figure out and tweak themselves so that they can do a ton of complicated tasks. That would be the geeks and power users.

lykwydchykyn

April 28th, 2010, 07:37 PM

Reminds me of this story from a couple years ago:
Is Money Useless to Open Source Projects? (http://www.codinghorror.com/blog/2008/07/is-money-useless-to-open-source-projects.html)

Or, similarly, this:
http://dneary.free.fr/gimp_bounties.html

lykwydchykyn

April 28th, 2010, 07:38 PM

If there are people like me too lazy to watch the video, you can find the OpenOffice Impress presentation here:
http://lunduke.com/wp-content/uploads/2010/04/linux-sucks-2010.odp

Is this just the same thing Lunduke's been saying for the last couple years, or is there anything new here?

EDIT: Ok, I read the slides. Nothing new. And frankly, a pointless waste of time.

Yes, everything he highlights as a problem is right on the money.
His "solutions" offer no real solution; he wants to be Alexander the Great slashing through the Gordian knot of FOSS development issues, but it ain't working.

For example: "There are too many audio systems. It's confusing. Pick one and go with it." How exactly do we "pick one"? Who gets to pick? What happens with all the software that suddenly stops working because it's based on an audio platform that didn't get "picked"?

Or for hardware issues:

Hardware Issues - FIXED
Distros need to stop revving X.Org so often.
If modern video cards work with a current release of a distro (or X.Org)... do not release the new version of that distro (or X.Org) if that video card is no longer fully supported.
Same goes for Wireless Cards and all other hardware.

Oh ok.... but wait, what if the new version adds support for some cards, and deprecates or has regressions for others? Who gets screwed over in that case?

Maybe I'm being unfair, maybe I'd understand his point of view better watching him talk, but from the slides and from what I've heard from him in the past it just sounds like the kind of mentality that says "if I shout 'FIX IT' enough someone will fix the problem".

aysiu

April 28th, 2010, 07:46 PM

I agree with you, but as I'm sure you know there is a chicken-or-egg problem here. Some of those things you mention are the stumbling blocks that prevent OEMs from pre-installing Linux distros or finding a niche for them when they do pre-install.
Well, Apple didn't get as far as it did by peddling Mac OS X to Dell, HP, Gateway, and the rest.

Apple created Mac computers.

Likewise, desktop Linux won't be successful if it relies on half-hearted efforts from Dell, HP, or Asus to preinstall and properly market Linux. We need something exactly like Apple... but more open. Originality is fine. Originality is not, however, the most important thing about open source. The most important thing is freedom.

If Ubuntu set up its own physical (brick and mortar) stores (maybe just one or two to start out with in New York and London) selling one or two very well configured and heavily tested good-looking machines that were guaranteed to work with all future versions of Ubuntu (no kernel regressions, no broken xorg drivers), don't you think Ubuntu users would be lining up in droves to get those?

Here are the major problems with current Linux preinstalled options:

Major OEMs Sometimes they offer. Sometimes they don't. Asus started off with Xandros on all its Eee PCs. Now where is Xandros? Dell used to have at least three or four Ubuntu options. In some countries, all they have is one netbook. In other countries, they don't have any Ubuntu options. HP used to have a Ubuntu netbook. No more. The Linux options tend to be pricier than similarly configured Windows options. This leads the vast majority of Linux users to prefer buying Windows preinstalled and then installing Linux themselves. Major OEMs sometimes have support staff who don't even recognize they offer Linux options. And the very Linux offering webpages will recommend Windows 7. The Linux options will sometimes not have options the Windows equivalents have. The Linux options aren't well-tested. Sometimes the Linux options don't support upgrades (9.04 preinstalled works fine, but if you upgrade to 9.10, no go). Minor OEMs: Virtually unheard of. Can you trust this business? Who are they? System76 ships to only the US. Even with smaller OEMs, there is problem that a lot of them simply don't exist for a lot of countries. They just don't look that cool... or they are simply rebranded Windows laptops. There is no physical store, and usually it's hard to find reviews (professional or otherwise) of the products. Find me, for example, a thorough review on the Hoverboard from ZaReason. I can't find one. They also can't price competitively, because of economies of scale. The only viable solution is for a company that has the money to do it (Canonical, for example) to set up a store, a real store, and sell uniquely designed Ubuntu preinstalled laptops. I don't mean an Asus laptop with the Windows key replaced with a Ubuntu key. I mean a uniquely designed Ubuntu laptop. Just as Macs are uniquely designed Apple laptops.

