Linux Distributors Work Towards Desktop Standards

WebHostingGuy wrote to mention an MSNBC article discussing a move by several Linux distributors to standardize on a set of components for desktop versions of the operating system. From the article: "The standard created by the Free Standards Group should make it easier for developers to write applications that will work on Linux versions from different distributors. Linux has a firm foothold as an operating system for servers -- it's popular for hosting Web sites, for instance -- but has only a few percent of the desktop market."

Comments

Yea like they will ever agree with anything

[jellomizer]

After the talk there will be 2 Major Faction. While one may win. The Second one will go Screw you and make their own design in-spite of the the talks. That is the problem with Ego Driven Software vs. Profit driven. While they both have their advantages and disadvantage. Ego Driven Software while the Code my be better quality but have a much harder time agreeing with other people. But Profit driven Software tends to be more consistent but software quality tends to be a little lower.

Re:Yea like they will ever agree with anything

[mugenjou]

and microsoft is a combination of both? they have the low quality of profit driven software as well as egoistic recreation/bastardization of standards just to be as incompatible as possible with the rest of the world.

Re:Yea like they will ever agree with anything

[tsa]

That is exactly what I thought when I read this. This sort of thing has been tried before but so far it never worked for Linux. Let's hope this time people will lower their testosteron levels a bit for the greater cause, because having three major desktop operating systems is way better than the two we have now!

Re:Yea like they will ever agree with anything

[Pneuma ROCKS]

egoistic recreation/bastardization of standards just to be as incompatible as possible with the rest of the world

That's not egoistic. They break several standards to maintain a strong grip on the market, and also because it's sometimes very costly to live up to them word by word, so it's cheaper just to take the easy way. This is all maximization of profit. I think GP is very right.

Re:Yea like they will ever agree with anything

[Jugalator]

Egoistic as a company, yes, not egoistic per individual/grouping, which can be a problem in open source organizations.

Re:Yea like they will ever agree with anything

[asuffield]

You're half right. The bit you got wrong is that the profit motive does not inspire people to produce consistent software. Most commercial software is just inconsistent, with everything around it and sometimes even with itself. This happens because each piece of software has a different project leader, and nobody in management above them understands enough to impose a single vision on the whole system. Given a choice, an individual project team will usually attempt to differentiate their project from all the others, in the hope of getting more money and/or recognition.

So the conclusion is probably that different software created by different people is usually going to be different. That's probably a good thing and you should just get used to it. Nobody can invent a single way to do things that is right for every piece of software you might want to use in the future.

Re:Yea like they will ever agree with anything

[inode_buddha]

I would argue that their ego is directly tied to their market share. Bummer.

Re:Yea like they will ever agree with anything

[Anonymous Coward]

I dont think the same applies to everything, just look at the profit driven US mobile phone network versus the 'ego' driven European networks.

I think this is the closest analogy I can think of

Re:Yea like they will ever agree with anything

[Bloke down the pub]

So the conclusion is probably that different software created by different people is usually going to be different.

But it's more likely to be more different when it's produced by volunteers, simply because a business has more power to compel developers to follow standards or guidelines.

That's probably a good thing and you should just get used to it. Nobody can invent a single way to do things that is right for every piece of software you might want to use in the future.

This is plainly and simply wrong. While no one way may be intrinsically any better than the others, making things different just for the sake of it adds no value at all and smacks of developer arrogance to me. All it does is confuse users - which no doubt feeds that arrogance and sense of being "733t".

For example, it doesn't matter whether an electric plug has flat or round pins, but it's useful that they match the socket in the wall, and that a TV has the same shaped one as a microwave. Sometimes a standard is right simply because it is a standard.

Re:Yea like they will ever agree with anything

[techno-vampire]

This is plainly and simply wrong. While no one way may be intrinsically any better than the others, making things different just for the sake of it adds no value at all and smacks of developer arrogance to me. All it does is confuse users - which no doubt feeds that arrogance and sense of being "733t".

And that, ladies, gentlemen, geeks and geekettes is the great value of the Microsoft "look and feel." All Winderz programs have their menus in the same place, in the same order, and the keyboard shortcuts are all the same. If you know how to use Winderz at all, you can probably find your way around any Winderz program well enough to get started using it.

Say what you want about the Evil Empire of Redomond, they do Get Things Right on occasion, and this was one of them.

Re:Yea like they will ever agree with anything

[schuster]

You also start to hint at the other major problem. We'll assume for a second that they do agree on a standard. Now you're going to have to have the standards body approve every piece of software to make sure that it's compliant. Ease of use is a philosophy. Creating this kind of standard requires that the developers to want to be consistant. If this does actually happen, the freedom of toolkit choice that the open-source community values so much will almost completely cease to exist. If the developers/users want a new toolkit, they'll just have to wait until the standards body approves it. Now there's the issue of how well applications integrate with each other. There's just too much to deal with and I really don't think the community can make it happen.

reasons why

[fl!ptop]

interesting that msn bills move as 'making the operating system compete better with windows' instead of 'making it easier for developers to write applications that work on different flavors.'

i would think the former is a result of the latter, instead of the other way around.

I don't know what they are on about

[MichaelSmith]

I can run KDE applications under fvwm and Gnome, as long as the runtime libraries are there. I don't see why it is hard to have QT and GTK libraries on each system.

The only remaining issue is cut and paste with rich content but the article doesn't talk about that.

Re:I don't know what they are on about

[jellomizer]

It is an issue of consistency. If I am running Gnome I know when I am running a KDE app because it looks a feels slightly off. The same if I am using a straight X11 App. Linux for the desktop is not about Window Managers. It is about giving Developers tools to make their Apps Desktop Friendly, And ability to make sure Linux Apps look good no matter what WM you are using.

Re:I don't know what they are on about

[Sweetshark]

I don't see why it is hard to have QT and GTK libraries on each system.
Because:
- its ugly design
- it involves lots of code duplication
- it sucks on lean platforms (for example Maemo)
- it doubles your chances of being hit by a security flaw
- it produces a lot of unmaintained basic infrastucture code (like VFS) where the implementation is the spec.
- standards are a Good Thing

Re:I don't know what they are on about

[moro_666]

i wouldn't start my kde if it would look and behave like gnome ;)

  now about the `issue` itself, redhat is dragging along a bunch of people to push some kind of one-standard-for-all (cough-cough-bs-cough-cough-profite-cough). they want to unify some things (the article didn't really elaborate what ...), and therefor make all the stuff more the same.

  i don't know about you, but if i'd want everything to look the same, i could aswell choose osx or winblows (nah, not really win, it's not ...). i chose linux many years ago because i wanted it to behave like i want it to and to look like i want it to. i don't want my desktop to look'n'feel like it suits a redhat salesman.

  i understand that this will help to push linux into the streets blabla, but is this really what we all want ? or is this the beginning of the end of linux as we know it ?

  we have seen many items out on the `market` that were supposed to unify and standardize linux (various package managers etc.), none of them have succeeded. for the broadband, i hope it succeeds, for my own's sake, i hope they fail terribly and give it up.

Re:I don't know what they are on about

[Homology]

I don't see why it is hard to have QT and GTK libraries on each system.

Because: - its ugly design

It's not an issue of bad design, but of neccessity. As long as different applications are using different libraries, you end up installing those libraries.

- it involves lots of code duplication

There is no code duplication involved, however, there is some overlap of functionality

- it sucks on lean platforms (for example Maemo)

If your computer is very limited, then you don't want to run either KDE or GNOME.

- it doubles your chances of being hit by a security flaw

Care to elaborate on that?

- it produces a lot of unmaintained basic infrastucture code (like VFS) where the implementation is the spec.

Just because some random Linux distro is offering both KDE and GNOME does not imply that KDE or GNOME stops maintaining their own code.

- standards are a Good Thing

Yeah, yeah, sure. Except when the "standards" sucks.

Re:I don't know what they are on about

[IamTheRealMike]

The standards in question are things like "how do I install a menu item in a way that works across distributions" and "how do I distribute C++ apps in ways that don't randomly crash" and "what libraries can I expect a Linux system to have".

The whole "as long as the runtime libraries are there" catch is what it's all about. It's not reasonable to expect people to deal with dependencies.

Re:I don't know what they are on about

[Homology]

The whole "as long as the runtime libraries are there" catch is what it's all about. It's not reasonable to expect people to deal with dependencies.

It's the package maintainers job to deal with library dependencies. I a Linux distro is unwilling to do this, why should I use it in the first place since it is obiously of low quality?

Re:I don't know what they are on about

[DrXym]

I guess the issue isn't what widget set you use, but the fact that you have two stacks underpinning both of them with little sharing of libraries or settings. When KDE & GNOME learn to play nice and share the same settings, desktop files, theme engine, and behaviour is the day that Linux move on from pointless widget wars (that no end user cares anything about) and focus on delivering a high quality desktop.

Re:I don't know what they are on about

[Kangburra]

i wanted it to behave like i want it to

I agree with this, but also when I choose that way I don't want some wizard putting it back or an update changing the setting I explicitly chose.

Linux is not perfect but it's not the worst option.

Re:I don't know what they are on about

[wysiwia]

Or say it with a single statement:

"It's the first top inhibitor of the Linux desktop adoption"

See http://www.osdl.org/dtl/DTL_Survey_Report_Nov2005. pdf

and http://linux.slashdot.org/comments.pl?sid=183801&c id=15179906

O. Wyss

Re:I don't know what they are on about

[IamTheRealMike]

World of Warcraft is not in most distros, is it low quality? Compared to say, bzFlag?

Don't get me wrong, I love a good game of bzFlag, but saying "If a Linux distro is unwilling to do this, why should I use it in the first place since it is obiously of low quality?" isn't going to fly with most people ...

Re:I don't know what they are on about

[AusIV]

Different distros may have different ideas of how to handle dependencies package distribution and dependencies. I'm a relative noob to Linux, and I already know that debian uses one kind of package, gentoo another, fedora another, and there are several other methods of distributing. I think part of the problem is that each of these distribution techniques has its own advantages. You're not going to get people who use Gentoo to download and use binaries (at least not for everything), and you're not going to get people who use Ubuntu to compile everything they want to install.

Not wanting to meet someone else's standard may lead to a lack of ability to maintain dependencies, but that doesn't mean the method of distribution is necessarily low quality.

Re:I don't know what they are on about

[Spit]

Apple doesn't have this, some apps are the white look, some are the brushed metal. Windows apps often look quite different. Why should X developers do any different? The app works doesn't it?

Re:I don't know what they are on about

[jZnat]

Well, it would lighten the load on tracking library dependencies, that's for sure. It seems that one of the easy ways of figuring it out is trying to compile the package in question in a chrooted environment and resolving missing libraries as you find them, but that's also a shit shoot if you're looking for its optional libraries without futzing around with the source code.

Re:I don't know what they are on about

[jbengt]

No, the need for standardization is not for the user's benefit, it's for the developer's convenience in not having to tweak and compile dozens of versions.

Re:I don't know what they are on about

[NutscrapeSucks]

Well said. The fact that the Linux end-user even knows what "toolkit" they are running in a given program is geek-marketing gone insane. Talk about a totally irrelevant choice-point foised on the poor users.

A typical Windows install has dozens of toolkits in use, and the user is none the wiser. This is mainly because most common settings are shared in the registry, providing a rough consistancy -- not because the toolkits is are so awesome.

Re:I don't know what they are on about

[munpfazy]

I think part of the problem is that each of these distribution techniques has its own advantages.

Yup. Very well said.

One of the things I like best about my distro of choice (slackware) is that its package manager doesn't try to keep track track of dependencies. In my book, that's a very good thing. I've never met a package manager that can deal with dependencies reliably, and have wasted days trying to repair the screwups that happen when package managers get it wrong.

But, that's not to say people who prefer dependency tracking package managers are stupid or that their systems are "low quality." They simply want something different from a distro than me. Lucky for both of us, it's easy enough to accommodate everyone.

Re:I don't know what they are on about

[Anonymous Coward]

>I can run KDE applications under fvwm and Gnome, as long as the runtime libraries are there.
>I don't see why it is hard to have QT and GTK libraries on each system.
>The only remaining issue is cut and paste with rich content but the article doesn't talk about that.

Not really. After that, you need to address the issue that your end-user never have to learn what "KDE", "fvwm", "Gnome", "runtime libraries", "QT" and "GTK" are.

While I agree that the Linux community really does need to agree on ONE desktop, ONE way of doing thing, etc (like Apple and Microsoft), they need to wake up and realise that people don't give a damn on HOW their computer works, they just want it to, you know, WORK. Just like I don't really know all the details on how my car engine and transmission works, but I can drive it just fine, like 99.99999999999% of the drivers out there. I'm not asking them to stop making specialized distros, but the desktop version has to be constant.

And now, let's watch the open-source elitists mod me down as "troll" because they don't quite grasp what I wrote above.

Linux may be free, but even free stuff is useless if nobody uses it.

Re:I don't know what they are on about

[kimvette]

By that token, Windows is a bad solution because you have applications which directly access the Windows API (analagous to X11 controls), MFC, .Net, and various other libraries - and then you have third-party libraries like Stingray.

If you find that having to have lots of libraries installed to support various GUI elements is a bad thing, you might want to check out Amiga, BeOS, or just plain old DOS - because even OS X is not going to give you what you want.

This problem is likely to never go away - if, for example, you didn't have a tree control which allowed for drag & drop and you really, really needed one, you'd likely look for a third-party control which is prebuilt, or develop your own custom control - and now you've just added a new library. Need an outlook-like bar which supports Drag & Drop? Now you've just run into the same issue - another library or extending an existing one which does "almost" what you want.

You could of course stick with what you get out of the box, but good luck building a modern-looking application that doesn't bury functionality under three levels of submenus and obscure shortcut keys.

Re:I don't know what they are on about

[Anonymous Coward]

...make sure Linux Apps look good no matter what...

Define "look good". Right.

And that is exactly why monotheists are so off base. When they shout "we need standards", they really mean "do it my way".

Here's a little history lesson. Dictators are bad. Freedom is good. You'd think it would sink in after a while...

Re:I don't know what they are on about

[AnyoneEB]

look good: follow whatever desktop look&feel theme the user has set.

Re:I don't know what they are on about

[Sweetshark]

It's not an issue of bad design, but of neccessity.
nonsense. Is it necessary to have two or more implementations of a VFS? No. Is it necessary to have two or more message bus systems? No.

If your computer is very limited, then you don't want to run either KDE or GNOME.
Of cause you will want that - portability is some damn fine thing. The PDA and the moblie phone are merging and are beginning to become really useful. Even Microsoft realized that and is pushing .Net for that cause like there's no tomorrow.

There is no code duplication involved, however, there is some overlap of functionality
Both are involved, which is even worse.

Care to elaborate on that?
Less code, less vulnerabilities without loss of features.

Just because some random Linux distro is offering both KDE and GNOME does not imply that KDE or GNOME stops maintaining their own code.
When was the last major release of Gnome-Vfs? Even though the Gnome-Wiki says it really needs a basic rework? Multiple projects are a Good Thing in the beginning, when you still need to find out which ideas are the Good Ones. After that, its better to settle on a standard, and to build implementations that are compatible and exchangeable.

Yeah, yeah, sure. Except when the "standards" sucks.
There is not much space for creativity in the basic infrastructure needed for a desktop enviroment. This is the reason for projects as dbus, HAL, poppler and Desktop-VFS.

Re:I don't know what they are on about

[shutdown -p now]

i don't know about you, but if i'd want everything to look the same, i could aswell choose osx or winblows (nah, not really win, it's not ...). i chose linux many years ago because i wanted it to behave like i want it to and to look like i want it to.

Precisely. But wouldn't it be wonderful if you only had one look&feel config to tweak for all DEs and widget toolkits, rather than a separate one for each that we have now? Uniformity does not mean lack of freedom of choice. And default KDE settings can still be very different from default Gnome settings - but when I change them, I want the changes to apply to all GUI software on by box.

Re:I don't know what they are on about

[Anonymous Coward]

i wouldn't start my kde if it would look and behave like gnome ;)

Seriously, this is the biggest problem with desktop Linux these days. There are a few programs that have no real competition, are gtk-exclusive, and have gotten seriously mucked up by the ego tripping apple wannabes running gnome. Hopefully Konq will flesh out with KDE4 to be a real alternative to Firefox and someone will fork Krita (or something else, ANYTHING!) to compete with GIMP.

Particularly GIMP is a problem, since it isn't just the file dialogs anymore. It seems like the GIMP developers are more interested in fucking over the interface completely than in making any moves to fix the places where it really needs improvement. At the point I'm just rebooting to windows again to use my old copy of Photoshop since it drives me too fucking nuts to put up with anymore. Which really makes me sad, since I love the superiority of an X desktop, and KDE specifically.

And for all their talk about "ease of use" I have yet to find someone who hasn't been using Linux for a while who can even figure out wtf is going on when gnome crap pops up when they use my computer. I don't see why GIMP should integrate fine with windows (I've seen several people have no problem using it in windows) but is such a fucking nightmare on Linux.

I propose a new rule, no one who puts windows, mac, or anything else over Linux should have any fucking say in the development of Linux applications. Especially not ones like the GIMP where there are plenty of alternatives on other systems and NONE for us.

I have to ask...

[MobileTatsu-NJG]

This question is going to seem rude, and I apologize for this, but why didn't this happen years ago? I'm asking out of curiosity, not as a jab at the community. It seems to me that this sort of standard would have been quite valuable as soon as GUIs became prevalent with Linux.

Re:I have to ask...

[wysiwia]

Simply because nobody realized the real reasons until OSDL published its survey in Dec 2005 (http://www.osdl.org/dtl/DTL_Survey_Report_Nov2005 .pdf). Even today nobody wants to take the necessary steps as outlined here http://linux.slashdot.org/comments.pl?sid=183801&c id=15179906.

O. Wyss

Re:I have to ask...

[Pneuma ROCKS]

I think GUI, despite being prevalent for quite some time, have been very, very low in the priority list of Linux developers. The community has focused more on the low-level, kernel and architecture areas, and the rest has suffered from it. IMO, GUIs in Linux have always been an afterthought, and that's the reason they suck so much (again, IMO).

This sheds light in a key problem with open-source software: developers will work in what they want to work, not necessaily in what needs to be done.

Yeah, mod me down, see if I care.

Re:I have to ask...

[metamatic]

Once upon a time there was one major desktop environment for Linux, KDE.

Some people didn't like the licensing of the Qt toolkit KDE was built on, so they started work on GNOME, which was to be the truly free alternative.

The licensing of the Qt toolkit was changed. However, out of ego the GNOME developers continued, and so we ended up with two major desktop environments.

Re:I have to ask...

[Rich0]

There are a few reasons why LSB compliance and the like isn't considered critical by many devs:

1. Most distros use package managers - which let you put all of an applications files wherever you want, and thus the application doesn't need to know where the right place to install itself is.

2. Most distros compile their own binaries from source. So, if the original developer linked his binary version against glibc-2.0, and RedHat has v2.3, they'll just recompile it themselves, and distribute their own version.

It works just fine 99% of the time. Really, the only people who need to have all kinds of standards that are locked down are folks who distribute closed-source apps - since the distros can't help them out by rebuilding their application in 47 different ways. Then it matters what versions of various libraries are installed, and where everything can be found (no automake/configure/etc directly on the target platform).

As a result, many volunteer distro developers find it hard to care - why should they donate a ton of their time just to increase the incentive for companies to not release their source code? They do linux for fun/satisfaction, and it doesn't really matter to them whether name-your-favorite-software-vendor provides closed-source apps for their distro - they'll just write free competing versions of the same.

Re:I have to ask...

[Anonymous Coward]

It still depends on your opinion of the current Qt license.

