Slashdot Mirror
OSDL To Start Pushing on Desktop Linux
Psyke writes "The Australian Financial Review is reporting that 'IBM, Red Hat and a consortium of computer makers backed by the likes of Dell, Hewlett-Packard and Intel will push to move the Linux operating system out of the back office from next year.' and 'Meanwhile, the OSDL, which has largely worked on improving Linux's ability to run large servers, said it would work on improving Linux's performance on ordinary desktop computers.'" The article itself is a little off- those companies are working *through* the OSDL of which they are members - along with a number of additional companies as well.
The only two problems with desktop linux that needs
to be overcomed are
Pre-installation on new PC by default on mainstream computers
Support by the mainstream computers' builders.
Strange I thought Redhat had just abandoned the desktop. If Redhat are going to push Linux out of the back office, where are they going to push it too?
SCO / Microsoft sponsored conspiracies!!!
Oh wait. Wrong article.
Especially when RedHat was one of the companies backing away from the desktop in favour of enterprise installations. In the coming chaotic times for the desktop, I expect that we'll see lots of mixed signals as companies change directions or move in multiple directions at the same time. (In the last chaotic times, look at Microsoft's sudden shift from OS/2 to Win 3.)
One line blog. I hear that they're called Twitters now.
Of course Linux* is based on Unix. It may not be derived from the sources of Unix, but the idea and the running of it most certainly is based on Unix.
*I'm prepared to accept arguments from the GNU/Linux crowd here.
Cheers,
Ian
...when the number of Linux desktops reaches the critical mass where hardware manufacturers have no choice but to support linux with drivers, etc. If you'd asked me a year ago I would have told you that that day probably is a decent amount of time into the future. Now, I say we will most likely soon start seeing the signs on the horizon.
Often we hear people talk about how "linux isn't ready for desktop". Bah. Nonsense. I would rather say; Some people aren't ready for the linux desktop. It might not be as easy as Windows or OSX, but nothing really stops you from using linux instead of windows, except for specialized applications only available on that platform.
The only thing needed for success is a distro as easy to set up as Mandrake/Suse/Redhat, with the ease of upgrading of Gentoo or Debian. Maybe Fedora is the choosen one, tho I doubt it.
Ceterum censeo Microsoftem esse delendam
Seeing Linux distros as an option pre-installed into major OEMs systems would be great. Desktop linux has all the tools that the average consumer needs (especially those buying from OEMs). Not to mention that consumers might be more willing to use linux if they didn't have to actually buy Windows first.
Other things like, I dunno being able to buy software off the shelf in Best Buy, taking it home and simply running some autorun or one-click installation process, regardless of which distribution of Linux you are using needs to happen...
Easier configuration, better transparency for applications and functions, a common clip-board and many more things...
Even corporate desktops need a more cohesive system to be able to install, then later update individual packages without needing to install a dozen other packages, requiring a dozen other packages to be installed, which require still more packages to be installed.
I have personally run into issues like that when wanting to update things like the version of Mr Project that came with Red Hat 7.3 to the latest release of Mr Project in order to take advantage of some new features. Since there was not a single Red Hat 7.3 Binary package made available, I downloaded the 'generic' RPM and found that I needed to install a dozen things to be able to install it. So... I downloaded the tarbal source and found the SAME problem.
I love Linux, but it just sucks that I am unable to take a piece of recent software and install on an OS that is NEWER then Windows 98SE without having to update dozens of other pieces of software, when I can still take that old Windows 98SE and run MOST every piece of software that has recently become available. That is one of the largest usability issues keeping Linux from taking desktops over very easily.
Not everyone wants to update their ENTIRE OS all of the time. Why should people take a handful of hours to most of a day every 6 months or so in order to migrate to the latest release of their chosen Operating System? They shouldn't have to do that. They should be able to install it and update the pieces they need to and then when and IF, they have the time, then they can upgrade to the latest release. There should also be no major issues with doing so...