Of course, if Ubuntu doesn't pony up, there are supposed to be Chrome OS netbooks by the end of this year. If those use a Linux kernel, we have one kind of Linux preinstalled, even if it is just a web browser and not a full OS.

lykwydchykyn

April 28th, 2010, 08:12 PM

Well, Apple didn't get as far as it did by peddling Mac OS X to Dell, HP, Gateway, and the rest.

Apple created Mac computers.

Arguably, they didn't get as far as they could have because they didn't peddle to independent hardware vendors. There's a reason the PC dominated the market in the 80's and 90's, and heaven knows it's not thanks to a superior operating system.

Likewise, desktop Linux won't be successful if it relies on half-hearted efforts from Dell, HP, or Asus to preinstall and properly market Linux. We need something exactly like Apple... but more open. Originality is fine. Originality is not, however, the most important thing about open source. The most important thing is freedom.

If Ubuntu set up its own physical (brick and mortar) stores (maybe just one or two to start out with in New York and London) selling one or two very well configured and heavily tested good-looking machines that were guaranteed to work with all future versions of Ubuntu (no kernel regressions, no broken xorg drivers), don't you think Ubuntu users would be lining up in droves to get those?

There's something to that, and it could possibly work. But it's risky, because the minute Ubuntu goes into the hardware business, it goes into competition with Dell, HP, et al. Even if those vendors are half-hearted on the desktop front, they're much-needed allies on the server end where Canonical is trying to get a foothold.

Here are the major problems with current Linux preinstalled options:
<snip>

All excellent points, but would a fledgling Canonical hardware division be in a significantly better position than zaReason or system76 to price laptops competitively or offer variety? My guess is that these would still be "boutique" computers competing directly against Apple, probably at close-to-Apple prices due to scale, yet offering much more limited ISV/IHV support.

Still, I'd want to buy one, if I could afford it.

Dragonbite

April 28th, 2010, 08:22 PM

If there are people like me too lazy to watch the video, you can find the OpenOffice Impress presentation here:
http://lunduke.com/wp-content/uploads/2010/04/linux-sucks-2010.odp

Thanks for that link. Work doesn't like YouTube, but they don't block these sites.

And marketing matters, big time. Look how many units of the iPad Apple has sold just because it's cool. It doesn't matter there's no webcam, no 1080p video, no Flash plugin, no USB ports. It looks cool! So people get it. And it's from Apple. So people get it. I even want to get one (well, not the first generation, of course). If Apple has taught us anything from the success of the iPod, iPhone, and iPad, it's that features and specifications are secondary concerns to most people.

People want a whole product that functions well at least in very basic tasks. That's the masses.

Yup! Apple focuses on the "user experience". If it causes a potential problem then it isn't included. If they need to lock something down to make things work seamlessly then they pull out the Crazy Glue.

Freedom and flexibility are sacrificed for the "user experience" and obviously they do this successfully.

lykwydchykyn

April 28th, 2010, 08:32 PM

T
Yup! Apple focuses on the "user experience". If it causes a potential problem then it isn't included. If they need to lock something down to make things work seamlessly then they pull out the Crazy Glue.

Freedom and flexibility are sacrificed for the "user experience" and obviously they do this successfully.

I don't see that approach doing well with FOSS.

First, because the vast majority of existing users would reject such an approach.

Second, because there's always a way to make it happen with FOSS. If they throw stumbling blocks in the way, all it will accomplish is to reinforce the idea that it's hard to get simple things configured on Linux.

aysiu

April 28th, 2010, 08:36 PM

All excellent points, but would a fledgling Canonical hardware division be in a significantly better position than zaReason or system76 to price laptops competitively or offer variety? My guess is that these would still be "boutique" computers competing directly against Apple, probably at close-to-Apple prices due to scale, yet offering much more limited ISV/IHV support.