I don't particularly like being given the choice of releasing free software or paying money to one organisation (why should it just be Trolltech that gets paid for use of their library, what about the developers of X11, or the kernel, or the many other libraries that are in use by the program?).

While I like free software and tend to use it over closed software when given the choice, I strongly object to forcing a license on a developer for dynamically linking to a library. The developer should be able to choose what license they want in those cases. And Trolltech of course got to do that, building on top of X11/whatever else that allowed them to choose their own license, and I fully support their choice to do as they will. I just don't support using a library with such a license as a 'standard' component of a desktop.

Re:I have to ask...

[metamatic]

Well, sure, if you disagree with the GPL, you're not going to like the current Qt/KDE licensing. But if you disagree with the GPL, why the hell are you using Linux anyway?

Re:I have to ask...

[Anonymous Coward]

Why the hell do you feel the need to apologize? Your question/criticism is totally valid. This AC was finally fed up with half-assed, almost-but-but-not-quite working, GIMP-It's-different-but-better, EMACS-makes-better-programmers, QT-is-not-GPLd desktop operation. Linux is a fine server OS but the desktop sucks donkey balls. OSX is what Linux could/should have been years ago. Got an apple and wont come back.

Re:I have to ask...

[Kjella]

Well, if I get to pick my preference it goes like this:

1. Make it work
2. Make it work well
3. Make it work fast
4. Make it work fancy

Besides, I don't think it's wrong to say that for many Linux has been and is only a server OS and they couldn't really care about a GUI. But somehow I get the impression that those most deeply into inner mechanics of kernels and drivers might not be the most qualified to make good GUIs either, so I'm not sure it's suffered that much. And even so, GUIs are really subjective and hard to make an authoritative source for, my impression is that OSS has focused more on flexibility than design. For example, look into some of the "photoshopifier" plugins of GIMP. They basicly redo the entire UI. So go ahead, come with your suggestion for a good UI - my bet is it'll be quite different from someone else's idea of the same.

Re:I have to ask...

[Anonymous Coward]

I don't disagree with the GPL for use in applications. I quite like the GPL, but I much prefer the LGPL. In regards to why I use Linux - it has better hardware and application support than FreeBSD. And for the most part, the more popular libaries are LGPL, BSD, or similarly licensed.

Re:I have to ask...

[Hydroksyde]

You think the Gnome developers continued "out of ego"? how about the fact that

• abandoning the gnome project would mean discarding several years of hard work, for no good readon
• Some people, for one reason or another, prefer Gnome over KDE

Re:I have to ask...

[Anonymous Coward]

I have seen the arguments go back and forth on the gnome vs kde thing for many years in the end the ultra simplified breakdown is gnome is free as in freedom(develop what you want sell it if you can.) and kde is free as in demo-ware(give it a try, if you like it and want to develop commercial apps... well it's gunna cost ya!) This also answers the question as to why kde is chaotic and inconsistent. Just look at the flat 80's era line art images all throughout the latest kde release. I opened Koffice and had a flashback to windows 3.0 who the heck is the supposed artist responsible for that crap? Same with the default background artwork in konqueror, kontact, etc... yuck! Don't get me wrong I love the applications that the free "qt" community produces and I use them all the time. I just happen to think the K in the Desktop Environment stands for krap.

Re:I have to ask...

[Pneuma ROCKS]

I'd say that the point where we differ the most is that, for me, 4 would be distributed among 1 and 2. Agreed, if you're building server software it is clear that the GUI will probably not even be installed, but I think TFA focuses more on the desktop market. If I'm doing an app, I do it right, and that means that if it has a GUI, it better be a good one (consistent, easy, intuitive, clean, etc).

I use the Gimp, and I hate its interface. You eventually get used to it, but it makes the learning curve so much harder. I'm not some GUI guru, but I know that something that can never be missing is consistency. I can go to any (modern) Windows box in the world and I will know where to find Start, Control Panel, etc. Even in a new and unknown Microsoft application I will know how to use menus, how to configure things, what alerts I can expect from certain actions, where buttons are located, and so on. It has been my experience using Linux GUIs that even the most trivial alert window requires me to inspect it to realize what happened and to find the button or option I want to click. Most users will give up right away after something like this, and that is unacceptable.

Re:I have to ask...

[metamatic]

abandoning the gnome project would mean discarding several years of hard work, for no good readon

Date GNOME project started: August 1997.

Date Qt licensing changed to open source: November 1998.

First GNOME release: March 1999.

Date Qt licensing changed to GPL: September 2000.

But anyhow, they could have just put their efforts behind Harmony. But no, they had to reinvent the wheel.

This is nice, but...

[PenguinBoyDave]

Dell, HP, Toshiba, etc. etc. STILL package Windows with every new PC that leaves the shop. I have seen no indication that they plan on changing that any time soon. Sure...Dell might say he likes Ubuntu, but I'll believe it when the first Dell ships with Ubuntu and a Ubuntu sticker on the front where the Windows sticker used to be...you know "This PC specifically designed for Ubuntu Linux."

I don't know too many people that are going to go out and buy a while-box PC (other than geeks) and load Linux, when for about $400.00 you can get a fully rigged-out Dell with Windows and all the goodies.

Re:This is nice, but...

[Anonymous Coward]

"a fully rigged-out Dell with Windows and all the goodies."

With all the oem installed crapware that slows the system to a crawl and with the value add products such as demo software to keep the adware/spyware/viruses/worms/malware at bay. Indeed, who would turn down such an offer..?

"One big things that's difficult is consistency"

[wysiwia]

Yes and consistency can only be achieve by standardizing. Unfortunately this doesn't only hold true for the desktop, it's equally or even more important for the applications. So far Jim Zemlin, executive director of the Free Standards Group, doesn't seem to realize this else the FSG would have already standardized on a single set of application guidelines as outlined in wyoGuide (http://wyoguide.sf.net/). Since this isn't the case so far we still have to wait for the breakthrough of the Linux desktop.

If anybody is interested in a Linux desktop and don't want to wait much longer, he should persuade the FSG to come to terms and at least delve and evaluate wyoGuide.

See also http://lxer.com/module/newswire/view/54009/index.h tml

O. Wyss

A key step to the $150 laptop

[rqqrtnb]

Once Linux distributors get their act together, it won't be long before semi-disposable laptop computers will become available. The Nick Negroponte hand-cranked, third-world computer will spawn a commercial version.

Linux as the OS, Open Office, Mozilla, a few other key apps, and with no "Microsoft Tax", and no headache in installing Linux on a used Win-Tel machine, plus a few "styling options," these machines will sell like hotcakes!

Then, of course, the virus writers will shift to more fertle grounds, and all the bad that goes with the good...

Re:A key step to the $150 laptop

[jZnat]

And then the virus writers realise it is so much harder to even write a user-level virus that can do anything in a Unix environment, so they say "fuck it" and move back to pwning Windows boxen.

Finally!

[yootje]

What Linux needs is standardization. Having 921034 options to choose from is sometimes a good thing, but sometimes you have the feeling: why don't they just work all on 1 fantastic piece of software?

Re:Finally!

[Coryoth]

Having 921034 options to choose from is sometimes a good thing, but sometimes you have the feeling: why don't they just work all on 1 fantastic piece of software?

Because the worlds open source developers are not a giant slave pool designed to do your bidding.

Open source will always be chaotic and involve a great deal of duplication because that's that nature of the beast. The gain you get from that cost is much more open software that's developed rapidly and tends to work as a free market for ideas: the better ideas eventually win out (though that may take some time). If you want something different then you want Apple or Micrsoft with their rigid top down control structure which ensures that everyone is working toward a single unified goal (as much as is possible), and all the work is directed. The upside is consistency and a unified vision, but the downside is that the whole thing is more locked up, an often slower development cycle, and a tendency to get hit with the same stupid mistakes release after release after release just because it appeals to the guy at the top.

It's a choice and you can pick the software ecology that suits your needs. Just don't go expecting one to behave like the other on your whim - there are deep fundamental philosophical divisions about how to develop software (to let it evolve from the bottom up, or direct it from the top down) that are largely irreconcilable.

Jedidiah.

Re:Finally!

[tomstdenis]

First off Linux is a kernel.

As for the distros, yes there is redundancy. It's annoying. I tried to tell Redhat and SUSE to merge but they refused. For the most part outwardly they're all the same. You get some un-optimized heavily modified Kernel that you can't trace back to the vanilla and a plethora of pre-built tools with whacky --enable-* flags set. It's annoying and highly unproductive.

As for the options, keep in mind unlike [say] Windows a Linux based distro can target a variety of actual real world "work scenarios". This is why there are many projects out there that can "confuse" the landscape.

As for being complicated to choose between... use Gentoo. It handles 99% of all dependency problems while letting you use the latest and greatest built with the options you want enabled. How many packages do I have installed? Around 400 to 500. Can I name half of them? Not even close. Can I easily add a new package or remove an old one? Yes.

And then to those who claim bloat ... I'm using ~3.1GB of space for what I consider a fairly well equipped workstation (many tools such as GNU CC chain, GDB, various mem checkers, tetex, X11, Gnome, openoffice, etc).

I can get a basic workstation (with devel tools, X11, openoffice, etc) in around 2GB which is much better compared to Windows which on its own is 2GB. Then you have to install 6GB of MSVC, another GB of Office, another GB of Miktek [to get real work done], etc ,etc, etc. A complete Windows workstation takes ~15GB of disk or so.

Part of the reason why Linux distros [specially Gentoo] can be so small is the use of shared libs. Which is odd because Windows is largely based on DLLs.

Tom

Re:Finally!

[JanneM]

why don't they just work all on 1 fantastic piece of software?

Because there is no one answer to what makes a piece of software fantastic.

When intelligent people can reasonably disagree on it, don't be surprised - or dismayed - when the end result is several divergent designs. That is truly a case where any one of the designs are good, and importantly, better than a compromise between them.

Re:Finally!

[TERdON]

why don't they just work all on 1 fantastic piece of software?

Because there couldn't be such a thing - it's an oxymoron.

Basically, the requirements of the piece of software would be heavily contradictory - dead-easy to use, but still incredibly powerful. Few such programs exist - because they are virtually impossible to make.

Example: file managers. On the one hand, you have explorer, finder, nautilus et al, which all are at least relatively easy to use even for a newbie. Many find them far to little powerful, especially on /., where the favourite probably is raw /bin/bash, which is far more powerful, but also really hard to learn.

The same principle holds for most other software. Either you make an easily usable, or a powerful version. The powerful version will, by definition, need a lot of learning on the part of the users, and thus can't be easily usable.

When you try to unite these two conflicting requirements, the most likely outcome is one of:

1) Cluttered interface, which intimidates the newcomer
2) Clean interface, but with all powerful features hidden away from sight so the advanced user has to look for them.
3) Millions of settings in an unmanagable settings dialog, toggling the different features on and off.

Conclusion: One software normally can't be the great software - not for every single user. The shifting requirements different individuals have will without doubt make them prefer different software - and that isn't really a bad thing. If everybody ran the same software, there wouldn't be as much incitement for developing new, powerful features!

Re:Finally!

[wysiwia]

Open source will always be chaotic ...

There's nothing against chaotic but does it also have to be bad? I'm not against anybody doing anything in OpenSource but when it comes to standards there's no use for bad. So far the FSG mostly created sound standards but unfortunately that can't be said of the Freedesktop.org and even worse of the desktop architects of OSDL. Especially the Portland initiative of the OSDL which was just created for this task and which should know better do not even try to tackle the first top inhibitor of a desktop Linux adoption (http://www.osdl.org/dtl/DTL_Survey_Report_Nov2005 .pdf), not even after I've told them. So my hope is the FSG will take over this problem and do a better job.

O. Wyss

Re:Finally!

[Cereal Box]

It kills me how on the one hand you guys got absolutely nuts about web standards, document standards, etc. -- "just code to the standard and it'll magically work!" is the mantra around here. But as soon someone says that the Linux desktop or Linux distributions need to standardize on this or that the tired old "but that would stifle choice!" line gets trotted out.

Hmm, maybe that's what Microsoft thinks. They break standards because standards would just limit their choice...

Re:Finally!

[Anonymous Coward]

"...not even after I've told them."
--------------

Oh, yeah, Otto -- you're surely the greatest, we know. (Just not as convincing as you should).

Re:Finally!

[strider44]

It's simple: Standards for communication and interoperatability = good. Standards for implimentation = bad.

For example HTTP is a communication standard which is good. HTML is a communications standard which is good. However having one master implimentation of the HTML standard (say if we all decided to standardise our web browsers on IE) is bad - the implimentation might for example be insecure or corrupt, or it might just work badly. You might have one person who wants to display the HTML in all it's bright <blink> glory, and another person who wants just the plain text. Choise of implimentation is good, but having no standards for interoperatability is bad. Microsoft seems to want to standardise on an implimentation but not on communication.

Re:Finally!

[techitch79]

"Basically, the requirements of the piece of software would be heavily contradictory - dead-easy to use, but still incredibly powerful. Few such programs exist - because they are virtually impossible to make."

i know i'll be modded a troll now, but i just can't resist:
- did you ever see a Macintosh? or a Garmin GPS unit? or a TiVo? or an iPod? or a BMW iDrive?

writing dead simple yet powerful software is no wizardry. you just have to concentrate on functions and workflows which are used by the customers in 90% of the time, and hide the power features elsewhere (keyword: RTM). if you present all the functions at once, you'll end up with a chaotic GUI. keeping the focus on the main functions will allow the people to use your product in an intuitive way, with an extremely short learning curve.

Re:Finally!

[Coryoth]

It kills me how on the one hand you guys got absolutely nuts about web standards, document standards, etc. -- "just code to the standard and it'll magically work!" is the mantra around here. But as soon someone says that the Linux desktop or Linux distributions need to standardize on this or that the tired old "but that would stifle choice!" line gets trotted out.

Oh I'm all for standards for Linux, it's just that, given the development model, its going to have to be standards for protocols, communication, etc. and not standards for implementation. Have a look at Freedesktop.org standards, or LSB, or Portland from OSDL, ar even things like Tango icon themes - they all make sense because they sketch how interaction should work, without prescribing too many details of how to implement it. What you are asking for is standards in the sense of "there shall be only one toolkit used, and all implementation must follow these strict rules". If we were to apply that sort of standardisation ot the internet then rather than having standards for HTML and CSS and leavinbg it up to the browser makers to decide how to implement them we'd have one mandated browser and all other browser development would be barred.

Standards are nice. Let's use them - where they make sense. If you want truly rigid standards then there are perfectly good options for you: OS X for example. Trying to enforce the same top down control over the open source development world is like herding cats - it just won't work. So if you don't like OS X then take your lumps because OS X is what Jobs has mandated it to be and if you don't like that tough. If you do like OS X ... then why are you complaining, you already have what you want?

Jedidiah.

Re:Finally!

[Dis*abstraction]

Exactly. It's perfectly possible to make a UI powerful while at the same time making it easy to use. Grandparent's excuses for lazy UI design are what lead to travesties like Azureus.

What many OSS developers miss is that "ease of use" doesn't only help some hypothetical Joe Average, it helps themselves as well. Even developers use email and browse the web. And well-designed development tools save time and stress, freeing up energy to focus on... development.

Re:Finally!

[munpfazy]

Especially the Portland initiative of the OSDL which was just created for this task and which should know better do not even try to tackle the first top inhibitor of a desktop Linux adoption (http://www.osdl.org/dtl/DTL_Survey_Report_Nov2005 .pdf [osdl.org]), not even after I've told them.

Okay, you've posted this link three times, and its still not at all clear to me what you're talking about.

In the report to which you link, the first on the list of "top inhibitors of Linux desktop adoption" is "Application support" and discusses the need for several specific proprietary software packages such as Photoshop and AutoCad.

How, exactly, do you expect a standards body to fix this problem?

Re:Finally!

[Anonymous Coward]

Hold on a bit. "why don't they just work all on 1 fantastic piece of software?" is not "why don't they do my bidding?". He questioned their focus and why they don't do it a particular way.. he wasn't making demands. Your post came off as snobbish, which is somewhat typical of the Linux community sadly. I won't assume you meant it that way, though.

You have an excellent point about open source and consistency. At the risk of stating the obvious, Linux in general won't ever feel as focused as a closed-source company's product. Linux has always been about functionality over looks, but damn. Some of the new screenshots of Gnome are looking really hot.

Re:Finally!

[TERdON]

did you ever see a Macintosh

I wrote my post on my iBook, does that suffice?

And yes, the pattern holds even for Macs. Why do you think the pros don't use iMovie, but FCP? Why was one of the first things I did to put the terminal in the dock? Apple does a really good attempt at cracking the gordian knot, but ultimately, you have to handle the difference in demands on complexity somehow.

I'm not saying that's always bad, just that it means that a software just not can aim to be best for "everyone" - if you choose a specific target group, and make your software ultimate for them it will somehow be better (for that group at least). Ideally, every user should have its own system integrator/programmer that tailored for their specific needs, but that's impossible due to lack of programmers (hey, someone would have to make custom solutions for them as well) and need of interoperability.

Re:Finally!

[Anonymous Coward]

??? I find the interface of Azureus considering the amount of power it provides. Maybe you haven't used a recent version?

Re:Finally!

[flosofl]

How the fuck is this a "Troll"?

The only thing I can think of is you must have pissed off some thin-skinned, acne-ridden basement dweller and he (she? - odds are against it) is actively targeting you.

Re:Finally!

[wysiwia]

Okay, you've posted this link three times, and its still not at all clear to me what you're talking about.

Isn't it amazing that Linux fans don't grasp this? It seems impossible for them that even the majority of the Linux users still wish for Windows-only application (better said none-free applications). Sure enough none-Linux users all don't wish for none-free applications. The sole exception to this rule is so far Mozilla/Firefox. Doesn't this fact give you something to think about?

Unfortunately while the fact, thanks to the OSDL survey and thanks to other hints listed here (http://lxer.com/module/newswire/view/54009/index. html) is now proved there aren't many hints about why is it so. I've already suggested to OSDL to do another survey which hopefully lists the reasons but they failed to do anything so far.

Since I'm working on this subject since about 2000 I'm quite sure I've found a possible solution albeit I agree I can't prove it myself. But any sensible person understands that it's related to the look&feel of a Linux desktop and that definitely mostly means the look&feel of the applications. Just think why any other free application than Mozilla/Firefox hasn't gotten a bigger market share even if they cost less than the commercial applications?

How, exactly, do you expect a standards body to fix this problem?

Isn't that obvious? Assumed you don't believe me then the first task is to verify my reasoning, start a discussion about why and maybe do another survey. I'm sure the result will be just what I said but it seems you all need more reason.

If the first step proves I'm right then evaluate if wyoGuide is the way to go or if there are better ways. Maybe try to improve wyoGuide if it doesn't fit in all points. I again don't fear the outcome of this step.

Another if, if step two proves right, then these guidelines should be established and published as official Linux application guidelines. So far this won't have much implication except than to state the all the current used guideline failed and should be scraped.

So far the above three steps haven't solve the application problem but they set a corner stone where the solving may begin. On the other side these three steps don't cost anything except some time. Sure enough the real work starts now but without this official announcement this work won't start.

O. Wyss

Re:Finally!

[Anonymous Coward]

It's simple: Standards for communication and interoperatability = good. Standards for implimentation = bad.

There's no need for an "implimentation" [sic] standard. A desktop standard can specify what functionality the desktop must provide in a specific and detailed manner without requiring any particular implementation.

Re:Finally!

[wysiwia]

For anybody reading this it's obvious what an idiot you have made yourself, posting such a statement as Anonymous Coward but since I don't know why, you might simply not have understand.

No I'm not the greatest and maybe not even the greatest application guideline guru but I'm the first one and so far the only one who realized that without these guidelines, there's no hope for the Linux desktop. I can assure you the year of the Linux desktop will be announced year after year without these guidelines.