If you ignore the other uses of a tool, does that make the tool less useful, or you less useful?
What happened with IBM telling the UK gov that Linux wasnt ready for the desktop?
Also will the choose a certain linux distributing? Or just linux in general. Cause normal consumers would be able to choose for themselfs you know! They've going to have to have a list of suitable linux distros cause some of them are no wear near usable for newbies... And i assume thats what they are trying to do when they say they are pushing it for the desktop?
Giving IE users a taste of their own medicine since 2005 - http://pods.-is-a-geek.net/
Maybe we'll finally get a reliable copy/paste. Year or two... what the heck. We can wait. Its not like Longhorn is any kind of competition, is it?!
this is all almost like a movie. all you need to do is replace those involved with some more exciting entity, and you got the next billion dollar blockbuster:
Sco: liken to sauron, but with no real power
torvalds: some kinda of wizard
red hat: that land of humand you are sure you can trust
bsd: the dwarves that can kick anyone's ass but are more content with chillin in the mountains
Ibm: the elves that youre pretty sure are on your side...
and so on
xao
xao
http://TheHillforum.hopto.org
That doesn't mean that Windows will automatically die and Linux takes over the desktop. In means that you get a period where companies, developers and users all try to second-guess each other. Expect weird shifts as people try to flock towards the winner of the moment.
One line blog. I hear that they're called Twitters now.
"Red Hat, the leading supplier of Linux to business, also said it would produce major enhancements to its distribution of Linux, which would make it easier to use the operating system on corporate desktops."
Because i swear i read a couple of days ago that RedaHat will stop its RedHat Linux line, and stick with the RedHat Advanced server
The lunatic is in my head
Well, it is not like Redhat is the only company supporting Linux. They might say what they please, but imho players like IBM and Intel want linux to succeed, and more often than not, they get what they want.
I think the grandparent's last paragraph is interresting. Combine the power of some of the "more geeky" distros with the easy of use of others. Might be a winner.
---
The combined human population is enough to feed every living tiger for app. 28000 years.
Apple successfully brought Unix to the desktop with MacOS X. It's a very pleasing and easy-to-use GUI on top of unix.
It gives you the easy usability Joe Sixpack needs ("It just works."), yet still gives you access to a console (Terminal) and developer tools for technical people to do technical things with it.
If some group out there could slap on an OS X-like GUI on top of Linux that looks, runs, and plays as well as OS X does ontop of Unix (for non technical people and technical people alike), and have the OS be free, Windows would be done for.
Up, Up, Down, Down, Left, Right, Left, Right, B, A, START
It is difficult and wasteful to try to market products at too large a market. So, "Linux for the Desktop" is probably an unattainable and moving goal.
:-). Value: 20%. Linux: some work to do. Windows: perfect.
This is how I see the real market segments for desktop computers, their percentage value, and how well Linux fits. I apologise in advance for doing zero research and just basing this on my experience of the field, but... hey... this is Slashdot, exactly the place for uninformed opinion.
Here goes.
1. "Small Office" use. Value: 20%. Requirements: edit/print documents, spreadsheets, graphics. Web. Email. Music. Linux: perfect. Windows: perfect.
2. "Medium Office" use. As above, but add support for exotic hardware such as notebooks, scanners, DVD burners, whatever. Value: 10%. Linux: some work to do. Windows: perfect.
3. "Large Office" use. As above, but add integration with enterprise information systems, currently done mainly through Exchange and Office macros
4. "Cybershop" use. Value 10%. Requirements: web, chat, email, office, VoIP, p2p, trivial (re)installation, efficiency on cheap, old systems. Linux: perfect. Windows: too expensive and complex.
5. "Game boy" use. Value 15%. Requirements: support for latest video, audio, and large software library. Linux: needs work. Windows: perfect.