Still, I'd want to buy one, if I could afford it. Significantly better than ZaReason and System76 just by source of funding alone (unfortunately right now it's Mark Shuttleworth's pocketbook and not Canonical's record net profits).

It would be a tough sell for some people against Apple, but if Canonical could implement a slick design and some unique hardware features, it may be able to take some of the ex-Windows users away from Apple just because a significant portion of Windows users are turned off by Mac OS X's interface. Ubuntu may, with Lucid Lynx, be mimicking the button placement and colors of Mac OS X, but it is not really changing the functionality of Gnome. The Gnome desktop environment in terms of function and workflow is still a lot more like Windows than it is like OS X (Alt keys for menus, last closed window closes the application, windows are managed through a window list and not a dock, etc.).

And, as you've said, people who already use Ubuntu would probably be thrilled to get a Canonical-sanctioned laptop, especially ones that real Ubuntu users can try out in a store and get their hands on (even if, at first, it would be Ubuntu users in only one or two major cities).

If Canonical did this (physical store, guaranteed-working Ubuntu computer, unique hardware design), I would definitely start saving up money for it.

mihai.ile

April 28th, 2010, 09:02 PM

I would definitely buy it BUT, it has to be a good laptop. Something like the Dell XPS M1330 series that I have (got it because I knew it was 100% supported directly by dell), but it could have even a better design/quality.

dmengo

April 28th, 2010, 10:04 PM

Apple gives users the absolute best "out-of-the-box" experience in my opinion. Take a Mac out of the box, plug it in, and you'll be up and running with dozens of high quality applications ready to go and displayed prominently on your dock.

Windows 7 on a Dell for example is quite lackluster due to all the third-party junk applications and trial products that are included. Uninstalling this garbage will give you a productive operating system.

Linux on the desktop still trails behind both Mac and Windows on the "out-of-the-box" experience for the user. Too much tweaking and configuring is required. Few non-technical users want to be bothered with these tasks.

lykwydchykyn

April 28th, 2010, 10:13 PM

Linux on the desktop still trails behind both Mac and Windows on the "out-of-the-box" experience for the user. Too much tweaking and configuring is required. Few non-technical users want to be bothered with these tasks.

Which distros have you tried? Not every distro requires the kind of tweaking Ubuntu does.

And are we talking OEM installed Linux? Because otherwise it's an unlike comparison.

gemmakaru

April 28th, 2010, 10:17 PM

I know you are really clever / important / influential people but my own humble opinion is this. I would rather choose my own hardware and choose a distro based on its merits then install it in the way i want. The only preconfigured systems I have ever brought were laptops and they get wiped and reinstalled as soon as possible. Then again I am not one of those normal users you are trying to attract. I have no influence what so ever and I sure won't affect market share. I don't think linux sux and i think it gets better every generation. Cue someone telling me that because market share is increasing right?

dmengo

April 28th, 2010, 10:21 PM

Which distros have you tried? Not every distro requires the kind of tweaking Ubuntu does.

And are we talking OEM installed Linux? Because otherwise it's an unlike comparison.

Well if you're talking "clean installs" then Windows 7 and Mac OS X 10.6 have the upper hand, with Mac OS X being the easiest to install. (Obviously Apple has the advantage because they make both the software and the hardware.)

Ubuntu 9.10 and Linux Mint 8, while probably the most user-friendly Linux distributions are a little more difficult to do clean installs and get everything working as quickly as possible.

Put it this way, if an average, non-technical user ever has to touch the command line for any reason you're sunk. That product is intended for another audience.

lykwydchykyn

April 28th, 2010, 10:40 PM

Well if you're talking "clean installs" then Windows 7 and Mac OS X 10.6 have the upper hand, with Mac OS X being the easiest to install. (Obviously Apple has the advantage because they make both the software and the hardware.)