Think why I do this and get tried to be insulted as you did. You're not the first and probably won't be the last. So better think first the next time.

O. Wyss

Re:Finally!

[shutdown -p now]

Example: file managers. On the one hand, you have explorer, finder, nautilus et al, which all are at least relatively easy to use even for a newbie. Many find them far to little powerful, especially on /., where the favourite probably is raw /bin/bash, which is far more powerful, but also really hard to learn.

You are right. But why do we need several Explorer clones in Linux? Do they differ enough from each other to warrant two completely different codebases?

Re:Finally!

[Slithe]

Maybe it is too expensive for him?

On the desktop and haven't looked back...

[i_want_you_to_throw_]

I have tried using Linux on the desktop MANY MANY times and always found myself stymied by getting printers to work and so forth. I have always been adamanat about using it for servers where it's very much worth the time to figure out Linux to have the benefits of it as a server product (bulletproof security, etc).
As a desktop product though I wasn't about to spend all day dicking around with trying to get it to work. That's was then.... this is now...

I have been using Linux as a desktop for several months now and it has flawlessly detected all my perpherals, and I Have now been able to spend more time doing development which is what I get paid to do.

Linux is getting better in this area and Linux is going to start making inroads. Slowly but surely...

Re:On the desktop and haven't looked back...

[tomstdenis]

Again this is an example of faulty logic.

You're saying because your printer manufacturer hasn't followed the 25 year old PS standard that Linux is broken? Why not buy "Linux ready printers" [some Samsung laser printers for instance].

After that driver install is easy and you basically print through CUPS.

But again, this is totally a manufacturer problem not Linux. It isn't like Linus can force manufacturers to include Linux drivers for their non-standard proprietary shit.

Tom

Re:On the desktop and haven't looked back...

[Kangburra]

You're saying because your printer manufacturer hasn't followed the 25 year old PS standard that Linux is broken?

No, but most people will think this. If you know enough to even know what PS stands for you'll have bought a printer that is capable of it.

To the general public things like printers may need the CD that came with them but not have to chase help forums to get a test page. Until this is changed, Linux is not ready for the desktop. So we have to complain to the hardware manufacturers until it does work out of the box.

Re:On the desktop and haven't looked back...

[neildiamond]

In response to the comment about why not use X printer just beacause it supports Linux...

First off, I used to buy only HP printers. A year ago I bought an Officejet 7110 to have the ability to print AND scan in Linux. (It didn't fax in Linux, but 2 out of 3 ain't bad.) Scanning from Gimp in Linux was fantastic for bulk scanning projects. I really miss that. However, the printer was never quite right (as in flaky). The print quality (when working properly) was identical to the HP printer we had 4 years earlier. Finally the whole thing crapped out right after the one year mark. I did some research and found out that Cannon had the best all around multifunction in my price range (Pixma 780). I can't scan in Linux which pisses me right off. Cannon won't provide drivers for printing. I think that some CUPS drivers supposedly sort of work, but I downloaded proprietary drivers from Germany that only let me print in draft mode (though are fine as a shared printing device using Windows drivers from another machine).

Just noticed today that there are hacked scanner drivers that I will have to test out... http://pixma.schewe.com/

Anyway, I based my purchase decision on...
Cannon
Lower cost of printer,
Lower cost of ink
Better print quality
Fax really works all the time!
scanning in Windows only (for now)
Printer is rock solid

vs. HP
Most functions supported in Linux
Lesser print quality
Fax half-worked (has this annoying power save feature that has it crap out after about 8 hours of idle on a UPS that cannot be disabled)
More expensive ink
Printer was flaky
more expensive purchase price for printer

I think most sane people (not necessarily on Slashdot) would agree that full Linux support isn't everything if the product still kind of sucks.

Re:On the desktop and haven't looked back...

[tomstdenis]

Again wrong wrong wrong.

Just because your printer handles PS doesn't mean you can just dump a PS file to it and have it print. Many use proprietary encodings of the PS data as it's sent to the printer.

As for installation ...

emerge -uD cups
cd samsung-driver ./install.sh

[I think it was called install or setup...]

The installation is GUI DRIVEN, you pick your printer and how it's hooked up and it installs the driver for CUPS.

From opening the box to printing [locally] you're looking at all of about 5 mins of work at most. Getting remote printers is a bit tricky [mostly figuring out CUPS] but still not that hard.

My old job has two Samsung LPs hooked up to CUPS in round-robin fashion. Not exactly super hard.

Just because HP, Canon and the others can't figure shit out doesn't mean "Linux ain't ready". The technology support in Linux is there, has been there for a long time.

This is the same as the game debate. Linux has long supported video drivers such as those from ATI and Nvidia. It also supports sound, keyboards and mice.

The only reason there are not a lot of good FPS games for Linux is because nobody is writing OpenGL engines. They're all too gung-ho on DX.

Let's see... SDL + OpenGL == portable game engine which handles keyboard, mouse, timers, graphics, network, fonts, etc and would be portable [even work in Windows].

Tom

Re:On the desktop and haven't looked back...

[LocoMan]

it's making inroads but it still needs some more work done, IMHO. I'm trying linux for the first time now on a second computer I have around (Ubuntu, to be more specific), and while I was pleasantly surprised at how usable was right off the bat (even the internet worked), it has dissapointed me so far at how many times I have had to resort to command line from the beggining (to get MP3 working), and how hard it was to get something I would have though would be simple as installing a video driver to get decent 3D. At least in my experience trying to install them, following the instructions on wiki.ubuntu.com got my a BSOD (well, it was a screen telling me X couldn't be loaded... but it was blue.. :) ) and sent me to command line only, managed to find some more instructions to uninstall them from there too and could get back on the GUI, then downloaded the drivers from nvidia.com and tried to install them... and failing because it didn't have what I needed for them. After 2 HOWTO's (about 10 pages worth of instructions) I finally managed to install them after 7 failed attempts (it wanted gcc.. then it wanted the RIGHT version of GCC, then it wanted some system variables, plus several other stuff I did from the HOWTO's that I have no idea what they did). At least that was my first experience with linux. My computer doesn't have any weird stuff, and the video card is a nvidia 5500, which I would consider fairly mainstream, yet it took me 2 nights dicking around to make it work (in 3D, at least). And that comes from someone that's fairly computer literate and not foreign to command lines (I grew up in the time when DOS was king and we had to edit autoexec.bat and config.sys files all day to make our games work). I completely agree it's getting better, but at least in that area it needs some more work... 3D and MP3 playback at least (I haven't been using linux much yet and those are the only roadblocks that I've found so far) are things that are very mainstream by now and it should be a lot easier to get working... I understand they can't be included working by default because they're not "free", but maybe include a wizard to make it work without having to resort to searching for HOWTO's and command lines. That said, though, I'm still pleasantly surprised with it, and while I probably wouldn't use it myself on my main work computer (I'm a video editor and graphic designer, all the better tools for that are windows or OSX only) given more time with it, once MP3 playback is working, that is, I'd probably recommend it to someone that only needs a computer for simple browsing and email.

Re:On the desktop and haven't looked back...

[Trelane]

Just because HP, Canon and the others can't figure shit out doesn't mean "Linux ain't ready".

Actually, HP has done a terrific job of supporting Linux

These are at least Open, if not Free Software packages, and included in your distro (I've not found one yet that doesn't have them, what with them being FOSS and all.) To use them:

emerge [hpijs|hpoj|hplip]

and then the drivers will show up in your printer listing in CUPS (you have have to restart; I don't remember. I use the web interface; use whatever you're comfortable with). If appropriate, you can select the scanner in SANE; it just appears.

It is precisely because of this great support that I will be buying an HP Office Jet in the near future. It also makes sense on their part--the software is a way to sell printers, which is a way to sell ink. ;)

Thanks, HP!

Re:On the desktop and haven't looked back...

[c0d3h4x0r]

The parent poster's GNU/Linux experience mirrors my own. Even Ubuntu is unnecessarily difficult to configure.

Installing programs (that aren't yet specifically built as a package for your distro, which is rarely the case when you're trying to install, say, the most recent released version of an app) is a nightmare.

Installing device drivers and configuring X for 3D graphics cards is a nightmare.

Installing printer drivers and having them actually work in all applications is a nightmare.

Hell, just trying to find device drivers for many common devices is a nightmare. Half the time they are available only as source code you have to merge into the kernal, or only as binaries packaged for Red Hat, and you can't get them to work harmoniously with your particular distribution.

Trying to get something like video file playback working is a total mess, because there are umpteen different competing video subsystems or libraries with different applications relying on different ones.

HOWTOs aren't an answer, because often the information you find in them isn't quite right for your given distribution, and it just fouls things up worse than before you started following the instructions, and of course the HOWTO provides no insight about how to roll back the changes if you encounter failure.

All the Linux zealots keep cheering the improved usability of the Linux desktops. It's bullshit. Users still can't easily install the right programs or drivers, get devices working fully and reliably, make sense of asinine library dependencies that make "DLL Hell" look like a walk in the park, or figure out how to get basic scenarios (such as video file playback) working on their desktop.

Usability is about more than pretty point-and-click eye candy. It's about making tasks easy to accomplish, not just for newbies, but for expert users, too. It's about saving people time and hassle. It's about simplifying and making consistent the inner workings and structure of the system so that there isn't so much to have to learn and figure out. It's about making things painfully obvious to users so that they don't have to form their own inaccurate guesses through expensive and risky trial and error.

I'm sick of Linux and FOSS zealots claiming that something is easy to use just because mountains of documentation are available for it. Something is by definition not easy to use if you have to learn a lot of unintuitive stuff about it before using it to perform basic tasks.

If someone has to go buy a "Linux for dummies" book, or has to read a lengthy technical HOWTO document, or has to go post questions in user forums, to understand how the system works and to accomplish basic tasks, then the system is by definition unusable.

GNU/Linux systems in general are a complete disastrous mess of chaos. If the grand experiment that is GNU/Linux has proven anything, it's that top-down design is absolutely necessary when it comes to something as technically complex as software.

Re:On the desktop and haven't looked back...

[Kangburra]

Again wrong wrong wrong.

Just because your printer handles PS doesn't mean you can just dump a PS file to it and have it print. Many use proprietary encodings of the PS data as it's sent to the printer.

As for installation ...

emerge -uD cups
cd samsung-driver ./install.sh

What's this cd sam... crap. Click!

Most people can click stuff, typing is way too much effort.

Re:On the desktop and haven't looked back...

[Anonymous Coward]

You're right. Linux is far too difficult for anyone to ever want to use. Luckily you (and people like you) can still use Mac OS or Windows. Enjoy.

Re:On the desktop and haven't looked back...

[neildiamond]

See my other comment in this thread regarding HP Officejets. Yes scanning works, but the HP MFC printers suck in every other way. I got sucked in by that Linux support thing (which is nice), but I recommend steering clear of the Officejets.

Re:On the desktop and haven't looked back...

[Anonymous Coward]

The point is, if Cannon included Linux drivers with each of their printers, then they would work flawlessly in Linux. The fact that Cannon do not do so is entirely Cannon's fault, and not the fault of the Linux developers.

The mere fact that any hardware works at all under Linux, when the manufacturers themselves don't provide drivers, clearly shows the superiority of the Linux model.

Re:On the desktop and haven't looked back...

[dbcad7]

I understand your frustration. I have been into Linux for about 8 years now, and even I run into snags. I went through a stage where I collected and installed MANY different distros to try them out. For me, the Debian based distros (Ubuntu is one) offered the best experience because of synaptic (the gui for installing programs). Ubuntu itself was ok, but I ran into problems finding some packages that I wanted, so I went stock Debian.

The video driver problem for 3d that you talk about.. been there (with Nvidia anyway) and it is frustrating. But what you have to realize is that all these differnt distros are using different Linux kernels. If Windows was tweaked, upgraded, and changed at the same rate as Linux "at the kernel level" then you would have the same problems in Windows. If you are using Debian Sarge (stable) or Ubuntu with stock kernel, I would think that there would be no problem getting video drivers.. it's when you have to go for the bleeding edge (eperimental) stuff that you start running into headaches.

Some printers can be a nigtmare that's very true. HP creates Linux drivers for many of their products.. Kudo's to them. Those that don't, well that means somebody will have to write one, and there are many such. I think it a bit unfair to blame Linux for this. If a hardware manufacturer makes a product with drivers for a PC but not Mac, and you own a Mac do you blame Apple ?

I don't see how Synaptic is confusing for installing things. It's organized by catagory, and you can search for software by name or keyword. It's easier than going to "download.com", searching, downloading, and then installing.

It's been said many times.. start with a blank hard drive, installing and configuring Linux is not that much more difficult than Windows. Windows also has it's learning curve, and nobody knows how to do everything without learning it.. web, books, help files etc..

Re:On the desktop and haven't looked back...

[techno-vampire]

My computer doesn't have any weird stuff, and the video card is a nvidia 5500, which I would consider fairly mainstream, yet it took me 2 nights dicking around to make it work (in 3D, at least).

I'm not a Linux guru, but I do know why you had so much trouble. Nvidia drivers are proprietary and Windows only. People are working on reverse-engineering them, and have (more-or-less) working drivers, but the project is still going on. If they had access to the specs (Not the code, mind you, just the specs.) things would go a lot faster, but even those are closed. They'll have them working, eventually, but until they do, getting those cards working with any Linux distro is going to be hard. If you want to complain, and you should, do so to the OEM. If enough customers do, they might be more cooperative, although I doubt it.

Re:On the desktop and haven't looked back...

[Hydroksyde]

Nvidia drivers windows only??? I suppose it was the magical elves that live in the walls of nvidia's offices that wrote these.

Re:On the desktop and haven't looked back...

[techno-vampire]

Thank you, I sit corrected. I do know that until fairly recently these cards were problematical under Linux, but I don't have one and hadn't kept up with the issue.

Re:On the desktop and haven't looked back...

[Anonymous Coward]

From opening the box to printing [locally] you're looking at all of about 5 mins of work at most.

Bullshit.

I have been trying for several months, on and off, to get my Samsung ML-2010 to work with CUPS on Slackware, without success, dispite claims that Slackware is one of the distros that the printer claims to support.
(I get the error "Could not find a suitable printer!", a well-documented but unsolved error for various Samsung printer models.)
(The Samsung site is no help at all, as it stays stuck on the "Select a country" page, no matter what link I click.
It's the last time that I'll buy anything from Samsung.)

I have a workaround that involves booting up MS-Windows 95 and printing to file using the MS-Windows 98 driver (which I can't use to print directly to the printer because MS-Windows 95 doesn't seem to sufficiently support USB printers (and, no, I don't have a more recent version of MS-Windows, and I'm not giving Bill Gates any more of my money)), then rebooting Slackware and sending the raw file to /dev/usb/lp0.
In other words, I have to boot up MS-Windows 95 to create the SPL2/QPDL proprietary printer language file, then send the file to the printer using Slackware, so printing anything involves two reboots.
I wouldn't have to do this if the printer supported PS, HP PCL, or some other "normal" printer language.

Hell, if the SPL2/QPDL language were publically documented somewhere, I could write my own CUPS driver, but it isn't.

Standards wont make a difference

[gimpimp]

We've had standards bodies for a long time. LSB, Freedesktop, etc - none of will help increase market share. Sure, they make like easier for developers, ie a gnome icon theme will soon work on a kde desktop. But the single major problem on linux is dependancy hell. I have nightmares about this.

Repository based installation is NOT the way to go. Autopackage is just a pretty frontend around the same problem. Until we can install and remove applications as easily as OSX users can, we don't stand a chance.

If you were a new user to unix, what would you prefer:
A) open synaptic, search the thousands of packages, hope you find what you're after, install it.
B) download an app folder, drag it to your appliactions folder. go.

Without this ease of use, there's no chance. I still laugh at people who say linux is ready, whilst at the same time they can't install the latest firefox on their box because it depends on the latest gtk which depends on the latest glib, which depends on....

Re:Standards wont make a difference

[tomstdenis]

What are these dependency problems of which you speak ... .... uses gentoo ...

Biggest problem with some commercial apps is they insist on using C++ and all the bleeding edge features of GCC 4.x.y. That's why they have "portability" issues. Specially on platforms where the C++ internal symbols have slightly different names (re: Redhat, try running Synopsis tools in Gentoo).

It's true there is a few too many OSS libs out there where some could be joined into one larger lib (or just even one package with various shared objects) but that's also the flexibility of the system.

But in general there are standards for development as you've said. So really and problem working cross-platform is usually of their own doing.

You can easily "install/remove" apps in Linux. Gentoo is already there with the end user application. It just needs rollback capabilities and a shiny GUI wrapper.

Tom

Re:Standards wont make a difference

[gimpimp]

i've used gentoo, and though it's great for developers and people who like to get their hands dirty - it's not for everyday folk who just like to check their email with the latest Evolution, and browse the web with the latest Firefox.

Until linux has real drag and drop package management, not this searching through a 10000 package repo garbage, there's absolutely no chance in hell. And I won't recommend it to anybody either.

I love linux, use it every day on my pc's, but it's package management is trash. Mac users for example. If the latest beta of their chat client comes out, they can very easily download the package, open it, and run the app inside. it doesn't clash with anything on their existing system, it just works. On the other hand, I can't install the latest Gaim beta because it's source only. Then I have to worry about its deps. So depressing.

Re:Standards wont make a difference

[IamTheRealMike]

Repository based installation is NOT the way to go. Autopackage is just a pretty frontend around the same problem.

Well, autopackage was designed to deal with many of the problems repository based distribution has, so, I would strongly disagree with the notion that it's just "a pretty frontend around the same problem". We've put many, many times more effort into things like reliable/easy installs than making it pretty (though there is still much to do).

Without this ease of use, there's no chance. I still laugh at people who say linux is ready, whilst at the same time they can't install the latest firefox on their box because it depends on the latest gtk which depends on the latest glib, which depends on....

This problem affects any OS. You can't install Safari on MacOS X 10.1 either, if I remember correctly. It's true that Linux suffers this problem worst of all though, because there's no unified platform, and because there's no profit motive so little incentive for developers to go "the extra mile" to reduce system requirements. But it's a separate (though related) problem to how you install software.

Re:Standards wont make a difference

[IamTheRealMike]

Biggest problem with some commercial apps is they insist on using C++ and all the bleeding edge features of GCC 4.x.y.

GCC 4.x has been out hardly a year, almost no commercial apps "depend" on it. C++ on Linux is generally riddled with problems anyway but it's not the fault of people who use C++, it's the fault of the people behind the compilers and low level binary formats/tools ...

Re:Standards wont make a difference

[Florian]

download an app folder, drag it to your appliactions folder. go.

Unfortunately not. OS X programs often spread their files all over the file system, with a mess of binary configuration files, possible netinfo entries (akin to the Windows registry...), etc. There is no standard method in OS X to cleanly remove them - just deleting the application won't do the trick in most cases. Even Windows is superior in that respect.

Besides, downloading binary code somewhere from the Internet and installing it in your system is a security nightmare and practice that should be abandoned ASAP. I find the Linux/BSD model of providing all software in distribution-provided repositories blessed by the distribution's maintainers vastly superior to OS X, with unmatched clean and safe installation, removal and upgrading of software. (How, for example, do you upgrade all your Mac OS X software with one command or click?) I use both Debian and Mac OS X and find Debian vastly superior in this respect.

Re:Standards wont make a difference

[tomstdenis]

I dunno. I don't subscribe to the dumificiation of humanity. I agree that things shouldn't be harder than they need to be. But it isn't like "emerge -uD firefox" is so fucking hard to type. I mean look at how many people can hardly use Windows as it is. I think the trend should move towards "let's document our system and stick to standards".