6. "Serious home user". Value: 10%. Requirements: as for Small Office, but more solid, tighter on the budgets, slightly hackable, and with loads of free software. Linux: perfect. Windows: slightly too expensive, but otherwise perfect.
7. "Naive home user", Value: 15%. Requirements: as for cybershop, but with ability to plug in digital camera to download snaps of baby. Linux: perfect, with some limitations on range of exotic hardware. Windows: perfect, except for security.
Overall analysis: Linux can cover 60-70% of the market with nothing more than some good marketing.
Ceci n'est pas une signature
In my opinion, the current state of Linux is still not going to match what Windows 2000/XP now supports.
For Linux to succeed in the desktop/laptop market on a large scale, they need to do the following:
1. Standardize on the programs installed. That means no more KDE-vs-Gnome wars, Mozilla Thunderbird/Firebird Internet access programs as standard, and most likely OpenOffice as standard. I hope the Linux Standards Base project will aim for such a standard for "base install" of Linux.
2. It MUST have widespread hardware support. That means it supports the latest graphics cards, sound cards, network cards and I/O cards at full functionality of the device.
3. We need the equivalent of Microsoft DirectX on Linux to make it easier to program and access multimedia devices. Hopefully, the SDL project will fulfill this need.
4. It really needs support for the Advanced Configuration and Power Interface (ACPI) for truly automated system configuration and the ability to easily hot-dock devices through the USB and IEEE-1394 ports.
If Linux succeeds in these four goals, then I can see its adoption by everyone become much more widespread.
AT&T is like Sauron: they created The One Ring (Unix), tried to use it to exert a hold over thousands of licensees, but lost it inadvertently (to the public domain and the valiant Berkeley).
SCO is more like Gollum: they got a hold of the Ring, they're convinced it's "My Preciousss..." and will make their lives wonderful, yet they are essentially unable to do anything powerful with it. They are also schizophrenic, having one happy Caldera personality that wants to be friends with Linux users and one evil The SCO Group personality that wants to kill them all.
I'd be aiming more for the enterprise and business desktop. Better defined hardware configuration issues, productivity tools for Linux are already out there and it could be bundled with back-end services in a very attractive end-to-end package, both in terms of price and function.
This is really an exciting time in Linux development. It's fun to watch it coming together.
That's our life, the big wheel of shit. - The Fat Man, Blue Tango Salvage
...wouldn't it fall off the desk?
This isn't a troll, it's an honest question.
I'm looking at switching to Linux because Windows SUCKS. But Linux can't really be very good; I mean, almost all of the Slashdot editors (Slashdot being owned, of course by "the sinister OSDN keiretsu") admit that they use Mac OS X. If OSDN themselves don't use it, why should I? Why should I go out of my way to use something that even these guys don't think is worth the trouble? Why help code/debug/improve/write docs for/ my own operating system when I can have someone else do all the work for me, all I lose is a little freedom?
I noticed that Doc Searls, editor of Linux Magazine, uses a Mac running OS X, and he raves about it daily in his blog. The EDITOR of Linux Magazine doesn't use it! Why should I?
No, Linux may be TECHNICALLY ready for the desktop, but if even Linux' biggest supporters (save RMS, of course) use alternate (proprietary, prebuilt, corporate-made, 'Cathedral') systems, it really can't be that good. You don't see Steve Jobs running WindowsXP or Bill Gates toting a PowerBook. Why do the Linux gurus alone refuse to "eat their own dogfood"?
I just don't get it.
What too many hard core Linux folks forget is that diversity and choice in the desktop isn't what most of the market wants. The vast majority of users see a computer as a tool and don't care about all the nuances of GNOME versus KDE. They want an integrated package with a few tweaking options and are happy to have a vendor pick a single stack and deliver it.
Microsoft does this and has 90+% of the market. Apple tightly controls their stack, including tight hardware control, and while their share isn't growing, they've tuned to their users and hold their base. But most Linux folks are developers or hobbysists. Many care passionately about what are inconsequential differences between KDE and GNOME.