Ubuntu 9.10 and Linux Mint 8, while probably the most user-friendly Linux distributions are a little more difficult to do clean installs and get everything working as quickly as possible.

Put it this way, if an average, non-technical user ever has to touch the command line for any reason you're sunk. That product is intended for another audience.

I've seen better out-of-the-box than Ubuntu and Mint, Mepis, Mandriva, and Suse (the pay version, not OpenSuse) come to mind. But that's different from long-term friendliness.

I would contend that command line usage in Linux is a lot less required than people like to say. If I bought a Dell Mini 12 with ubuntu preinstalled at the factory, what would I need to use a command line for? Give me an example of something I would need to do on a command line during the course of normal usage on a pre-installed system (note: "usage" does not include fixing some obscure problem)

aysiu

April 28th, 2010, 10:50 PM

My point is that there is essentially no "out-of-the-box experience" for desktop Linux. There should be one if it wants to go anywhere.

aysiu

April 28th, 2010, 10:51 PM

I know you are really clever / important / influential people but my own humble opinion is this. I would rather choose my own hardware and choose a distro based on its merits then install it in the way i want. Who's stopping you?
The only preconfigured systems I have ever brought were laptops and they get wiped and reinstalled as soon as possible. Then again I am not one of those normal users you are trying to attract. Exactly.

Objekt

April 28th, 2010, 11:28 PM

I would contend that command line usage in Linux is a lot less required than people like to say. If I bought a Dell Mini 12 with ubuntu preinstalled at the factory, what would I need to use a command line for? Give me an example of something I would need to do on a command line during the course of normal usage on a pre-installed system (note: "usage" does not include fixing some obscure problem)

Does Dell configure their Ubuntu machines with the full Medibuntu package, or something like it? I've often wondered how vendors of pre-installed Ubuntu machines handle this potentially thorny issue. If you do the full Medibuntu thing, you're using the legally "dirty" libdvdcss package. That could put Dell in a tricky spot.

I think Dell's customers would be disappointed if they couldn't play DVDs on their computers, but I don't know of any legally "pure" way Dell can provide software to do that in Ubuntu. I would think you're pretty much stuck installing Medibuntu or at least libdvdcss yourself. If not the CLI, you have to at least use Synaptic Package Manager.

(Complex workarounds like running a Windows XP machine in Virtualbox don't count.)

aysiu

April 28th, 2010, 11:30 PM

Dell Ubuntu's come with legally licensed proprietary codecs, as far as I know. No need for Medibuntu.

Dragonbite

April 29th, 2010, 03:41 AM

Well if you're talking "clean installs" then Windows 7 and Mac OS X 10.6 have the upper hand, with Mac OS X being the easiest to install.

Windows 7 is much better about it, but Windows XP is horrible. Upon installation I have to go to a DIFFERENT computer to DOWNLOAD the Ethernet drivers! Thne there are the graphics, etc.

Windows 7 was only easier becuase I ran the preview and knew it was missing drivers for my scanner, my chip could not handle XP Mode (virtualization) and my graphics card was not up to snuff enough for Aero. Otherwise I would be stressing out trying to get these to work.

With Linux, though, you run a Live CD and find out then and there not only if things are recognzined but get an idea how well you like the distro too!

Does Dell configure their Ubuntu machines with the full Medibuntu package, or something like it? I've often wondered how vendors of pre-installed Ubuntu machines handle this potentially thorny issue.

I don't know if they use Fluendo or not, but they do use something legal. They would have too much at stake to not do it legally.

lykwydchykyn

April 29th, 2010, 03:48 AM

I don't know if they use Fluendo or not, but they do use something legal. They would have too much at stake to not do it legally.

A quick google search reveals that, at least back around 2007, they were shipping Ubuntu machines with Fluendo. They don't say now, but I am sure they ship with legal codecs of some kind.