Remember the days of the 200-page MS-DOS 5.0 user manual showing off all the commands with examples? What happened to that? For $300 [full XP pro] you think they could include a 100-page primer on using Windows. I mean it isn't like the CD cost them $300 and when they fully admit they're making money hand over fist you think customers have a right to demand more.

It isn't like there are not Linux books though. So if the user has to learn how to user their computer is that really so bad? It means they get better use of it and are not at someone elses mercy as to what they can run and how. I think that's a good thing. I could be wrong...

Tom

Re:Standards wont make a difference

[tomstdenis]

Where I work onsite [let's just say they like chess] they are using the latest and greatest C++ features of a "blue" compiler. Turns out porting to GNU and other toolchains is doable but not without some minor pains here and there. Do they need these features? Maybe, Maybe not.

My point wasn't to blame C++ but to say using the bleeding edge of compilers is not always smart. Similarly with other tools and libs that are fresh new.

You're right most OSS tools target GCC 3.2 or 3.4 in terms of "expected functionality" which is nice. But various commercial apps [like verilog tools] often use C++ and are built on platforms like Redhat where there are hidden renamed symbols. There is no reason why a userspace program like those of Synopsis can't run in Gentoo [or Debian or ...]. They don't because the C++ stdlibs are incompatible.

Tom

Re:Standards wont make a difference

[IamTheRealMike]

Hidden renamed symbols is a new one to me, do you have any more details about this specific issue?

Re:Standards wont make a difference

[moro_666]

so basically the thing you want, is full support. someone who would test all the dependancies and stuff when every package is released. and you want it for free.

good luck, i hope you get a ferrari for free too.

Re:Standards wont make a difference

[i_should_be_working]

If you were a new user to unix, what would you prefer:
A) open synaptic, search the thousands of packages, hope you find what you're after, install it.
B) download an app folder, drag it to your appliactions folder. go.

You forgot the part in B) where you search through the internet for the home page of the application. Then you read the home page trying to find out how to download it. Once you see the "download" link you go through a couple of pages asking you what version you want and what mirror you want to use. Then after waiting for the download you finally start the actuall installation.

Whereas with A) it's more like: Open Synaptic, use the search field to find the app faster than you would on the net, install it.

I prefer option A. It's more convinient for me and the repository based system has other benefits I'd rather not do without. I can see where you are coming from, but different people prefer different things. I'm just glad the distros agree with me (or rather I agree with them).

And for the record, it's not the distribution or Linux devs who are stopping app folders from coming to GNU/Linux. They already exist. Nothing stops someone from bundling everything a program needs in a self-contained folder. That's how most of the proprietary apps I use are packaged. Open source devs could do this with their programs too, but it would be more effort without much benefit when the distros are going to package it anyway.

Re:Standards wont make a difference

[baadger]

I agree.

You know sometimes I wish I could just goto Help -> Check for Updates in Firefox on Linux as easily as I can on MS Windows. It's laughable that the most well known of open source software doesn't function as seemlessly on an open source operating system as it does on a proprietary Microsoft one.

Hell, if my repository doesn't have the latest version of Opera (it doesn't) I say sod it and get it from the source, run Opera's 'install.sh' and i'm happy if it works (it does). Yet, theres no safe way to uninstall or manage that installation thereafter.

Microsoft's registry and filesystem arrangement isn't as pure as us geeks would like, everything thrown in a single 'Program Files' folder, the start menu and registry practically pissed upon, user documents stored in a subfolder of the user profile and settings folder on the same partition as the operating system ecetera ecetera.

The fact is though on Linux, you're forced to engage with the community to get what you want in the repositories, rely on using the distro flavour of the month to get the best choice, or get down and dirty with configure, make and the filesystem yourself. Some people never want to have to do *any* of that, and they shouldn't have to. How anyone can claim Linux will every make it to the average mom's desktop, without constant nannying by a geek (and yes lots of Windows users struggle by without one, and the spyware awareness situation is improving), unless they address these issues is just funny.

Re:Standards wont make a difference

[baadger]

But it isn't like "emerge -uD firefox" is so fucking hard to type.

Wouldn't it be easy to goto help -> check for updates in Firefox like on Windows? oh wait it's greyed out on Linux. What does "-uD" bit mean? Man page, manual page what's that? Why should I have to read this to understand it? Why is it called emerge? Why isn't this easier?

I'm a gentoo user too, but this isn't trivial to the n00b. The days of the 200 page MSDOS 5.0 user manual are long gone, in them days you didn't have the same joe blogg user on the computer trying to get their job done.

The fact is you shouldn't need a multi-hundred page manual for anything with a GUI.

Re:Standards wont make a difference

[tomstdenis]

I can't find the material right now. But if you try to run [iirc] Synopsis tools on Gentoo they'll crash. The trick was one of the __ functions was renamed. Apparently renaming it back in the ELF file is enough to get the program to work again.

If I had access to a Redhat box I could tell you.

I don't think it's huge problem but it is problematic.

So are the kernel patches [even Gentoo does]. I run vanilla kernels [always have] and I update when appropriate [or given the time]. But I mean what exactly is gentoo-sources-2.6.16-r3 in terms of a Redhat kernel anyways?

In essence all the right tools for good standardization are there. Just people abuse them to get "an edge".

Tom

Re:Standards wont make a difference

[javanree]

This might have been an issue years ago, but these days there isn't any serious "dependancy hell" anymore. Tools like yum sort that out. As long as you pick a sane combination of repositories things will "just work"
For Fedora (only one I'm familiar with), there's freshrpms , Dag and a few others that work great. For the distro I use (CentOS) I maintain my own repository, so all other users just have to click to get what they need.

And if you want one-click install, have a look at Klik, which is now available for many distro's already. Although I personally prefer RPM's (since it's easier to clean/upgrade) it's a good idea for novice users.

Things like LSB and freedesktop ARE making a difference, although some of it might not (yet?) be visible on the surface.

Re:Standards wont make a difference

[NorbrookC]

You know sometimes I wish I could just goto Help -> Check for Updates in Firefox on Linux as easily as I can on MS Windows.

It's not just checking for updates. I've found it's often much faster and simpler to use the Windows version of some OSS projects for evaluation. Yeah, yeah, yeah, I know, people are going to tell me how "simple" it is to do this with Linux and give me a set of procedures for their distro. What I want is to be able to download it, install it, and go. I don't want to be downloading distro-specific libraries, etc., etc., etc., particularly when all I'm trying to do is see if application X has the promised features/capabilities I need, and then have to reverse the procedure if it doesn't.

Re:Standards wont make a difference

[init100]

You know sometimes I wish I could just goto Help -> Check for Updates in Firefox on Linux as easily as I can on MS Windows. It's laughable that the most well known of open source software doesn't function as seemlessly on an open source operating system as it does on a proprietary Microsoft one.

This is because the "Check for updates" in Firefox relies on a well-known security flaw in Windows, namely that every user is running as an administrator, and thus has the authority to modify system files. In Linux, since you are not browsing the web as root, it is perfectly understandable that you cannot update the systemwide installation of Firefox from the browser window. By the way, package management systems would be screwed up if files they "own" suddenly are replaced by other files of (to them) unknown origin. What are they supposed to do if the user tries to uninstall the original package?

Re:Standards wont make a difference

[Homology]

[rant]

It is seldom I see that anyone is recommending Linux users to read man pages. I used to use Linux (SuSE, a few years ago), but quality issues and poor documentation made me move away from it. In general, the Linux man pages are of low quality (out-of-date, incomplete and buggy), if there are any man pages at all to read.

New OpenBSD users with a Linux background are unused to actually read documentation, and just post on a mailinglist without doing some research first. Considering the quality of Linux documentation, that is understandable behaviour. However, on OpenBSD, the man pages and other documentation is high quality and is expected to be read.

[/rant]

Re:Standards wont make a difference

[Kangburra]

so basically the thing you want, is full support. someone who would test all the dependancies and stuff when every package is released. and you want it for free.

This could be a good way to earn money in the OSS arena though.

Re:Standards wont make a difference

[tomstdenis]

Dunno about you. The man pages I read are usually POSIX or glibc functions and they are just fine. As for the various other random commands it depends. Most of the coreutils are well documented [e.g. "cp", "ls", etc].

The thing that is least documented would have to be /etc/conf.d/ entries. But mostly a quick google is all you need.

You have to keep in mind the "man-pages" package is actually a separate project on its own. It's not strictly part of the Linux realm.

Tom

Re:Standards wont make a difference

[cozziewozzie]

Repository based installation is NOT the way to go. Autopackage is just a pretty frontend around the same problem. Until we can install and remove applications as easily as OSX users can, we don't stand a chance.

We can do this already: Klik

The problem is that you end up with 200 versions of the same libraries, and the resulting memory and disk space overhead.

That's why this sort of installation is generally used for easy testing of things instead of a sane installation procedure.

Re:Standards wont make a difference

[init100]

Wouldn't it be easy to goto help -> check for updates in Firefox like on Windows? oh wait it's greyed out on Linux.

Read my previous comment about this for an explanation.

Re:Standards wont make a difference

[jeffehobbs]

OS X programs often spread their files all over the file system, with a mess of binary configuration files, possible netinfo entries

Wha...? What you say may hold true for server software .pkg installs, but these days those are few and far between, and by definition include a bill of materials that tells you where and what files are being installed (select "Show Files" during the install to see). Most OS X application software these days are .app drag and drop installs, and they will "spread their files" 99% of the time in three places:

1. ~/Library/Preferences/
2. ~/Library/Application Support/
3. ~/Library/Caches/

...and I have *never* had an app modify my NetInfo settings in the five (!) years I've been running OS X.

~jeff

Re:Standards wont make a difference

[grumbel]

### Biggest problem with some commercial apps is they insist on using C++ and all the bleeding edge features of GCC 4.x.y. That's why they have "portability" issues.

I have to disagree, while C++ incompatibilities between GCC versions are an issue, they are mostly an issue for distributors who try to link everything dynamically, for stand-alone packages that is hoverever not a problem, link statically and the problem is solved. The real throuble are mainly the Linux kernel itself and glibc, since those you can't solve by static linking or by including them in your package, and neither of the maintainer of them seems to care much about binary backward compatibility, so that the problem won't solve itself anytime soon in the future. And there is also the throuble with GCCs usability, which makes dynamic linking easy and static linking a nightmare to get right, which is why hardly anybody is using it in the first place, most simply don't know how.

Re:Standards wont make a difference

[asuffield]

Nothing stops someone from bundling everything a program needs in a self-contained folder. That's how most of the proprietary apps I use are packaged. Open source devs could do this with their programs too, but it would be more effort without much benefit when the distros are going to package it anyway.

Actually, it's not because it's more effort. It's because it is fundamentally a bad idea.

If you bundle everything you need into one blob for each application, then suddenly your system has installed several hundred copies of gtk, all at different versions. Obviously this is quite wasteful of space, but even that is not the real problem. This is:

A security advisory was just released for all copies of gtk before a given version.

What exactly do you do now? You don't know which of your hundreds of applications has got that code included in it. Even if you could figure it out, you now have to either rebuild all of those by hand (if you can), or go to each individual upstream developer and download an updated version from them. If you're a desktop user then you probably aren't going to get this done, so you'll be running with known security holes in some applications. If you're a sysadmin then you're probably going to find a new job.

I would say that the ability to install security updates in a reasonably painless and secure manner is one of the most fundamental tests of any distrbution method. Applications-as-self-contained-blobs fails it badly.

Re:Standards wont make a difference

[Homology]

The thing that is least documented would have to be /etc/conf.d/ entries. But mostly a quick google is all you need.

On OpenBSD it's seldom that I've to google for something that is part of the base install (and that covers alot). Most, if not all, config files are documented in man pages or other documentation available (like the excellent FAQ).

You have to keep in mind the "man-pages" package is actually a separate project on its own. It's not strictly part of the Linux realm.

This seems to part of the problem.

Re:Standards wont make a difference

[zerblat]

You know sometimes I wish I could just goto Help -> Check for Updates in Firefox on Linux as easily as I can on MS Windows.

If you're using Ubuntu, Update Manager will take care of the updating for you. You don't even have to ask it to check for updates, it does that automatically and notifies you if there are any updates. Plus, it works the same for all of your software, not just one application.

Other distros have similar things.

Re:Standards wont make a difference

[baadger]

For the most part most Window's apps bundle all the DLL's they need to run with the installer. This is inefficient when it comes to bandwidth, but it isn't necessarily the mess you suggest it is if the developer has done their job correctly.

For a start, these DLL's should be installed into shared location (Common Files, or the System folder). Secondly, most installers now warn and ask you that you are about to overwrite a file of newer version than is currently being installed, and all is well.

I don't see how you can claim Linux distributions are any better with the likes of "library.so.5" and "library.so.6" and related symlink mess that often entails. I can't even count how many times i've seen people have library issues on linux community forums and the like and the solution was to move a library here or add a symlink there.

Re:Standards wont make a difference

[drsmithy]

Remember the days of the 200-page MS-DOS 5.0 user manual showing off all the commands with examples? What happened to that?

It's now an online, searchable, cross-referenced, interlinked resource.

Books look nice on the shelf and are handy for helping get to sleep at night, but a good online help system is infinitely more functional.

Re:Standards wont make a difference

[swillden]

If you were a new user to unix, what would you prefer:
A) open synaptic, search the thousands of packages, hope you find what you're after, install it.
B) download an app folder, drag it to your appliactions folder. go.

I have both Macs and Debian Linux boxes, and (A) wins, hands down, almost every time. It is *much* easier for me to find and install software on Linux that it is on a Mac. Uninstallation is a breeze, too. The "drag-to-install" idea is nice, but to make it really nice Apple needs to add a central repository of software that I can easily search and then install from (drag 'n drop or otherwise).

OS X needs synaptic.

Re:Standards wont make a difference

[baadger]

The administrator priveleges dilemma is a good point. I'm absolutely infuriated with the way Windows forces me to have TWO administrator accounts on my machine because you need another in addition to the 'hidden' default 'Administrator' account. You cannot downgrade your user account to Limited User if it is the only account bar 'Administrator'.

There are advantages and disadvanages to both the repository and proprietary installer methods. With repositories I know atleast I will get updates eventually during a sync. On Windows I always know I can have upto date versions NOW straight from the vendor as long as I remember to check for updates myself... or use the applications build in update checker (which doesn't require administrator priveleges).

Perhaps Microsoft should implement a 'check for updates' API that proprietary vendors can use to deliver update *notifications* as a sort of plugin to add/remove programs and Windows Update. Call it Application Update or something. That would certainly solve alot of problems with updating on Windows.

Re:Standards wont make a difference

[1in10]

Option A, and I didn't even have to think about it. Installation management is the biggest strength Linux has over other operating systems. Instead of having to search all over the place for applications, and having no easy and consistent way to install, manage and update them, it's all done through a single interface that keeps track of everything.

Why on earth would anyone think this is a weakness?

Re:Standards wont make a difference

[Knuckles]

I fully agree with you, but it would be nice if clicking "update" from within Firefox would call a package manager function that spawns a window, asks for the pw, and looks for a newer (e.g. beta) firefox package in the distro.

Re:Standards wont make a difference

[Anonymous Coward]

OSX is crap...

You can install Firefox on windows 98. (and 95 osr2 i believe with some tweaks)
Most of the major free software can run in win98. It makes linux undesirable when you see you need to update all your distribution to get some newer apps.

Re:Standards wont make a difference

[tal197]

Repository based installation is NOT the way to go. Autopackage is just a pretty frontend around the same problem. Until we can install and remove applications as easily as OSX users can, we don't stand a chance.

You already can: 0install.net. More easily, in fact, because Linux will automatically fetch the dependencies and check for updates. Things have moved on since the days of centralised APT repositories where you have to be root just to install something.

Take a look at the screenshots!

Re:Standards wont make a difference

[tal197]

This is because the "Check for updates" in Firefox relies on a well-known security flaw in Windows, namely that every user is running as an administrator, and thus has the authority to modify system files. In Linux, since you are not browsing the web as root, it is perfectly understandable that you cannot update the systemwide installation of Firefox from the browser window.

It doesn't have to be that way. When you type 'firefox', you want your computer to run the software from http://www.mozilla.com/firefox. Setting this short-cut (associating this short name with the full URL) should indeed require admin access if it is to affect all users. But, if you let each user store this association themselves, so that typing 'firefox' expands to telling the computer "Run http://www.mozilla.com/firefox", there's no reason why the computer can't be smart enough to only download and store the actual program code once.

See 0install.net for an example of this way of running programs.

Re:Standards wont make a difference

[Rich0]

Couple of things -

1. On windows the bundled DLLs definitely cause problems. I'm sure I still have PCs in my home which are vulnerable to that gif/jpg/whatever vulnerability that came out a year ago or so (the one where the flaw was in a series of DLLs that got bundled and repackaged with just about everything). On linux you use shared libs (which support multiple installed versions) and you can dodge this mess.

2. The .so setup on linux is designed so that you can have multiple versions of the same library installed (thus discouraging every app from just keeping their own private copies). In theory if two verisons are compatible you can just symlink them. In any case, as long as you run configure it should link the app against the appropriate library version - the problem only comes when you go to install binary software without using a packaging system of some sort.

Re:Standards wont make a difference

[Rich0]

Well, the big players pushing LSB tend to be companies that don't want to let anybody see their source.

I'm sure RedHat/SUSE/Debian/Mandrake/Gentoo/etc would be more than happy to build and package Oracle and Websphere so that they work perfectly on every version of their OS. However, to do that they would need to have the source, and Oracle and IBM don't want to give that up.

For open-source software LSB isn't really a big deal. If they install their netscape plugins in the wrong place, then Debian will just move them to the right place before packaging them up, and Gentoo will patch their Makefile.

I agree wholeheartedly that distro-managed packages are far superior to proprietary apps. I can run one command on my linux install and find out if EVERY binary on the system has any security updates available, and automatically upgrade them all. On a windows PC I can run one check to see if the core OS is OK. I can run another to find out if MS Office is OK. I can download a bunch of 3rd-party apps to scan for other problems, or run nessus on my network to get a half-decent idea of what vulnerabilities exist - however fixing them is a mostly manual process.

With most distros you can even check your binary hashes and ensure that every binary on the system is OK. For windows this requires a bunch of add-ons (and thus more risk of breakage).

Which isn't to say that windows couldn't go the same route. But, windows is oriented around proprietary software, and so a package manager isn't going to be super-effective. Besides, MS can barely keep track of its own bugs - let alone everybody else's...

Re:Standards wont make a difference

[Anonymous Coward]

When you get right down to it, most Windows programs only put things in 3-4 different places as well. However, that's not the GP's point.

Re:Standards wont make a difference

[waferhead]

Actually, all that is needed is a background script that can (for example) parse the output of an attempt to run a program, detect "liblablabla not found, or prefererably "_somecall.o", hit the repos database, and prompt the user with:

"This application requires XXX to work, which is available on "repository Y"
Would you like to install it and try to run the application again?
(PGP keys match)

With yum/apt/urpmi existing capabilities (the latter 2 have almost always worked better for me BTW) this shouldn't be a huge project.

Re:Standards wont make a difference

[octopus72]

Firefox, if installed into home/$user directory, will notify user of new version and update if he agrees. Can't be simpler than that. Distributions can also be configured (Ubuntu e.g.) to automatically update when new version pops up in the repostory.

Re:Standards wont make a difference

[colinrichardday]

I don't expect expect it for free, I pay for my distributions.

Re:Standards wont make a difference

[WhiteWolf666]

Well, there's always klik:// , which, IMHO, one ups OS X. I say this as someone who is typing on my MacBook Pro (while laying in bed), with my PowerMac Dual 2.7Ghz sitting in the next room.