Each technical point about everything is debated and the choice is usually no choice but another splinter project or variation. So much god work, but also a tremendous amount of wasted energy. MS and Apple are businesses. They look at alternatives and make choices and compromises to meet market needs. Linux is a hobby. The purpose os to tweak, customize and change, not to have the same Linux as everyone else.
The Enterprise / back office stuff is different. There the IT staffs are customers. They do want to tweak and customize the stack. Even though most of the time they really don't need to, they have the skills and time and $$ to do this so that they get the kind of custom IT shop environment they want.
These folks are not the home user / desktop user. They are geeks just like the folks who make Linux. They speak the same language and often care passionately about the same minor and irrelevant issues. And since this is where the $$ are for Linux (Red Hat's recent announcements confirm this)this is where the paid Linux folks will spend their energies. A different Linux is worth $$ here.
Why can't people accept that Linux on the desktop is just like APple. There is a niche market, geeks/hobbyists, and they want Linux on their desktop. The rest of the world doesn't care. Windows is just fine for them.
The Register has a cheery article Asian first-timers prefer Linux to Windows worth reading.
From the article:
So there. So what's all this about it being too hard?
They're talking about corporate desktops. Linux is arguably better suited to that desktop environment than Windows is. Not to say there isn't the occasional thing which needs careful consideration.
I'm currently managing several hundred Gnome desktops on Solaris for engineers at the moment but there's absolutely no reason it couldn't be Linux instead. Using the right architecture and using the workstation edition of redhat for the login servers and execution nodes you can scale to thousands of concurrent sessions fairly easily on very modest hardware indeed and with a significant saving in support and licensing costs.
Government of the people, by corporate executives, for corporate profits.
IF - and that's a big if in the linux world - you're a "Linux desktop for the rest of us" fan, you're in for a real treat.
:-)))
You'll have the pleasure of handling with a lot of gripes and user requests that you suddenly have to take serious.
The moment you say "Whip out Terminal" you'll have lost 90% of your user-base. Compare with W95 "Go to DOS" comments. Yeah, right, that's why users like GUI's, so that they can give commands.
Furthermore, you'll have the pleasure of contemplating a shitheap of philosophical and usability issues, resort to real-world testing, redesigns and what not ALL FOR SOME STUPID BUTTONS!
And this is only the enterprise desktop we're looking at. These people can still be trained (although after 20 years of mass computing we finally know training is the worst waste of your time and money since you could have put the effort in making a better application...).
As for manuals... I want to see the IT geek who for the first time explains to a group of account executives they should read the MAN pages.
I think, therefore I am...I think.
This is actually a real problem with many linux apps, both on desktops and servers. I just experienced it while trying to get perl-suid installed on a server, with deb (which is generally easier on install than RPM in my experience)...
.deb and .rpm to handle most common distros, as well as maybe source available.
So... base/stable version of perl-suid wasn't compatible with my newer Perl, so dipping into unstable I go, but then I also need to update Perl since unstable/testing perl-suid are newer than my Perl. So perl gets updated, which requires a new libc6...
This breaks the ogg/vorbis packages I have installed, so they have to get updated too... in total about 32 packages to install, 21MB.
Now, realistically this would be a HUGE pain in the butt for somebody with little experience to install for a single progam, on a regular basis. However, at 21MB you can probably fit most common updates on an install disc, and have your system automatically update on install as needed (windows does it with DLL's). The big thing is you'll probably need both
So really, a common package format would be really nice. It would also be great if I could update my "unstable" packages, without moving entirely to unstable (I'm hybrid, and my servers stay as far in "stable" as possible).
For deb, maybe an updater that gives the option of "upgrade to "stable" if current "unstable" package installed is = available stable, or keep "unstable" packages at current "unstable" version. Most games, etc, nowadays have online updaters, so it shouldn't be a huge problem to have something similar to keep the required packages up-to-date, without mucking the rest of the system
For the last 2 months I am trying to convert my desktop in the office to linux.