This is the kind of thing you can do with a preinstall that you can't do shipping free CD's around the world.

aysiu

April 29th, 2010, 03:50 AM

From the Dell Ubuntu page (http://www.dell.com/content/topics/segtopic.aspx/ubuntu?c=us&cs=19&l=en&s=dhs&~ck=anavml):
After extensive development and testing, Dell PCs featuring Ubuntu are now shipping with version 9.10. In addition to the DVD Playback offered in version 7.10, the newly available version 9.10 offers even more peripheral options such as ATI Video Graphics, Dell Wireless, FingerPrint Readers, HDMI and even better Bluetooth and MP3/WMA/WMV support. And that's the US site, so, yeah, no need for Medibuntu. And, yeah, probably licensed codecs.

magmon

April 29th, 2010, 04:01 AM

Put it this way, if an average, non-technical user ever has to touch the command line for any reason you're sunk. That product is intended for another audience.

I have never touched the command line for setting up ubuntu. Most of the programs you could ever want are in the repositories or on getdeb. The only times I can ever remember touching the command line are when firefox 3.5 came out and when installing caffeine.

dE_logics

April 29th, 2010, 04:01 AM

Problem is the opensource developers are hobbyists, hackers. This is the reason why Linux is a superb piece of software but hard to work on. They are good at implementing very hard things... like new protocols, io schedulers etc... but don't like implementing software for ease of use.

For this purpose we need to 'employ' developers and make them do the job forcefully.

For e.g. let's take on a suggestion of 'Super deb packages' which's like a single msi file for windows... after having on helluva debate with the dpkg developers, they said I need to contact the apt developers for this... since it's basically a collection of deb packages.

Anyway, I will advice apt team also.

Buovjaga

April 30th, 2012, 10:13 AM

Bryan Lunduke gave two back-to-back presentations at Linux Fest Northwest, first a critical one and then a praising one.
See the videos:
http://lunduke.com/?p=2953
Why Linux Sucks (As Usual) (http://www.youtube.com/watch?v=Sh-cnaJoGCw)
Why Linux Does Not Suck (Not Even a Little) (http://www.youtube.com/watch?v=BfLqLK7VdQY)

Many of the same points as in Artem Tashkinov's list of weaknesses:
http://linuxfonts.narod.ru/why.linux.is.not.ready.for.the.desktop.current.htm l

I'm excited about the boom in Linux gaming and what it will bring. I do creative work and that's why I'd like to see The Linux Desktop happen - can't do serious work on a mobile device..
Was nice to see Lunduke deal with the problems with funding open source projects, I've been thinking about the subject intensely for some time. The rise of crowdfunding platforms seems to be improving the situation and, importantly, making open source teams think about organizing and marketing.

Btw. isn't Open Build Service (http://www.open-build-service.org/) a sort of solution for unified packaging? At least in this situation of "religious war".

Version Dependency

April 30th, 2012, 04:07 PM

Yea...seen his previous presentations (he rants about the same things ever year). Dude has been talking about the need for Xorg to go forever. I think we all can agree on that one.

And having a single unified packaging system instead having .deb packages, .rpm packages, Arch packages, etc...yea, we can all on agree on that one but it's NEVER going to happen. Even if you got every distro that exists now to move to one package system, someone will come along and create a new distro with a new packaging system just to be different (better).

Wasted effort and duplication of work is a big problem. Go look at distrowatch sometime and search for lightweight Ubuntu-based distros that use LXDE. There's a bajillion of them. Do they not all realize they are working on the same project?

Ok...end of rant.

zombifier25

May 1st, 2012, 11:51 AM

And having a single unified packaging system instead having .deb packages, .rpm packages, Arch packages, etc...yea, we can all on agree on that one but it's NEVER going to happen. Even if you got every distro that exists now to move to one package system, someone will come along and create a new distro with a new packaging system just to be different (better).

Wasted effort and duplication of work is a big problem. Go look at distrowatch sometime and search for lightweight Ubuntu-based distros that use LXDE. There's a bajillion of them. Do they not all realize they are working on the same project?

Ok...end of rant.
http://imgs.xkcd.com/comics/standards.pnhttp://imgs.xkcd.com/comics/standards.png
A good knowledge of XKCD comics is essential for elite GNU/Linux users.