Furthermore, I think you exaggerate the problems of "dependancy hell". I run SuSE on most of my systems (including my MacBook Pro), and I don't experience RPM hell anymore. I use the "Smart Package Manager", and in recent history I haven't experienced dependancy problems, even for packages I download. Furthermore, unlike OS X, RPM or APT based system will effectively "clean up" after uninstalled applications, while in OS X you'll have crap in /Library, /Library/Application Support, $HOME/Library, and other places.

Here's the problem with Linux packages: More developers need to make statically compiled binaries with decent quality widgets! If you install an application from your distribution's online repositories, your golden. Dependancy problems are resolved automagically, and you always remain up to date. If you install a statically compiled RPM manually, you have no dependancy problems, and as long as the software complies with some freedesktop base rules you'll get mime-type configuration and program menu entries. The problem is independant Linux applications that are not statically compiled and NOT properly setup on a distribution by distribution basis. This isn't a problem that will go away, not as long as developers don't provide static packages. There's no "legislating" around this (in terms of desktop base), unless you specifiy a "base" minimum configuration, in which case you'll be excluding window managers, and other common segements of Linux users.

Really, the answer is either a) use your distribution's package management system, exclusively, or b) use statically compiled packages.

Note that Windows applications are generally statically compiled, or rely exclusively on common Microsoft runtimes. It's not so hard to build a "standard" Linux package, either, if you assume the existance of Gnome and GTK2.

Re:Standards wont make a difference

[WhiteWolf666]

Not arguing with you, but two points you may be interested in, if you are interested up-and-coming package management.

1. Look at the "Smart Package Manager". It's surprisingly good. It uses APT repositories (among YUM, APT-RPM, ZYPP, YAST2, whatever), and is actually a good deal smarter than "apt". You can run them side-by-side, as well. Smart can be less strict about "broken" system configurations, which is quite useful. It also load balances repositories based upon response time.

2. Look at klik:// . It's similar to OS X's .app folder/application scheme. You aren't restricted to statically compiled apps either, IIRC, but most klik:// apps tend to be distributed like that. It needs a recent kernel version to really work well (FUSE support is import, IIRC.) All of the apps's settings, libraries, whatever, are contained in the "folder", and you start the "app" by double clicking the folder. You "uninstall" the app by deleting the folder. This is a good way to distribute binary-only apps like Skype.

Re:Standards wont make a difference

[WhiteWolf666]

Doesn't your distro do this automagically?

SuSE keeps everything up to date, automatically. I assumed that most big distros/package management schemes did something like this.

As it is, I don't have to check for updates to ANY software on my machine. It's all done by SuSE. Well, except for Cedega; I wish transgaming would setup an APT-RPM repository.

Re:Standards wont make a difference

[theCoder]

It's laughable that the most well known of open source software doesn't function as seemlessly on an open source operating system as it does on a proprietary Microsoft one.

It's laughable that because of Windows pathetic software management system (if the uninstall registry keys even count) that programs have to worry about how to update themselves. No, wait, it's sad. There's no good reason that an application should have to build in logic to check for and then update itself. That should be the packaging system's job. The fact that FireFox doesn't update itself on Linux is not a bug -- it's not its responsibility, nor should it be.

Re:Standards wont make a difference

[techitch79]

"Unfortunately not. OS X programs often spread their files all over the file system, with a mess of binary configuration files"

you're a bit misinformed. 95% of the OSX apps do not spread any files and do not need any installation. also, all the required files are stored in a single .app directory (except the preferences, which are kept under $HOME/Library/Preferences).

"possible netinfo entries (akin to the Windows registry...)"

i just browsed through my entire netinfo db with Netinfo Manager, and couldn't find any application-related entries, except for a few special user groups (ftp, daemon, clamav, mysql, etc). there are 154 apps in my Applications folder.

"There is no standard method in OS X to cleanly remove them"

yes, there is:
- applications installed using drag'n'drop can be simply deleted. you may also want to delete the application's preference file under $HOME/Library/Preferences.
- applications installed using an installer can be uninstalled by running the installer again, some installers even create an uninstall shortcut.

Re:Standards wont make a difference

[skinnygmg]

people are lazy. they don't want to read the manual. even if they get a free one. people (like me) who will read the manual, will go to borders and get what they need.

Re:Standards wont make a difference

[Knuckles]

I have no fucking idea what I possibly could have been thinking. Thanks for pointing out how stupid I was.

Re:Standards wont make a difference

[Dis*abstraction]

You're forgetting all the cruft under ~/Library/Application Support/ and ~/Library/Caches/, as well as the systemwide /Library/ equivalents and anything else the program may see fit to spew around your system (e.g. */Library/Safari/, */Library/Logs/, */Library/Scripting Additions/, */Library/Contextual Menu Items/, */Library/Fonts/ in the case of Microsoft Office, and on and on). And that's just for document-based applications, not even including system utilities.

I have to agree that the solution lies in abstracting the process of installing and uninstalling programs. There'd have to be a user-friendly, non-obnoxious way to pick and choose what data you want to keep (should deleting Mail.app also get rid of ~/Library/Mail/?). I'm not volunteering to program it, though.

Re:Standards wont make a difference

[i_should_be_working]

For Debian based systems there is gdebi.

If you have it you can just double click a .deb package and it installs. I've only tried it twice and those were packages for which I already had the dependencies installed, but supposedly it works as smoothly as if you had used apt. Actually I think it does use apt, it's just that it can be used for a .deb package that you already have sitting on your hard-drive.

Re:Standards wont make a difference

[asuffield]

For a start, these DLL's should be installed into shared location (Common Files, or the System folder). Secondly, most installers now warn and ask you that you are about to overwrite a file of newer version than is currently being installed, and all is well.

Yes, this is in fact a better idea than the applications-as-one-blob idea. I'm guessing that you haven't seen it (as used on macosx). They don't install their libraries into a shared location, they keep one copy in the directory with each application. It's basically like static linking.

Of course, the failure of the Windows method is that incompatible versions of a DLL are frequently given the same name - they don't have an equivalent of the unix soname scheme. So installing a new DLL version can break existing applications.

Independent library packages and dependencies are still a better idea than either of them (regardless of whether you do it like Debian, Redhat, or Gentoo) - assuming of course that the vendor gets it right.

Re:Standards wont make a difference

[Zanth_]

Agreed, however, in terms of the parent's example...until Ubuntu decides to include the latest stable version of a program into their respository, you can't just go through Update Manager and get it. I'm still waiting on a modern version of Firefox, I'll have to wait until Dapper Drake unless I want to install from source.

Re:Standards wont make a difference

[baadger]

1. The only way to stimulate a solution to that problem is to whinge to the developers/vendor. I imagine that you're perceived vulnerability to that vulnerability is indeed less than you imagine. Which DLL's were these? Why would any software bundle Window's DLL's? I would hazard a guess that the only reason these problems don't occur in the Linux world is the package maintainers act as filters for mess and crap like that.

2. 'msvcr60.dll' and 'msvcr70.dll' is almost as elegant, not to mention DLL's have the capability to host embedded version information just like exe's (no idea if .so's do) so if you find an incompatible version your application can always keep on chugging until you find a suitable one. Failing that extravagence, it occurs to me that if DLL's are both backward and forward compatible why not simply overwrite the old version and keep the filename the same?

Re:Standards wont make a difference

[killjoe]

Firefox does do that. It updates itself.

As for your other point. If you subscribe to an unstable repo you will always get the latest software. Beware though it's called unstable for a reason.

Re:Standards wont make a difference

[Rich0]

1. Ok, I took the time to look it up. This was a very big deal when it came out, and the list of applications that required patching on that bulletin was on the MS ones - a bunch of other vendors had issues as well. Of course the solution is to bug the developers - but that is only because there are no other options, since most windows software is distributed binary-only. On linux most distros would have striped out the local copies of the libs.

2. Windows has embedded versions in filenames in a few cases, as you've pointed out. However, this has problems. What happens when a new version of the DLL works 90% of the time, but not 100% of the time? In theory all apps should use the newer DLL (so you don't want to put the version in the name), but maybe a stubborn app might need to link to the specific version. On unix the solution is to have libabc-1.2.3.so with symlinks from libabc-1.2.so, libabc-1.so, and libabc.so - then you can link against whichever of the 4 is most appropriate (the last one working with any installed version). The problem with the windows technique of forcing the version in the filename is that it only works if you planned on breaking compatibility from the start - if somebody discovers the incompatibility later then you're up the creek.

I believe recent versions of windows have used a few tricks to let each application think it has its own set of dlls, which partially solves this problem, but it still results in having a lot of extra versions floating around that may or may not be necessary, and which could contain security holes.

Basically the windows model is much more distributed, while the linux package model is more top-down. The latter is better for getting all your security patches from one source, although the windows model scales better (with its flaws), since each vendor carries its own weight - the only problem is that if the vendor drops the ball nobody is around to clean up.

Re:Standards wont make a difference

[Arker]

Actually proper Mac applications DO reside entirely in a specially tagged directory called an 'app bundle' which you can install by simply dragging somewhere on your harddrive, and cleanly uninstall by simply dragging to the trash. Unix applications in Mac land, and system components count, are the major exception. They are tracked in a database similiar to RPM, and can be cleanly uninstalled as a result of that, although for some odd reason Apple doesn't include a tool to do that. But OSXPM, Free Software, handles it nicely.

Re:Standards wont make a difference

[Arker]

Yet, theres no safe way to uninstall

Ummm sure there is. 'Make uninstall.' Easy.

Re:Standards wont make a difference

[Anonymous Coward]

HA HA HA HA HAAAAAAA!!!!!!!! ROTFLMAO! Lin$ux better than OS X? Man, oh man, what kind of crack you are smoking because it sounds like it kills brain cells pretty damned fast. Idiot.

Re:Standards wont make a difference

[baadger]

wouldn't that assume you've kept the installation files?

Re:Standards wont make a difference

[init100]

Firefox, if installed into home/$user directory, will notify user of new version and update if he agrees. Can't be simpler than that.

I know that. This discussion was about the greyed-out update alternative in Linux, which applies to system-wide installations of Firefox

Distributions can also be configured (Ubuntu e.g.) to automatically update when new version pops up in the repostory.

Is that really true, or are you referring to the once-a-day automatic update cron job? The former method would require interrupt-based functionality (server notifies you) while the latter uses polling (you periodically check the server for updates).

Re:Standards wont make a difference

[init100]

use the applications build in update checker (which doesn't require administrator priveleges)

You can check for updates as a regular user, but you would surely need administrator privileges to install them, unless the application is installed in your home directory (or equivalent).

Re:Standards wont make a difference

[earthbound kid]

It needn't be that tricky. Every application in OS X already has an info.plist describing it. Just put some information in there about which files to uninstall. When the app is dragged to the trash, ask the user, "Do you want to delete supporting files and preferences as well?" and put a details button on it to list which extra files will get binned.

Re:Standards wont make a difference

[sirambrose]

I would agree that bundling every shared library with a program in order to make it portable is a bad idea. I believe that selective bundling is the right way to portably develop software. If the LSB standardizes glibc, gtk, dbus, openssl, and all the image libraries, then software can be compiled against those standardized interfaces and any additional libraries can be bundled with the application. For example a video conferencing application might bundle a library for SIP that changes interface frequently while using standardized libraries for everything else. Using the standard versions of common libraries gives the application developer most of the security and space savings of compiling against each distribution's libraries while preserving portability.

Re:Standards wont make a difference

[WhiteWolf666]

Oh, your sarcasm totally flew over my head. I had no fucking idea what I was thinking, huh? :)

Thanks for not whacking me with a big, heavy stick. Or sharp, pointed object.

Re:Standards wont make a difference

[Arker]

Well, you should have, but no, actually, it doesn't. If you were shortsighted and got rid of them, you can download the tarball again, untar it and make uninstall.

And, if you run into one of those silly ones that don't have a make uninstall target (fortunately that's quite rare,) you can use a switch (-n I think? man make) to do a dry run install. This will tell you exactly what was installed, and where, to do a manual uninstall from.

Re:Standards wont make a difference

[Knuckles]

I feel honored and and it's really tempting to let it stand, but seriously, when I wrote this I was not sarcastic, and when you called me on it, I was looking at my drivel and couldn't believe it, or even understand how I arrived at writing this.

Re:Standards wont make a difference

[octopus72]

Is that really true, or are you referring to the once-a-day automatic update cron job? The former method would require interrupt-based functionality (server notifies you) while the latter uses polling (you periodically check the server for updates).

Cron job.

Re:Standards wont make a difference

[baadger]

Good info, appreciated.

Re:Standards wont make a difference

[Keen Anthony]

Remember the days of the 200-page MS-DOS 5.0 user manual showing off all the commands with examples? What happened to that? For $300 [full XP pro] you think they could include a 100-page primer on using Windows.

Better than that, I remember the days of the Microsoft DOS technical manuals - the green books that had the original Microsoft logo on them. The books explained every command just as the white books did, but also provided detailed technical info to include assembly language examples and caveats. Yes penguins, there was a time when Microsoft inundated you with useful documentation!

I have an idea of what happened. DOS (and later Windows) became ubiquitous. Non-technical users became the most common kind of users, and they would report issues that were clearly addressable with a read of the manual. Producing those manuals was costly, so it was a major annoyance for the company to catch flak for things that were really non-issues.

Microsoft discovered that many consumers flatly refuse to read any manuals. Maybe while dumpster-diving for tech, Microsofties discovered chaches of DOS manuals, but no diskettes. Since consumers tend to respond more favorably to full color thin pamphlets (ie, they read them), MS went with those. So now, we get a thin guide to get us started and a lot of marketing speak to make us feel smart for buying the product, and we look to the computer book after market for support.

Why doth the rumours continue?

[Psychotria]

Unfortunately, those added software libraries differ among Linux distributors, making it hard to know if an application like a word processor will function on a particular Linux computer.

What a load of rubbish...

When I read a comment like this, I have to question a) the qualifications of the article author; and/or b) their motives. Any assertions made in the article need to be critically examined and their validity questioned after such false hoohah.

Re:Why doth the rumours continue?

[Shivetya]

don't take the word function to literally. This could be as innocous as "does it look right". Akin to running a window31 version of software on XP. It might work fine but it certainly doesn't look or "feel" the same.

The motive is to better the chances of Linux being on the desktop. This means that your going to have some badly worded observations that say one thing to one group and something entirely else to another.

It is the casual user who controls the market. The buy what is sold to them and what is sold to them is the simplest software at the lowest price point. Windows is filling this bill because for the most part its insanely easy to use for the common user. Most will have passing familiarity with it from work or school. Throw them on OS/X and I bet a few would be confused initially but would adapt pretty quickly. Throw them on Linux and they should adapt as well, until they sit down at another Linux system using a different desktop and they will be confused as to what Linux is.

If standards like these are not in place it will forever be "vendor's Linux" or some marketing gimmicky name. Eventually people may restrict their development to a paticular vendor and their extensions. Its best to set the ground level and keep developers and users focused there. Otherwise you will continue to be fractured and just meander in the low % of users.

Re:Why doth the rumours continue?

[mikesd81]

That's partly because, Linux, created in the early 90s by Finnish programmer Linus Torvalds, is really just the kernel, or core of an operating system. For a Linux computer to perform meaningful tasks, more software needs to be added that does things like presenting a graphical user interface.

I'm sorry, but I thought KDE and Gnome where just that? Personally I like KDE, but that's besides the point.

Also, I would think a big reason that it's not in the desktop market is Windows and Mac had the money to advertise and hire people to work for their company to id the way they want before Linus created the kernel. Linux just kind of came late into the game and now it's playin catch up and this standardization could potentially be a leap forward.

Standard don't remove freedom

[MarkByers]

i understand that this will help to push linux into the streets blabla, but is this really what we all want ? or is this the beginning of the end of linux as we know it ?

No. There will always be distributions that do it their own way despite what any standards organisations say. You will always be free to use these distributions. No-one can force standards into Free software (if you try, people can fork), but you can make the standards so good that distributions (and their users) want them. If people don't want them, they won't be successful.

Re:Standard don't remove freedom

[wysiwia]

...but you can make the standards so good that distributions (and their users) want them.

That's exactly what I had in mind when I created wyoGuide (http://wyoguide.sf.net/) albeit others have to decide themselves if I've come close enough. If I almost get no complains might be a hint but don't know for sure.

O. Wyss

Re: .sig

[Anonymous Coward]

http://angryflower.com/aposter.html

ONLY a few percent?

[penguin-collective]

A few percent desktop marketshare is what Macintosh has. Seems to me that the "fractured" Linux desktop is doing pretty well already.

Re:ONLY a few percent?

[Anonymous Coward]

http://marketshare.hitslink.com/report.aspx?qprid= 2

That's because Linux people like to make statistics up.

Mac OS has 4.37% marketshare according to that. Linux has 0.32%, pretty soon there will be more people running OS X on Intel (0.08%) then Linux.

Re:ONLY a few percent?

[Anonymous Coward]

The numbers you cite are complete bullshit. Sorry if you can't figure out for yourself why, but that kind of blind unthinking zealotry is apparently typical of Mac users.

Re:ONLY a few percent?

[IntlHarvester]

Their Mac numbers are definately higher than normally seen in webstat surveys, probably due to having Mac-biased sites in their pool.

But I've never seen any sort of webstat averages that puts Linux above 1% -- it's usually around 0.5% which is basically below the margin of error in these things and barely above bot-noise.

Re:ONLY a few percent?

[penguin-collective]

A lot of desktop Linux usage is not in places where people surf a lot (at least not to the kinds of sites that publish statistics), places like school labs, university labs, government, and corporate desktops. Many people also use Linux for many desktop applications, but do so by running the actual applications on a server (thanks to X11).

Furthermore, web site-based statistics consistently undercount Linux users (there are many reasons why Linux users are misidentified as Windows users, but few reasons why OS X users are misidentified as Windows users, and no reasons why Windows users are misidentified as anything else).

Overall, OS X is clearly still more widely used among home users, but probably not among desktop users.

OS X will probably focus further and further at the upscale home market; commercial desktop use, as well as nations like China and India, will be a battle ground only between Windows and Linux desktop systems.

What is this article talking about?

[Anonymous Coward]

Even on MS Windows, apps break on library updates (SP2 anyone?). I'm conflicted, I refuse to install PAM and other such garbage and the good thing about linux is that I don't have to. OTOH while I can't stand design by comittee this may be worthwhile if it stops KDE/Gnome apps requiring so many libs and daemons to run under usable DE's.

uhm what?

[Anonymous Coward]

lots of profit driven software companies are full of problems caused by egos, i mean after all, what drives profit?

the only way to have 'no ego' software is to have anonymous contributors.

I see more flying chairs...

[HangingChad]

...in Redmond tonight.

Re:I see more flying chairs...

[RobertLTux]

gee if on first boot you where given a wizard that had a pair of buttons labeled

[apply defaults and close} [make it look like windowsXP] at the start and this did things like make the Kong/Natilus window act like a "My Computer" window and build a "control Panel" link you would see conference tables fly

As a devout Linux desktop user...

[STDOUBT]

I have to say I don't care.

I don't care about all my apps having
a unified 'look 'n feel' (boring!)
I don't care there's no standard 'base-system'.
I like the variety!
Little things like cut/paste, I can work around.

If the home user wants to run Linux, let them
take the time to learn what it's all about.
Efforts like ubuntu which appear to aim at dumbing it all down for Joe Sixpack are IMHO, rather insulting. Fine -make your easy-peasey, unified system. I'll keep the chaos/versatlity thank you very much.

Re:As a devout Linux desktop user...

[jc87]

Efforts like Ubuntu which appear to aim at dumbing it all down for Joe Sixpack are IMHO, rather insulting. Fine -make your easy-peasey, unified system.

No offense , but if i had mod points left i would rank you a troll for that sentence! Why?