I am currently almost there, but the biggest problem that is holding me back at the moment is that certain business software is not available under Linux. The GUI for my accounting software package does not run under linux (it might under WINE, but not nativly). My Order Management System does not run under Linux, My Billing Application does not run under linux.
All the rest works via open office, outlook web access and other applications, but the biggest problem is that unless the core applications of your company are available for Linux, you can not switch.
So all this talk about hardware should be supported, USB should plug and play etc etc means nothing when my software does not work.
Rigolo
I've always had a problem with the idea of competing to get into desktop usage. Not in that we shouldn't be able to, but that the goal seems to be to compete with microsoft by copying them. Both parts of that seem wrong. This article http://news.google.com/url?ntc=04SL0&q=http://www. onlamp.com/pub/wlg/3971 talks about some of the ideas I've been thinking
I think there should be a new definition of desktop, and in true Linux fashion, it should support multiple definitions! one might be the way MS defines it, but everyone (including MS, look at how they are handling Longhorn)) don't agree the current one is very good. Linux and X-win is so flexible, I hate seeing projects like gnone and KDE shorting out all that flexibility to force every app to work with thier definition.
Me, I think of a desktop like a real desk: it's always there and I put things on it to use them. I don't need a desktop to use them though...In terms of apps, I'd like to be able to put a document on the desktop to work on it, but not have the document tied to to the dektop/PC I was using to work on it. X-win allows for this kind of idea already, but there's just no apps/setupsd to take advanatage of it. KDE and gnome are fine, but they assume that every app is only running under thier framework and thier framework can't do it all.
OK, done with the rant....for now....
AB HOC POSSUM VIDERE DOMUM TUUM
So, I was correct all along... Microsoft Windows *IS* based on Macintosh. It may not be derived from the sources of Macintosh, but the idea and the running of it most certainly is based on Macintosh. The look and feel is definitely there. The similarities are definitely there. Reading the history of Bill's "borrowing" (read: cloning) of Apple's GUI goodies adds weight to the geneology, regardless of what the judge says (that there are only so many ways to do similar things).
Linux has the "look and feel" of Unix, just as Windows has the "look and feel" of Macintosh.
Because they look and feel the same, are they necessarily related? My Chev is 90% the same as your BMW, they have big metallic/fiberglass bodies, four wheels, and perform the same functions, hell, they're even based on the same scientific principles! But nobody thinks they're derivatives of each other any more than waffles are based off toast. We all KNOW there is some swapping of ideas in the industry, and it's acceptable. I'm only so tall because I stood on the shoulders of giants, and all that. Is my stapler a derivative of your stapler, and if it is, what does it matter? I'm sure the code in Linux is more different from the code in Unix than the process to build my stapler is from the process to build your stapler, but nobody's demanding the stapler companies throw out all the good ideas in the stapler world and come up with a working stapler without using any previous technology to build on, efficiency and practicality be damned.
But assuming you're correct that things that borrow ideas are derivatives, I find that now, with the advent of Mac OSX, which is based on Unix (FreeBSD and Mach), I think we can accurately say that since Macintosh is based off Unix, and Windows is based off Macintosh, Windows is based off Unix, which means that Windows *IS* Linux. There is no escape from the Redmond Beast.
What I never really understood is why there is no such thing as an enduser gcc-frontend.
It might just be a small app where you drag&drop your tar.gz. The app then asks "Do you want me to install this package?" The beast could then run the "./configure && make && make install" stuff.
In case of trouble it would just fire up the console, so nothings lost, but a lot is won. I know it wouldn't work with any file but with quite a lot of the latest source-packages it would work fine and help desktop users to install software without even seeing the console that scares them oh so much.
Maybe I should start this project although I'm a lousy programmer. What do you think?
cu,
Lispy