A) Ubuntu does not make users dumb , i for instance learned to edit/make my own sources.list , to create a working xorg.conf for my Radeon 9250 , to create a grub boot floppy , a few bash commands , etc.. a friend of mine after a few months even started to use Gentoo!

B) What is the problem with users not knowing how their computer works? Do i need to know how a engine makes the wheels spin so i can drive a car?

If you dont like distros user-friendly dont use them, just as simple as that , is the good thing about Foss , there are always alternatives and forks!

Re:As a devout Linux desktop user...

[Hiro Antagonist]

I'm going to take some time away from proving the Cauchy estimates (is there anything that man *didn't* discover?) and write a response.

As a former Linux desktop user and full time Unix sysadmin-and-programmer, I care very much about my apps having a unified look and feel, because I want people to USE THEM. So, if you're only writing applications for your own personal use, and never plan on distributing said applications, you can stop reading here, because we aren't even on the same planet, much less a congruent ballpark.

I don't write software that just fits my needs; I write programs that help other people solve problems, too, and this is mostly because software has a kind of magic to it. Unlike a physical tool that solves to a physical problem, like a hammer or a chainsaw, I can give the entire world access to my software tool without having to spend a dime. I can fix a whole worlds' worth of the same problem, freeing up other people to work on more interesting problems, through open source software.

That's why this stuff is so awesome; not just because you can see what's under the hood, but because it isn't necessary to reinvent the wheel every time you want to do something.

So, I, presumably we, write software to solve problems. Part of this is to put food on the table, but part of the reason why programmers love to write software is that we are natural toolmakers; we love to write logical tools that solve problems, and we share these tools because it gives us a kick that a big pile of people out there use, like, nay, depend on the tool which we have created -- otherwise, why make it available in the first place?

Still here?

So, we want people to USE the software that we release, not just let it sit there. Now, if we're writing system tools, they can be cryptic and terse, because other people like us are the ones who need them -- it's not expected that Joe Sixpack is going to know what to do with fdisk, setattr, or mDNSResponder. But a GUI application is another beast; it is designed, in all likelihood, to be something that takes a whole bunch of low-level computing tasks (disk I/O, shared memory, codec libraries) and turn it into something easy to use for a specific task (an MP3 player application).

Because GUIs sure as hell aren't for daemons or partitioning tools.

So, when we write this GUI application, we want it to be self-explanitory. It should be easy to use. And part of that is that our application should look like every other application running on the computer. It should exhibit similar behavior unless there is some reason for it to act differently than other applications, and if so, there should be some sort of warning that said application does so.

And this isn't just for Joe Sixpack -- context switching is expensive no matter who does it. When you switch between applications that act differently, your brain needs to make a cognative 'jump'. The bigger the difference, the bigger the jump.

So there's a good reason for consistency in user interface design.

On top of that, needing five different widget libraries is hugely inefficient; why should I need to give disk space to qt, GTK, wxWindows, motif, and whatever library-du-jour that some coder decided to use? Why can't there just be some uberWindowKit that is themeable and flexible enough to get out of the way, so that I don't need to install five different window kits?

This is the one bone I have with the way people develop software under Linux; too much of the 'not made here' syndrome. Which is why we have hundreds of X toolkits, web browsers, word processors, XML parsers, and so on, but only one good browser, a crappy office suite, no desktop publishing or photo-ready image editing. People spend too much time writing windowing toolkits because 'theirs is better', rather than contributing to an existing project, and so the real things that keep Linux from being adopted on the desktop are left to corporate interests -- like Mozilla and OpenOffice.

It

its an impossible job

[Anonymous Coward]

the best 'maintainer' in the world cannot keep up with the hundreds of sublibraries and different versions of sublibraries required for the various gui toolkits in linux.

programmers have an even harder time because they have to get the header files and development and/or debugging libraries for all this stuff. if two of their sublibraries use a 3rd sublibrary but each requires a different version... oi vey. they usually just copy all the source code into their own project which bloats it and locks out improvements/bugfixes/securtiy-patches made to the 3rd sublibrary.

not to even start mentioning the quality assurance issues. if you really want to test a program, to professional standards, you have to test it, as in actually run it, and actually do this on the different versions of libraries out there. so redhat, mandrake, suse, ubuntu, kubuntu, and so forth and so on, each have different libraries, and different locations for configuration files, and different settings used on some basic things like apache or mysql or X11 or KDE or gnome. so you have to test against all that stuff. and different versions of those OSes use different versions of the libraries. so you have to test against all that as well.

the less labor involved for the distro maintainers, and the less labor required for programming, and the less labor required for testing, the the more time people can spend writing useful programs instead of trying to make their program compatible with 5 different flavors of the libraries shipped on different linux versions, and the less time they have to spend learning the quirks and testing the 3 or 4 different ways of packaging programs.

if you want to make crap programs and leave it up to everyone else to install them, its not a big deal.

if you want to make programs that people can install with one or two clicks, like Windows has had for years, and macintosh, then you need more standardization.

apt-get install blah blah, yeah great. but it works differently for suse, redhat, etc. and besides, alot of times apt-get just wont work, or packages will be available in several different versions that are confusing.

thats because its hard to be a distro maintainer / packager and its hard to get everything perfect.

having more consistency would make everyones job a little easier and this would have reverberations throughout the whole system.

it has been gettin more consistent over the years.

Discussion..

[kahrytan]

I've got two questions for all of you Slashdotters.

What will it take for Linux to become a mainstream desktop operating system? It's the billion dollar question.

To help with that question, ask yourself this.

What do 97% of all computer users do on their computers?

1. They research information for school.
2. They talk to their friends via AIM, Yahoo, Googletalk, Trilian, etc.
3. They send emails to their friends and coworkers.
4. They use it to play games.
5. Watch or Listen broadband content. (Movies, Music, TV)

Linux needs an Open Source Standard in dealing with ALL graphics cards. Plus, a basic way in rendering graphics.

Linux applications need a good commercial look. GAIM looks to ugly for mainstream desktop.

Re:Discussion..

[DementialDuck]

First of all. I'm from Chile, and my English may be a little poor. So sorry if my English sucks. Yes. They need a very good support for all IM protocols including all features. I agree with you about GAIM. I'm a MSN Messenger user, because I'm from Chile and it's popular here. AMSN sucks, it's very three times slow than MSN Messenger from Windows, they emoticons sucks, it's webcam support is very poor. I think that Linux need a very good HAL. And if it's possible a HAL that support windows drivers, this way I don't worry about I'm in Windows or Linux. Much people may think 'but this is imitating windows', but Microsoft imitated Apple and Apple imitated Xerox. The good ideas deserves to be copied. The GUI models from Windows and MacOSX and the associated software are a very good base to start. About standards, yes that's a good idea, this way we can atract hardware developers supplying an API much easier for use that will make great operational costs savings. It's the same from the desktop, a very good standard for applications saves time in strategic decisions. A manager when decides to make a software for Linux, he lose time thinking about what graphical API to choice, GTK or QT?, that involucrates a cost of investigation for helping the manager to choose. Microsoft has only one API, one desktop, one IDE, one suite. Hardware vendors choice Microsoft because for them it's a fact standard.

Re:Discussion..

[STDOUBT]

What will it take for Linux to become a mainstream desktop operating system? It's the billion dollar question.

It will take an ubuntu-like distro with a name that isn't goofy and that comes loaded with all the multi-media codecs ready to go, along with a corporation to provide hand-holding... Oh, and a nice fat price tag so sheeple will think they're actually exchanging value for value. Personally, I want Linux to flourish in *business*. I use Linux 100% at home, and I honestly don't care if it ever goes "mainstream".

Linux needs an Open Source Standard in dealing with ALL graphics cards. Plus, a basic way in rendering graphics. It's called VESA.

Linux applications need a good commercial look. GAIM looks to ugly for mainstream desktop.

If you say so. But speaking strictly for myself, "commercial" is not a quality that I long for in my Personal Computer.

Re:Discussion..

[DementialDuck]

The discussion talks about who to make Linux take a more greater piece of the desktop market pie. I you don't worry about that, good for you.

VESA sucks, it don't support 3D Features.

You're comment doesn't offer an answer.

If we wan't to make Linux popular in the desktop. Is very good first to make a study what people like. I all software design, the requisites are primary.

Re:Discussion..

[broeman]

Go and ask that to Novell, Red Hat among other commercial Linux distributions, but what has Linux (or slashdot.org) to do with this?

How is it difficult to do task 1, 2, 3, 4 and 5 in a Linux distribution?

I guess Linux has a standard for graphics cards, or it would be hard for Nvidia or others to make drivers for the kernel.

Why don't you help making graphics for GAIM and others if it doesn't look good?

If some distributions want to have standards, fine with me, but this has nothing to do with hackers (appearently a slashdotter too) who likes Linux for its freedom. Call me an ideologist if you want, but this is why I choose Linux.

Others choose Linux because they want a gratis alternative to Windows. This is not what the most of the community including the developer's focus are on, so of course those people are anoyed and even frustrated. Somebody must have told them, that they can get a gratis OS, with free support. "Just download it on their site, ask questions on IRC/forums and you save a lot of money!", "they do it for free! use them! they don't care", "just annoy the hell out of them, and they will answer all your questions!" (which I read in comments some days ago here on slashdot)

I have seen many people leaving communities because they got abused again and again by freeloaders. It is really sad :(

Re:Discussion..

[Ignominious]

Q. Why isn't Linux a drop in replacement of Windows?
A. Linux doesn't want to be a drop in replacement of Windows.

Linux desktops already do all those things you mention, but don't expect them to be able to support all proprietary undisclosed/patented data formats and hardware.

If you think Linux software lacks a 'basic way in rendering graphics' it's probably because OpenGL, Cairo and countless non-3D libraries don't spend much money on promoting themselves.

Software for Linux is not lacking in capability.

Maybe if you want software with a 'good commercial look' then perhaps you should use good commercial software.

Applications

[walterbyrd]

Nobody runs an OS just to run the OS. It's all about the apps.

It doesn't take much, just one killer app, to sink linux as desktop candidate. That app could very well be a game.

As to your list of "What do 97% of all computer users do on their computers?" You seem to refer only to home users. Business users are a huge part of the desktop market. IMO: Linux fails even worse in the business sector. I know about OpenOffice, but there is *much* more to it than that. There are thousands of third party apps that just don't run on Linux.

Linux bundle files

[wysiwia]

B) download an app folder, drag it to your appliactions folder. go.

I don't remember when I suggested in the Linux kernel mailing list about creating Linux bundle file support in the kernel but it must be at least 4-5 years ago. Since nobody didn't recognize it's value then I haven't insisted.

I hope I don't have to repeat in a similar fashion the same in another few years about how to make the Linux desktop successful.

See http://lxer.com/module/newswire/view/54009/index.h tml

Re: I'll take A! - Repositories are The Answer

[anon mouse-cow-aard]

If you were a new user to unix, what would you prefer: A) open synaptic, search the thousands of packages, hope you find what you're after, install it. B) download an app folder, drag it to your appliactions folder. go. download from where? How did you find the place to download from? You did a google to find it first, on what term? the name of the software? the purpose? If the first step is to do a search in synaptic (btw... Adept in ubuntu is nicer) that looks only for packages for your distro, version, and architecture. It is far simpler. And from now on, all your security updates are applied automatically, so the user does not have to worry about the latest and greatest.

• A) is more:
  1. Start up a software repository browser. (Click'n'run, Adept, Synaptic, whatever)
  2. Do a reasonable search, to find the package, read among a few descriptions.
  3. click on the install button. (which downloads and installs.)
• B) is more like:
  1. google search to find the web site of the software.
  2. figure out where the download link is.
  3. Try to figure out which package is correct for the computer (This step completely defeats most ordinary folks.)
  4. Download it somewhere... (OK, it's on my desktop, now what do I do with this package thing?)
  5. drag it into the applications folder (That assumes that you know what an applications folder is... again, inexperienced people will not know or worry.)

Folks hear about downloading, and expect to download, and application developers find packaging a pain, a barrier to distribution, but once people look at it critically, it is really about what people are used to, not about what works better. Downloading random packages off the net is a bad idea on any OS. Getting supported packages from a repository that tracks your OS is the right idea. Vendors of proprietary software should (and the good news is that many are) simply provide repositories for distros that provide for this kind of automatic updating.

People who say that repositories are not uptodate are not reasonable. Most people want software that has undergone some testing, want software to update itself automatically once it is installed, want the correct version for their system to be chosen automatically (i.e. asking people to be able to answer the question "on glibc 2.1 based distributions..." is too much.) The the software provider cannot find the time to perform proper packaging, and will not arrange for updates to be easy to do when there are security issues or improvements available, then you should not install the software unless you are prepared to do that sort of support on your own. That is a choice that most people do not think about.

Making repositories easier to deal with is the thing to concentrate on. For example, A missing piece right now would be to have an XML ''download selector'' which would contain a list of repositories for various distros, that frontends for apt/yum/whatever could just download and automatically select the appropriate repository for a given distribution. ISV's would just create the XML file (and the requisite repositories behind them.) And the whole manual download/install process would disappear. That would be a big end user improvement with only a small change existing tools.

My main problem with Linux

[Bin Naden]

Linux is a great operating system. The main problem with adoption of this operating system in the mainstream is all the small things that makes its use difficult "out-of-the-box" and how most installation in Linux is so difficult. Here is a list of problems I have encountered with linux pretty much in the order I encountered them when I started dicking around with it: Installation stage: 1) Ooops, you have no hard drive! I had a SATA hard drive and teh distribution I was attempting to install (Mandrake linux) had no support for it. After trying a few distribution, only Fedora at the time seemed to support that hardware. 2) Hanging as the installation starts. Caused because of my newish graphics card. Therefore, I had to go through the text installation mode. Hooray! Linux is installed!!!! WOOOOTTT! 3) Horror! My nice graphical interface hangs at startup! That will teach me to have an ATI graphics card. After dicking around with the ATI drivers for a few days, I finally managed to make it work. 4) Oh noes, now my usb optical mouse doesn't work, I have to find out how to get it to work. After much googling, I find out that I must make some changes to the xorg.conf file and I therefore happily go twiddle with the settings in there. 5) Yippeee, my mouse works and I have a nice graphical user interface, now let's listen to some mp3's. First of all, let's mount my NTFS drive that contains all my MP3's.... NOOOOO!!! NTFS is not recognized :(. At this point, I get fed up with Fedora core and don't touch it for a few months. 6) On the recommendation of my friend, I install Gentoo Linux, follow all the instructions with a custom kernel and reboot. 7) Oh noes! Horror, I have no hard drive. After much googling, I find out that SATA hard drive are under the "SCSI" category for the kernel options. 8) No networking now! I must now rebuild my kernel with the Reverse engineered NFORCE 4 driver! 9) Now, do an "emerge kde-meta" and wait 10) 10 hours later 11) Yeah, I can actually mount a NTFS hard drive now. And I can play MP3's too. Now let's go check out those video clips on launch.com. ... ... 12) Try to get the totem plug-in to work in firefox. It finally sort of works but it's ugly and launch doesn't work anyways. But at least I can now watch my porn. Fiew... 13) After a session of "relaxation" let's now try to install half-life 2 on this machine. ... I won't even go into how much effort I put into this and never got it to work anyways. FOllowing that, I put linux aside a couple of months and decided to try some other time. 14) I got sick of ATI being so buggy in linux and causing my system to crash after every logout and shelled out the cash for a NVIDIA graphics card Therefore, the biggest problems to adoption of linux is: bad driver support (which ubuntu has fixed to a big degree but it still never got my graphics card right) and bad multimedia, gaming integration with the OS. The driver is fixable eventually if you tweak enough however it is inexcusable that an operating system will give you so much trouble in trying to play multimedia on the internet. I want to be able to watch my videoclips on launch.com dammit! The issue of gaming is being somewhat worked out with CEDEGA and wine but both are very buggy and iffy solutions. I got Deus Ex to work on my computer with Wine but all newer games refuse to install or get bizarre errors. Of course, if there is already a way to get all this to work in linux, let me know.

Re:My main problem with Linux

[caller9]

Right on! I'm sure someone will chime in "oh you didn't know SATA is a SCSI device for linux, lmao." But really most people don't know this and other crap until they've put several days of googling under their belt. If I had $1 for everytime I've had to jerk around with xorg.conf I could buy windows XP Pro. I've tried every flavor of Linux, all with moderate success. Ubuntu is so far the easiest to use. You can actually set out to perform something and complete it in most cases without needing to scour forums for that string you have to change on line 321 of the config file of a related application.

The biggest gripes I have are with closed source applications either not existing or existing in some either half-breed or reverse engineered way. Want to play flash? sorry not available in 64 bit. You can either do this complicated chroot hack or similarly use 32bit libraries to run firefox32 and the associated flash plugin. So I can view flash now, in a hobbled browser that I have to close the 64 bit version to load. Embedded windows media file? get ready for some dead-end googlin' but you might be lucky, I wasn't.

It gets tedious when after several months of this "learning process" you're now an intermediate user with understanding of the guts of the system...just to play a freaking animated piece of crap on a webpage, watch a movie trailer, or any of 100 other common user tasks in Windows that a preschooler can do.

This is the barrier to desktop adoption, the elephant in the room when discussions are about standardization of UI. The reason nobody uses linux is because it doesn't do what they want to do, click the icon, get the info/chat/send email/view porn/steal music right out of the box.

That is where one company can shine, create a great UI that answers all of those problems and figure out a way to take down the proprietary web standards that don't port a linux solution with an alternative that is free and interoperable. The company that does that will be a dictator of the standards by default given their install base.

Re:My main problem with Linux

[Bin Naden]

That's exactly what I mean. What Linux needs right now is a platform that will accomplish everything that joe sixpack wants to do right out of the box. I think that a script like automatix for ubuntu is a pretty nice feature in that it does take a lot of the guesswork out of installing the most common software needed. However, all the functionality in this script should be included in Ubuntu by default.

Also, what Linux needs the most is to have a media player that is well integrated inside firefox and that plays wma and wmv files. Either that, or make it possible to integrate windows media player inside firefox using wine.

The second most important thing to do for linux is to pour more resources into developing Wine. We need a windows emulator (I know it's not an emulator) that works flawlessly right now at least for games.

Re:My main problem with Linux

[quentin_quayle]

Unfortunately, your main problem with Linux will never be known to Slashdotters. That's because of your main problem with posting: lack of paragraph divisions.

Re:My main problem with Linux

[Bin Naden]

incidentally, my main problem with slashdot is that formatting using whitespace is not preserved ;)

Re:My main problem with Linux

[Anonymous Coward]

Oh wow. I should have expected the Linux snobs to mod this as troll. Too bad it accurately describes many people's experiences with various Linux distros.

Re:My main problem with Linux

[Gryle]

Your main problem seems to be that you don't understand the concept of paragraphs.

Re:My main problem with Linux

[Lauwenmark]

What Linux needs right now is a platform that will accomplish everything that joe sixpack wants to do right out of the box.

This is exactly what distributions like (K)ubuntu aim to achieve - by default, they provide what's needed to surf the web, type a letter, watch videos or read emails. Now, it is true that the default programs chosen for each task sometimes is a poor choice - but that's not a problem of platform definition, but rather one of "selecting the best-suited software for a given task".

I think that a script like automatix for ubuntu is a pretty nice feature in that it does take a lot of the guesswork out of installing the most common software needed. However, all the functionality in this script should be included in Ubuntu by default.

The problem is that "different people have different needs". (K)ubuntu aims at providing a "working basis" that can be fine-tuned by the user. This is not very different from what happens under Windows: the first thing one does after (re)installing Windows is downloading and installing those optional-yet-unescapable tools like Adobe Reader, Flash/Shockwave, or an antivirus software. It is hard to blame (K)ubuntu for not doing something Windows itself doesn't do.

Also, what Linux needs the most is to have a media player that is well integrated inside firefox and that plays wma and wmv files.

It already exists. In Debian and (K)ubuntu, this is contained in the mozilla-mplayer package. You may also need the Windows codecs (package w32codecs).

Either that, or make it possible to integrate windows media player inside firefox using wine.

Again, this already exists, although as a commercial product, which is called CrossOver Office. I cannot say how well it works with Windows Media Player 10, though, as I never tried that.

The second most important thing to do for linux is to pour more resources into developing Wine. We need a windows emulator (I know it's not an emulator) that works flawlessly right now at least for games.

The problem of Wine (or its games-focused spinoff Cedega) is not about the lack of development resources. It is mostly about the obfuscated Windows API/behavior. And each time a new version of a popular Win32 API comes out, a lot work has to be done by the Wine developers to (1)spot the changes, (2)understand how they impact the API and (3)implement them. Also note that games are actually the hardest apps to provide the 'emulation' for. Don't expect flawless support "at least" for games through Wine - if anything, perfect games support is what will occur last.

Novell, DBase, Lotus, WP, DR.DOS, soon AOL/Intuit

[WindBourne]

What did/do they have in common? All were/will be killed off by a company that is profit motivated and uses illegal actions. The ego stuff was done by these companies and then one low quality profit company has managed to kill all but 2 and they are just a matter of time (Intuit is doing ok, for the moment, but turbo tax is slowly being gutted and their targeting markets will happen soon with MS stuff; Intuit will start a slow downwards).

As a devout Windows desktop user...

[6Yankee]

As a non-Linux-user who can find his way around Windows (including the command prompt) very well, thank you very much, I find the "Joe Sixpack" remark "rather insulting".

It's not just Joe, it's people like my Dad - people who are willing to take the time to learn, but get frustrated at the idiosyncrasies and need someone like me to hold his hand now and again. And yet, I won't go near Linux - precisely because the last time I did there were so many idiosyncrasies that I gave up and reinstalled Windows.

A unified look and feel is critical - I get frustrated with things like MSN, ICQ and Media Player that hide the standard Windows buttons. And if you really want some fun, try Bryce. Eek.

Cut and paste isn't a "little thing" - it's fundamental, basic. It should just work, the same way in every app. Inability to copy from a given app (whether because it's a different keyboard shortcut or because the developers just plain forgot it) is one of the things that is guaranteed to piss me off.

I'm glad this is happening, though I have to agree with the poster further up who expressed surprise that this hasn't been going on for years.

IMHO, the Linux community needs to be doing a better job of convincing people like me that we can do everything we need to (even using WINE if we have to), and that it isn't going to be like pulling teeth.

I also need to be convinced that the tech support calls from Dad aren't going to be even more painful and frustrating before I'd recommend it to him. And, brave though he is, he's not going to go near it unless he knows he has ready access to someone who'll help him out.

Sorry, I want to use Linux and tell everyone else to, but I just can't. The rate it's going, I'd give it five years. At least.

Don't reinvent the Desktop

[Anonymous Coward]

I'm waiting for a 100% clone of the Windows or Mac Desktop. With similar window widgets and program groups. I don't want to relearn

Next thing is to use the font from windows to create all widgets - Arial -width of 1. Why is gnome using a bold fat arial-esque font.

Just copy for F'Sakes.

Re:Don't reinvent the Desktop

[Ash-Fox]

The Linux Gods have heard your call and bring to you this.

Cynicism is popular on /., but not always right

[swillden]

After the talk there will be 2 Major Faction. While one may win. The Second one will go Screw you and make their own design in-spite of the the talks.

History disagrees. While the Linux Standards Base and Freedesktop.org projects haven't solved all of the problems -- and probably aren't fully adhered to by any distribution -- they have already made a huge difference in the compatibility of Linux distributions, and I think efforts like this are exactly what we need to continue pushing interoperability forward.

I say this, by the way, as a developer who just finished developing a cross-platform, commercial, binary-only application for Linux. The app I was working on definitely pushed the limits of the interoperability, since it was an authentication system that replaced key system components, and in spite of that it went very smoothly. The differences between the half-dozen Linux distros I had to tweak the package for were very small. Actually, the more difficult issue was making things work in spite of customizations the admin may have made -- I just had to punt on that one, making the installer intentionally brittle in the face of unanticipated modifications to, for example, the X startup scripts, and then providing the admin with the ability to customize the installer to adapt to local changes.

After my experience of the last year, I wouldn't have any hesitation about developing more "normal" applications to run on multiple Linux platforms, and I expect initiatives like this one (which is from the same consortium that brought us LSB) will continue to reduce the platform differences that cause problems. I think we may even be able to get to the point where app developers may actually be able to target LSB (or whatever its successor is called) rather than having to tweak for individual distributions.

you mean...

[penguin-collective]

You mean like in the commercial world, the Apple faction said "screw you" to the Microsoft/IBM faction and did everything their own way? Or like the MS Office group said "screw you" to both the MFC and Vista groups and keeps violating GUI guidelines on Windows? Or like Mac developers can't agree on a consistent toolkit (there are half a dozen different ones in common use), consistent look, or consistent installer?

It's good that the Linux desktop is being unified further, but it certainly has to fear no comparison with other platforms. You can start complaining again once Apple and Microsoft sit down together and decide on a consistent place for the menu bar. (KDE at least gives you a choice.)

What the fuck is wrong with Slashdot these days?

[MrHanky]

Just a couple of years ago, we would have discussions about how to do cool things with software. These days, nothing seem to matter because it's all about market share? Why should I care about market share? Why should you?

I thought this was a site for nerds, not for people who play the stock market. What's important for us is that we have cool software to play with, not that we have cool software to sell to random people's grandmothers.

Why don't you guys just fuck off back to business school with all your fancy ideas. Come back when you've made your own distribution with app folders and whatnot.

Real news at the end of TFA

[standbypowerguy]

At the end of TFA I found the following quote: "Installation by the user is easy..." Imagine that! An acknowledgement that linux installation is easy published in a major media outlet. Hopefully, this will encourage some folks to try linux. Installation of any OS may be beyond the "joe sixpack" crowd, but IMHO, most linux distros' installation routines now rival or exceed Windows' simplicity, and you don't have to type in a long, cryptic CD key ;-)

Re: I'll take A! - Repositories are The Answer

[grumbel]

### People who say that repositories are not uptodate are not reasonable. Most people want software that has undergone some testing,

Testing is done by the softwares developer, not the packager. That software in repositories undergoes any real testing, beyond what the original developer does, is a myth, nothing more.

The real throuble with repositories is that they won't work at all for current software, they are great for yesterdays software, but if you read somewhere about the cool new version of some piece of software a repository based setup forces you to wait month or years till you will be finally be able to install said software, if ever. That just doesn't work very well for anything beside software which you don't care about.

Last not least repositories just don't scale, just look at games, multiple gigabytes installs are not something uncommon these days, the current mirror setup which most distros use would simply be stretched beyond what it can handle if just a very few of those games that windows offer would ever make it into a repository. Sure, its unlikly that those games will go into a software repository anytime soon, but it clearly shows that distro managed repositories just aren't good for all software.

For core parts repositories are great, for the rest they are something between just outdated and completly unusable.

What are you smoking?

[twitter]

The Second one will go Screw you and make their own design in-spite of the the talks. That is the problem with Ego Driven Software vs. Profit driven.

Oh, we've never seen that in the non free world. Have you ever seen a free software advocate hire Madonna for a release or make a speech by projecting their head onto an 80 foot screen? Do you remember a little anti-trust trail where a parade of computer industry giants testified about how often and hard M$ would screw them?

Ego Driven Software while the Code my be better quality but have a much harder time agreeing with other people. But Profit driven Software tends to be more consistent but software quality tends to be a little lower.

The free software people are agreeing about things that matter, the rest is called choice and it's nice to have. I can drag and drop files from various Gnome programs into KDE programs and vice versa. For example, I can take a thumbnail from gqview and drop it onto GIMP and GIMP will open the file for editing. Can the crappy Windoze driver software that came with your camera talk to photoshop the same way? Can you change your window manager and expect the same result? Can you drag things from your browser the same way regardless of source computer, sftp, ftp, samba, etc? Agreement and code sharing is much better in the free software world where it counts.

On Debian, I have a choice of excellence. I have a spectrum of applications for just about any task and can pick the one that works best for me. I don't want them all to work exactly the same way. All I really care about is that it's free and it does the job the way I like to do it. Only M$ would try to turn this choice into a disadvantage.

Re:What are you smoking?

[willyhill]

Ah twitter, I'm really looking forward to your reply to the other post. Really looking forward to it.

Only a month late this time

[smooth wombat]

This article sounded familiar and sure enough, it was rejected on March 28th and has been sitting in my Journal ever since. While not the exact same article the concept is the same.

In my case it was Neil McAllister who penned the writeup for InfoWorld. For Neil's take on the subject, you can read it here.

Never let it be said that providing folks with recent information was ever a strong suit of this site. Unless you're counting the dupes.

Free Software Installs are Much Easier.

[twitter]

If you were a new user to unix, what would you prefer: A) open synaptic, search the thousands of packages, hope you find what you're after, install it. B) download an app folder, drag it to your appliactions folder. go.

You have skipped a common step. You go to Google and look for the package that does what you want. This is something that's actually harder to do in the non free world, where people lie their ass off in the trade rags. Once you have the package name, installing software with synaptic, apt-get, deselect, kpackage manager or any other tool is less likely to screw up than the commercial software. Because free software users are free to share code, your distribution will take care of dependencies without you ever seeing it. Dependency problems come when you quit co-operating and act more like the non free vendors. Non free applications, self compiled source and other "extras" are much more difficult than just using the standard package in the stable repository. You can do it, and it can be fun, but the other way is much easier for people who just want to get their work done.

You can use the free software tools to find packages if you don't want to dig through reviews and screenshots. Synaptic, kpackage and apt all have search functions that work. My favorite is "apt-cache search term [terms]" because you don't have graphical overhead. Following this with "apt-get install" is easy enough, as is installing all of the recommended and suggested packages.

The important part is that the process works. It's rare that an upgrade breaks anything, even in testing when using an "apt-get update". That's what you would expect from freely shared code == the end of dependency hell.

Cut & Paste

[xerxesdaphat]

Hang on, hang on... cut & paste? I know ctrl-v doesn't always work, but I haven't found one app in which good old middle mouse button doesn't work. Maybe I've just been lucky and not tried to use it when it doesn't work, but hell middle-mouse even works from firefox to vi. And at least for me, I prefer middle-mouse button anyway... ctrl-v also requires ctrl-c in addition to selecting the text. And as far as it being preferable due to it being a keyboard shortcut, well cut & paste is largely a GUI thing anyway (vi has it's own built-in cut & paste thing of course) so you're likely to be using the mouse and you got to use it to select text anyway.

Please correct me if middle-mouse is not universal.

Re:Cut & Paste

[Anonymous Coward]

Maybe people don't have middle mouse buttons. Maybe they don't know about it.

It took me a week to get used to it. It is better, but converts might not relize it's an option.

Re:Cut & Paste

[xerxesdaphat]

Left + Right mouse button emulates middle button. I even use it on my shite laptop excuse for a mouse touchpad thing. People don't know about it? How is that the fault of the developers? I don't know about you but I think that there has to be more FAQs and tutorials and what-have-you for using GNOME or KDE than anything there is for Windows. Maybe that's just because Windows doesn't need help to use because it's simpler but I don't know... there's no excuse for ignorance if you have an internet connection.

Convince "the RiverSide campus" and you are gold

[RobertLTux]

the day 95% of america is working or living 10 minutes away from a store that they can buy a copy of Linux is the day that a certain person will get to the point of ripping sheetrock from the walls and throwing the chunks (because he already threw all the chairs lamps tables phones ...)

Compare and Contrast

[Anonymous Coward]

Kernel - Torvalds
KDE, Gnome, et cetera - splintered

Now, do you see the importance of project management? It's a concept lost on most developers. Only a select few possess the acumen to evolve as such. Most linux software chains lack direction, and more importantly, influence. Entertain fanciful and fluffy clouds of ego and profit in your wildest imaginations if you must. They are just that. Why? OSDL provides the profit already. Torvalds the ego. You provide no real "insight" here.

Dream what you wish, but having followed|used linux for 10+ years, I came to the realization long ago it will always remain just a hobbyist OS. It's a server OS now solely because of Linus' project management. Desktop marketshare? Never. Plain and simple - you need Microsoft's business model, which Linus' himself puts into practice with an iron fist.

Your honor, the defense rests.

Re: I'll take A! - Repositories are The Answer

[anon mouse-cow-aard]

first:
The developer is not some lone wolf performing creative work in a vacuum. A smart developer will take all the help they can from other sources. The whole point of open source is that testing is done by anybody with interest. If you think no testing is done by packagers, then you know squat about Debian. In the Debian model, there are 'maintainers' (the people who put software into Debian packages) as a completely
separate group from 'upstream' (original developers.) The maintainers try to make the software work well within the distribution, make reasonable guesses about defaults and encourage consistency in how things are managed across packages. In addition, software in Debian repositories is built from source, and dependency management is done automatically. For example, a security issue is found in glibc, if the change causes breakage, it will propagate backwards. The maintainers will do first level diagnostics, and will be aware of general breakage and fixes applied to other packages which got similarly broken. They may try to apply a similar patch and feed the result back upstream. They can also collate multiple bug reports and act as first line support before problems hit the developer.
This saves an incredible amount of work for the developer. Basically, the maintainer is doing distribution specific support for the package.

A dumb developer will insist on providing their own binaries "to ease the support burden" will be stuck personally supporting every version of every distribution, including learning all the quirks, dealing with the different release schedule, etc... with no help at all from the open source community. Providing your own packages is more work than having someone else help.

2nd: Yesterday's software?

What the heck is current? Have you tried Debian unstable? KUbuntu Dapper? Kubuntu is running Xorg, KDE 3.5.x, 2.6.15, etc... Folks who are willing to deal with some breakage can use pre-release versions and get the latest and greatest. Folks who prefer stability can get that, but need to accept older software. Folks who want just a few packages to be newer will run a stable release with some specific packages updated (via places like backports.org.) There is absolutely nothing stopping ISV's from creating their own repositories. That is even the right thing to do for companies that have sufficient commitment/resources to properly support a distribution's packages.

As for multi-gig games... Complete red-herring. If anything, Debian is proof of the opposite, A complete distribution is several DVD's and it works quite well, can be obtained via a number of means (such as torrents, jigdo, and straight iso's as well as DVD's if you want.) Providing a package for download is no different than putting a package in a repository, except that it is far less useful. If your point is that Games should be distributed on DVD's, that's fine, you can put repositories on DVD's too. That's how Debian works too. Repositories also automate the package selection for architecture, so you will not have to maintain two separate multi-gig blobs for AMD64 and i386.

3rdly:
How do you download updates/patches for your multi-gig game? over the net? Well then there is no difference between that and a repository approach. It is even better to use
repositories, because a good packager will split up the game into many, many distinct packages. When updates to the same package are supplied, the repository logic will mean that you will not have to download old versions of packages that have been replaced by future patches. With traditional blob patch approaches, you need to download all the patches and apply them in sequence.

Repositories are simply the best way to distribute software. It is only that people have a download reflex from the windows world that is nearly as bad as always being logged in as the admin. Download-itis is a sickness from windows, not a feature. Linux does some things better, and this is one of them.

Re: I'll take A! - Repositories are The Answer

[Pastis]

Testing is done by the softwares developer, not the packager.

BS!

repositories are just a tool to make your packages available. There are various degrees of public exposure and testing.

The real throuble with repositories is that they won't work at all for current software, they are great for yesterdays software.

BS!

Either your packages are shiny new (and only tested by their developers) either they are old and ironed out. You cannot claim both.

That again depend on the repository/distribution you point to.

Debian has stable, testing, unstable, experimental and then developer repositories for particular bleeding edge packages.

I already run Ubuntu 06.06 on 6 different machines. Do you really think that by the time it is out in June, only developers would have had access to the repositories? There are already thousands of users who are willing to help ironing out the latest bugs in exchange for running the latest and greatest. Same for every distribution.

And finally

multiple gigabytes installs are not something uncommon these days [...]

Nothing prevents a company to put their games on a online separate repository. Games are huge mostly because of the multimedia content.
1- it is easy to move this part of the game into a separate package
2- this part doesn't change much after a game has been released. So it's likely to not require a redownload when a software fix for the 3d engine appears.

So clearly it will take time to download a game, but at today's speed, getting 5G using a 10M connection is going to be much faster than getting 600M at 56k as I used to do 6 years ago. And maximising your 10M+ connection will easily be possible by adding P2P infrastructure to access the large package repositories. And it will cost much less to the game companies to distribute their products.

Come back read this in 10 years.

Perhaps they could call it...

[dacut]

... the Common Desktop Environment, or CDE. After all, the last experiment was quite successful.

That'll happen

[Anonymous Coward]

when there is a single way to talk to a graphics card.

Ask NVidia and ATI nicely.

Re: I'll take A! - Repositories are The Answer

[WhiteWolf666]

B.S.

SuSE does a LOT of testing on its package. SuSE will patch sources if necessary to fix bugs that exist, especially in important things like the kernel, KDE, GNOME, Firefox, and binary drivers.

You can see these patches by downloading the SRPMs. The patches are included in the build process.

FTFA:

[StarKruzr]

but has only a few percent of the desktop market.

I would seriously doubt it's even that high. I'd be more inclined to go with "less than one percent," personally.

Maybe you should toke a bit yourself

[heinousjay]

Only M$ would try to turn this choice into a disadvantage.

Only a zealot would reply to a general comment about developer motivation into a pro-Linux/anti-Microsoft screed with no connection to the parent.

Also, the M$ thing stopped making the point you want to make somewhere around the time Penny Arcade lampooned it so effectively. (I would link to it, but it seems to have vanished into the magical rails implementation that plagues that site.) You win no hearts and no minds with such rhetoric.

I guess what I'm saying is: calm down a little, and you may find your message is accepted much more widely. I say this as someone who believes you have a good message, so I offer this as friendly advice, not as a troll intending to insult you or anything you believe in.

Questions about OS/X

[spitzak]

Although I certainly agree that single application directory is the way to go, I have not found OS/X all that easy.

At least for me, everything I have tried to install involves the following steps:

1. Download the file to the desktop.
2. The file is a .dmg, double click it.
3. A mysterious disk is created and the damn gui gives no hint as to the pathname so I can't peek into it using the shell. In any case, typically the window opens showing the disk contents (though I have at least once had to locate the disk on my desktop and double-click it myself).
4a. *maybe* there is that great application folder in there and I can drag it to applications. I can also double-click it right there and try it, a great idea that you failed to mention. In at least one case (LAME) I have not been able to get the program to work after dragging the file anywhere, I always have to reopen the dmg and double-click it there.
4b. More likely, there is a .pkg file there, I have to double-click that and it gives me dialogs, and it could very well be doing more mysterious things than putting the app in /Applications. This, you will notice, is the same starting place most Windows users are in after downloading a file.
5. After installing, I am left with a .dmg and a disk, and am unsure if I can throw them away or not.

Now perhaps I am an idiot and am somehow using this wrong. And it is obvious that the underlying design could really be made easy. But for some reason the Mac software is not practicing what it preaches. The only technical reason I can see is the lack of a standard for downloading an already-working populated directory tree.

Replacing so/dll files

[spitzak]

Replacing the DLL's is a BAD idea, and I think all newer Windows installs don't do this. You can put the needed copy of the DLL in your application directory itself, just as the original poster suggested. Don't put anything in there that seems to be provided by default (easily tested by trying to run your software on a non-development machine) Though occasionally I have had an end user need some mysterious xyzzy32.dll and had to direct them to the net to get it (all I can guess is that they threw their copy away or some virus did it), the base functions seem to be consistently provided on all versions of Windows, and the bundled .dll's are typically very application specific and not provided by Microsoft (so there won't be any automatic patching of them anyway).

The same solution can work for Linux. There is a switch to the linker (-Wl,--rpath,'${ORIGIN}' passed to g++) that makes it look in the executable directory for libraries. (in addition, readlink("/proc/self/exe") can get the information that should be in argv[0] so you can locate files other than libraries). To figure out what libraries you need to include, just like Windows, try running on a non-development machine, perhaps a different distribution.

The problem with Linux is that there is a little less stuff in the base, such as a version of Qt or gtk that is likely to link with your program. And these things are truly huge, it would be like including an entire copy of Windows with a windows app. My solution is to use fltk, statically linked, but that may not fly with all developers. If you do include large libraries, be sure to tell them how to rename them so they can get their normal versions (this may also work on Windows except it refuses if the program is running).

Re:"One big things that's difficult is consistency

[PitaBred]

Because unpopular software on Windows like Winamp religiously follows the Windows design guidelines . How the hell did this blatant turfing for his own, really tangental site, get modded up? People will deal with new and non-standard apps quite well for the most part. This isn't about the interface presented to the user, it's about the parts that are common to all desktops like menus and hooks to the WM.

Until such a time as...

[zogger]

..a major vendor moves into offering a linux desktop in a big way, and I don't mean two models hidden back in their website that takes 10 clicks to find. When/if that happens, whichever distro they pick will start to become an industry standard. They *aren't* going to be picking any of the random hobbiest distros you see below the top 5 at distrowatch, that is a given, so I think folks who really want linux to suceed on the desktop need to take that into mind. Someone (a big hardware company) is going to crack first, my best guess is HP. But who knows there, just guessing at that one. IBM got out of the lower end market, so they are out (pity, I think they could have pulled it off) There IS a market for an easy to use, works outta the box, supreemly GUI intensive joe user linux desktop, if it existed. Right now there are several that are close, but not quite. And six month release schedules do NOT make a stable release. joe user is not even close to being interested in jumnping through that hoop, they want it pre installed and not have to fool around with it for a few years, until such a time comes as they want a new machine. I'd call it three years instead of 6 months, just do incremental security upgrades and make the dang thing work. We geeks don't care, but that leaves the other 99% of the population that does care, and don't want to be bothered with the complexity of brand new twice a year, said complexity that always seem to come with older stuff that worked before now broken and new stuff pushed as full release which is still betaware for most practical purposes. Now apple does it near yearly, but they have a slightly different market. I see linux fitting between windows and apple in that regard, on the desktop I mean.

Re:Until such a time as...

[Keen Anthony]

Add to that the requirement that whatever Linux distro the big hardware vendor chooses to market with must itself be a major player with a highly recognizable brand that will impress consumers of all types. This leaves Gentoo, Debian, Ubuntu, Slackware, and pretty much anything French out of the running. Red Hat is the most obvious choice. I would have liked to see IBM pull it off too. I even liked Corel. I thought for sure HP might do a Linux distro.

The reason for this isn't software quality or even delivery. It's the fact that consumers like seeing big brand names written on their products. Look at PC hardware commercials. They always remind the viewer that the computer runs Windows - either the powerful media operating system, Windows XP or "the unstoppable" Windows 2000. There's no practical reason for this since there is zero confusion in the marketplace that *any* doesn't run Windows, except the Machintosh.

MFC

[Anonymous Coward]

I agree on the HP MFC, it is just dismal to get it all working on linux. I have succeeded twice, then after I reboot (storms common here, I shut the thing down) they STOP WORKING. No idea why, don't know what changes, don't even have the foggiest on where to go look, no idea which new and improved shiny driver to use, but allegedly it's easy as pie according to their website. Why does it stop working then??? I have to go through, rip everything out I can find that has the least bit to do with printing, start from hoop jumping scratch, download, follow instructions off several web pages and linux forums, and then it may or may not print two to four hours later. right now it is not printing, just sits there, after hours and hours fooling with it. I can fire up the printer install thing, nada, doesn't even show up in the menu. I know to go to that localhost CUPS admin page but it doesn't matter if the printer doesn't show up in the menu! And I know it CAN, seen it a few times, so WHERE DOES IT GO if I shut the machine down????????? It's like nothing is attached to the machine. I am not a guru, just a user, and this has gotten beyond annoying. I give up on them, going to just call it the stand alone copier from now on (that part works OK as long as ink is in it and you use good quality paper that doesn't stick to itself) and get another printer. If it is a cheap laser printer as indicated, then so be it but MAN I am gonna be pissed if it don't and they claim it will! POSTAL TIME! MUAHAHAHAH-HA-HAAAA!

I am just tired of hardware that doesn't work and it is pushing me back to apple, even though I don't want to go back there. I've never tried solaris, maybe that works. maybe a bsd might work. don't know, sick of trying based on one guys anonymous screen name anecdotal that his stuff worked on some hel;p forum. I want to support open source, I really DO, but it has gotten harder, not easier over the past few years. Every new generation of distro I try takes a more powerful machine and lots more ram, and in the end all I want to do is surf the web, do some chat and email, have the damn printer and camera work. I remember having a nice simple webcam on the mac, worked outta the box with the install disk, having nice video conferences with my relatives, got to see my nieces aand nephews and joke aroundwith, fun stuff! Worked! Now, nothing! No idea what to even pourchase that might work. I was doing all that, surfing, printing, chatting, etc, in the mid 90s on machines then, with GUIs that worked and were easy to use on macs, so why can't I do that now???

I simply got to the point that apples cost too much (no, I don't want a mini) and when they went to a "unixy" way of doing things I did NOT want to go there, if I had wanted unix then I would have looked around and found something like that, I wanted a GUI for non gurus and some things that worked out of the box. So then, screwed. I figured as long as I was going to be forced into unixy stuff I would go with what I could afford, that meant linux on cheap x86 hardware.

Swell, just swell......

What is funny though in ye older days I had no problems at all with parallel connected printers and linux when I first started using linux,plug em in, set the printer, mash print, swtuff printed. But now that they are all USB at the store I can't make things work. I thought USB was supposed to "fix" that but you can't prove it by me. It doesn't see my USB camera nor my printer, the only two usb things I have right now. I dread the day my serial port external modem gets hosed (all I can get is dialup), you can't even find those anymore, and no way do I want to rely on another bogus USB experiment to just get online. I still use a ps2 keyboard and mouse, ditto there, I will hate it when you can't get those connectors.

I would gladly trade 7/8ths of all the new eye candy and features and applications on any large linux distro for 1/8th that actually worked and worked well, and I would pay a reasonable fee for it as well.. I don't nee

It's about time

[doc modulo]

It's about fucking time. Everyone can see that this will improve adoption. There comes a time when you have to stop experimenting for a little while, tone it down a bit, just long enough to produce something which combines the best of all the experiments.

Trust me, I know this because I'm also guilty of overthinking and not doing, for far too long.

I'd also like to see appfolders as that is another thing that's obviously right in my eyes.

Look at this directory structure for just a minute, it's interesting:
GoboLinux' directory tree

Re:Until such a time as pigs with wings...

[Anonymous Coward]

You make some fine points, but I've seen those trends all before over the course of 10 years.

You remember Gateway? Dell's equal at one point. And guess what they did a few years back. Yes, that's right. They were the first major hardware vendor to offer the linux OS to Windows as the sole desktop. Dell soon followed on select workstations with a Redhat 4.x offering. Where is that promised desktop market share now, set in thrust by your major hardware vendor? And linux Gnome/KDE then was every bit on par with windows 2000 as the current linux offerings are to XP. So, the excuse of "in a few years once the linux [ insert random linux standardization effort here ] comes to fruition, you'll see" just doesn't fly. It's a hobbyist OS. Always will be, until such time one individual with real credentials (ala Torvalds) says, "listen up KDE and Gnome devs, this is what you're gonna do, or [ insert random threat here ] is gonna happen", which never will in my estimation.

There's nothing wrong or incomplete with partaking in a hobbyist OS. I'm a linux fanboi and quite unashamed of it. The Apple/Mac rank stands in line proudly and salutes their own just as proudly. Really, who cares about market share anyways. I could care less about choice on linux myself, whatever that pie in the sky philosophy some hold dear means. I use it to develop, contribute, and enjoy. Alas, market share is driven by ego, and not by the developers, but the linux users themself; some untapped psychological need to belong to a greater whole - the very definition of a nerd. Wear that label proudly, and care more about using those talents to help other nerds be more productive instead. Eh, leave the market share driven ego to the attention starved children who frequent the linux forums with an axe to grind across Ballmer's face.

A preview of the standard features:

[Ivan Matveitch]

(a) A printing system that actually prints once in a while; (b) a Microsoft Office imitation that would run slowly on Natalie Portman's beowulf cluster; (c) input-method support for every country in the United States; (d) most importantly, Richard Stallman almost approves at those times when an unusual mood of tolerance comes upon him.

Talk like this just frightens off businesses

[Bob_Robertson]

The monumentally stupid thing is that "corporate" doesn't have to do squat! It's like talking about "The Average Person", there is no such thing. Each organization can choose the sub-set of software that works best for them without compromising compatibility with anyone else. Is Linux like using "minority" labor, as long as you don't see Linux in the executive offices it's ok?

What arbitrary authority figure has thrown tablets down from On High declaring that there must be a "standard" desktop that everyone has to use? That's insane!

So darn it just go do it! Use anything at all that uses ODF, and you're sitting pretty with 100% compatibility with everyone else! How about plain text? That's head and shoulders above what tortured compatibility Microsoft EVER provided with their proprietary standards! That's what writing commodity software that uses open standards means! The entire confusion between KDE, GNOME and anything else that is being touted as "desktop" is pure FUD!

The differences between distributions is miniscule. Skype is a perfect example, they offer one static and one dynamic precompiled tar file for any Linux system that, as far as I can tell, "just works".

No wonder the US Government is getting support for building Berlin Walls across the land borders, these so-called "free" people are a bunch of frightened SHEEP!

Actually, I overstate it just a little bit. Real soup-to-nuts business in America (at least) has always been very conservative, meaning that they do not change easily. Look how close to failure IBM had to get before embracing Open Source, which has turned into a massive profit center. Those companies who knowingly adopted Linux "early" did so quietly, because of the competitive advantage it gave them in the server room.

Now it's out of the closet, and let's face it what Joe uses on his desktop isn't going to make squat worth of difference in Joe's productivity.

What is coming is the Vista upgrade and Office 12. Microsoft is finally talking about breaking backward compatibility, which has been the one saving grace of going with Microsoft "on the desktop" up to this point.

Now is the time for IBM, Novell, HP and the mom-and-pop integrator out there to push and push hard: Linux-based systems will give you all the compatibility you want, improve reliability and security, at a fraction of the cost for software and without having to upgrade your existing hardware at all, even for those companies still running Win98 and Win2K systems!

Even a needed service that cannot run on anything but Windows doesn't suffer, because Windows machines integrate flawlessly into Linux-based networks. This fact cannot be overstated!

Compare that to having to not only buy Vista and Office 12 software for every machine, but throw away or upgrade all that perfectly working hardware. I believe the bottom line becomes very compelling if examined as a package rather than one-off.

Linux migration, because of the fact that applications such as Firefox and OpenOffice.org run just fine on existing Windows machines, can be accomplished in stages or as needed without employees being unable to share files. This is an IMPORTANT factor, often overlooked. Migrating 300 employees in a weekend is daunting and very labor intensive. Migrating 20 a week isn't.

Lastly, there's "retraining". The super secret tool in the Microsoft toolbox is the myth that changing versions of Microsoft software doesn't involve any retraining. Balderdash! They know it, businesses know it, now it's time to make them realize that we know it too. Especially with the multi-generation jump from 98/2K to Vista and Office 12! Integrators must offer basic classes or basic materials on "Using KDE" or "Using OpenOffice.org" or "Using Koffice" or whatever is being offered as standardized packages to buyers.

Integrators must convince the buyers that the retraining myth works both ways. Not only does staying with Microsoft not avoid lost p

The Issue

[Anonymous Coward]

In my subconscious I spend a lot of time thinking about Linux and I spend a lot of time disliking the MS mon$opoly. I mean, I really do not like MS, the company, the people, the technology.

But Linux, per se, has a self-designed mega-problem. One person can not write an application and release it for use. Tax prep(?), school operations software. The Linux straight-jacket. The only Linux apps I have used are bundled with the distributions.

Can someone please tell me what this is about? It is the one issue the stops the whole idea of a linux desktop for the general public. What is the deal? Why can you not install any software in the machine?!!! It is entirely captive? This should be the main discussion, as far as Linux and the general public.

One a separate note, I pretty much do not participate at Slashdot. My previous posts are not ever posted.

Linux - a nice dream. I use MS at work. I do not like green eggs and ham. I do not like you if you do not post my work that I spend time typing. I have a $100k education. I have lots of debt. I work full time. I affect one helluva lot of computers where I work. What else do you want? Should I curse more and be smarmy, or just lay a bunch of hightech on you?

Why can no one anywhere autonmously install apps in a linux o/s? What is that about?

well, there's a reason to have ...

[zogger]

....a decent market share. A few good ones really. First, foremost and of utmost importance is hardware support, still in 2006 much less than perfect, especially for the average non guru. Computers today are a lot more than simple box with a cheap spreadsheet and word processor onboard, they have to do a variety of things and usually rely on additional outside hardware and things like newish video and sound cards, etc. you will have to admit, functionality there is less than perfect, BUT, if "Linux" had a much larger market share, the vendors would pony up with their offerings.

    Second is cost due to aggravation, how many people have torn their hair out over getting their boxes pwned, been cootified, etc and have to take it to the fixit shop due to a near monopoly and bad coding? This is a real world, affects businesses, home owner level users, etc, up and down and sideways in the computing food chain, and it's mainly because the "market leader" has done such a bad job. We all get to pay for that though, even if we don't personally use their products, we are still affected by them in day to day living because so many others do. I'll call it the hidden expensive headache tax.

  Another hidden cost-my taxes, your taxes, going in large part to buy single seat licenses by the millions probably for a variety of redmond brand products, a humongous real expense that is mostly unnecessary for the bulk of government useage.

A further advantage would be speed in computing advances. Basing our societal advances on what a single closed source lumbering giant company releases is..well, it is silly. Sorta lame really. Bad car analogy time. Would we be better off with one giant planetary car company, then just a few hobbiest cars on the side, like only a few percent? I sorta doubt it...

    We need better alternatives and the alternatives have to be adopted (somehow) in such numbers as to be significant. Ya, we are nerds, and can "deal with it" and have fun, but we as nerds have a sort of duty to help out the rest of the planet with what we think up and make and do as well. It's not written in any sort of law, but I think we all have a little service to humanity deal going, and helping nudge along something better/cheaper/faster/freer is part of it. Or *should be*, call all of that just my opinion on it.

UNIX wars redux

[Danathar]

I love LINUX...use it...endorce it...but...

The fact of the matter is, NOT having standardized methods for things like graphical installation of software (like MS installer) is a BIG drag on desktop adoption.

Having so many linux distros is good for competition between distros and innovation, but horrible for commercial software vendors wanting to create products that will be bought by many people.

Graphical installers that pull software from repositories are still (generally) too complicated. I have to hand-hack X11 config files to get multi-monitor configurations to work. Stuff still just does not work "out of the box" as well as windoze in many important respects.

Get ready...if Apple ever decides to use the LINUX kernel (unlikely) it should put a WHOLE lot of pressure on LINUX distros to clean up their acts.

you can flame me now...I have my asbestos fire suit on

Re: I'll take A! - Repositories are The Answer

[grumbel]

### The whole point of open source is that testing is done by anybody with interest. If you think no testing is done by packagers, then you know squat about Debian.

You don't seem to have used Debian for long enough, many packages, especially those build for other architectures then x86 are not tested *ever* by *anybody*, they run through and automatic compile process and thats it. If you try to start them they will crash, end of story. Sure, sooner or later you might get bug reports or not. But more then once faulty stuff as sliped into stable, sure its not in the core packages, but in anything that isn't essential you can find tons of those issues. And that are only bugs caused by the packaging process, bugs in the software itself, unless they are extremly critical aren't touched at all.

### What the heck is current?

If something is released today, the user should be able to use it today, not tomorrow in two month, half a year or three years, at it is the case with todays repository based distribution.

### Have you tried Debian unstable?

Do you know that those are called *unstable* for a reason? If you go unstable get no testing, since you are the tester, no security updates and no guarntee that the next dist-upgrade doesn't fry your system. Yes, I do use unstable and I know how to fix it if there is throuble, my mom however probally wouldn't. The need to use Unstable is part of the problem, not the solution.

### Folks who are willing to deal with some breakage can use pre-release versions and get the latest and greatest.

Thats exactly the point, you have to accept that *the whole* distro becomes unstable to run *a single* unstable package on it. Would said software be distributed outside the distro everybody could install it with zero risk. And yes, I know you can backport and stuff to keep the risk lower, the point is that the distributions have absolutly no real official infrastructure for those add-on packages, its all just patchwork, which sometimes works and sometimes not.

Re: I'll take A! - Repositories are The Answer

[grumbel]

###### The real throuble with repositories is that they won't work at all for current software, they are great for yesterdays software.
### BS!

Explain me how you install the latest Mozilla, KDE, gtetrinet, SuperTux or whatever on your stable distro *without* by passing the official Debian repository? Answer: You can't, you have to have to wait a year or two for the next stable or compile yourself, be lucky and find backports, run some cross-distro package somebody build, etc., see the problem?

### Either your packages are shiny new (and only tested by their developers) either they are old and ironed out. You cannot claim both.

A packages that has been tested by thousands still has to wait for the *whole* rest of the distro to get stable as well, which his causing *huge* details on release, and thats already the whole problem. Would distros have some infrastructures to get add-on/updated packages into their stable distros their might not be much of a problem left worth to talk about, but they don't have such a thing, only security updates make it into stable after release and nothing else, ironically not even bug fixes for non-security bugs.

### So clearly it will take time to download a game, but at today's speed, getting 5G using a 10M connection is going to be much faster than getting 600M at 56k as I used to do 6 years ago.

The throuble isn't downloading it, the throuble is setting up the mirrors. With 100GB it isn't to hard to find somebody willing to donate that, but if the repositories go to 2TB or larger it suddenly becomes quite an issue to provide that amount of space, so most mirrors would end up either being incomplete or completly shutdown. Try to image a repository that holds *ALL* software ever written for Windows, all games, all apps of the last 20 years, that wouldn't be so easy to get right and thats exactly why repositories have a problem with scaling up. Its not just server space, but also testing to be done when a stable release should be done.

You don't know what you're talking about

[Anonymous Coward]

I have to question your qualifications. You've clearly never had to deal with library inconsistancies.

It must have been some system admins who modded you so high; no real programmer would have.

Isnt it already out there?

[martonlorand]

Taking in consideration the popularity of Ubuntu and a few other distros I think more that 80% of Linux Desktops out there are running either Gnome or KDE. A couple XFCE and the rest I think is little over the percentage of geeks running alternate windows shells (ex:blackbox).

These two are pretty consistent and you can run a KDE app over Gnome and Vice versa if you have the libs installed. It might look a bit different, but if somebody cant live with that little of inconsistency shouldnt be let near a computer anywhere. Not like the windows programs are so-so much consistent in their looks.

Pretty much the softwares for these two are much more consistent IMHO than the apps for Win.Apple is a different storry, but I dont think incostitency is the most important problem Linux have to tackle to gain popularity on desktops.