Slashdot Mirror
Linux in 2004?
An anonymous reader writes "John Terpstra and Eric S. Raymond have started the ball rolling on LinuxWorld's poll of the community for what they think will happen in the world of Linux in 2004. Terpstra says 'I predict that during 2004 at least one significant USA government body will adopt Linux on the desktop.'" Depending on how you define "significant", this has already occurred.
I think this will be a(nother) great year for linux. Long live Open Source Software :)
It's not about having "good enough" software. It's about having feature-for-feature replacements that are open and secure. It isn't enough that sendmail, procmail, spamassassin, ical, etc. can be put together to implement most of the MS Exchange features; it's going to be the drop-in replacement that drives adoption.
Once people are used to using the drop-in replacements, they will be able to migrate away from closed and proprietary solutions. Until the drop-ins are available, Linux will not make huge inroads. (all IMHO, of course)
Overrated / Underrated : Moderation
Agreed. The power of apt isn't necessarily the tool, it's the repository that apt connects to. After all, it's the thousands of packages that are tested against each other that creates a cohesive system.
Overrated / Underrated : Moderation
Being low-cost is a good way to win people over. The software quality will speak for itself. If a distribution feels quirky, slow, and unpolished, that alone will make Linux look "cheap." If it is Done Right (TM), people will actually just think that Windows is too expensive.
Slashdot: Where people pretend to be twice as smart as they really are by behaving like children.
The kernel will get up to at least 2.6.10 by December '04, and KDE will probably release 3.3 (or 4.0) later on in the year as well, along with Gnome 3.0.
Slashdot: Where people pretend to be twice as smart as they really are by behaving like children.
# Which Linux application area do you believe will grow the fastest in 2004?
If not strictly meaning desktop applications, I'd say overall infrastructure. Web servers, mail servers, etc. And this will take place mostly in governments that can't afford MS licensing (it's already happening).
# Will 2004 *finally* be the year when Linux makes significant in-roads on the desktop?
No. The new X movements are just now gaining momentum, and it will take quite a while before it starts really biting into MS marketshare. I'd say 2006 maybe, like a previous poster. And that's *if* things go well.
# Which distributions will show the greatest growth in 2004?
I'd say Fedora (corporate), Knoppix (safety of cd distro), and Gentoo (great distro, great community).
# Will the SCO debacle slow Linux adoption over the next year?
No. I think it will die soon. It is just a matter of time before the whole thing is brought before a judge who is able to sort through the SCO lawyer crap, and when that happens, they'll throw the whole thing out.
# Will Tux finally get a girlfriend?
Yes. The hottie in Matrix 3. (he can have anyone)
# Or, make your own question(s) up...
Q: What is the single most annoying thing about the Linux community?
A: Irrational trash-talking about Microsoft. There are plenty of *rational* ways to criticize them, and people should stick to those arguments rather than ranting on and on about the same old tired issues. At some point the Bill Gates and Blue Screen jokes just lose their luster.
dmiessler.com -- grep understanding knowledge
While ESR seems to be very zealous and into the (GNU)/Linux scene, he's it's worst enemy. While Microsoft may spread FUD, people look at this guy and "wtf is this idiot doing? what's he talking about?" if i didn't know better, i'd avoid linux for the sole reason i wouldn't want to be associated with that nut.
"It is not how things are in the world that is mystical, but that it exists." -Ludwig Wittgenstein
Openoffice file conversions from MS Office work better. Yeah, they work pretty well now, good enough for probably 99% of files/users, but that small portion left creates a lot of headaches. Like it or not (I certainly don't) MS Office is the standard, and office app file compatibility is an absolute requirement for widespread adoption of Linux OTD.
Unfortunately, I think that all this SCO garbage is going to have an impact in the data center. I don't think very many CIO's are going to be jumping to adopt Linux. Even though it's all a bunch of bullshit it's still a risky move until all this blows over. No one wants to become a target for some sue-happy company that can't compete in the current market if it can be avoided. No flames please. Just imagine a CIO pitches Linux as a huge money saver. Then by some insane turn of events SCO wins and charges $699 a cpu. That cost-saving move ends up costing 3 times as much as before. It's a risk not everyone is eager to take. I may be wrong, but I just don't see Linux having huge success in the server market as long as SCO is still spouting off. In the long run I don't think this will have much of an effect on anything. Linux has made its mark and once things clear up I think many CIO's will consider it a very attractive option. On the other hand Linux has a huge opportunity on the desktop right now. The mainstream distributions are becoming more useable with every release. All the security nonsense with Microsoft can only help Linux as well. It seems like Windows never gets good press anymore and I'm not just talking about techie publications either. Every time there's a new worm it's on the front page of CNN. After people read enough of that they start to think maybe it's time to a less susceptible platform. Just my 2 cents.
With apt I know that if I install a program it will work or at least be given a configuration screen to edit the settings
Rus
Cheap UK and US VPS
I think there needs to be more unification and simplification over the way things are installed in , not only Linux, but the BSDs as well.
I think everyone agrees that rpms suck. Most of the good code comes in source tarballs - configurable for any *nix... but this is where the user experience falls apart. What person is going to want to dig out the command line to compile source code, and will he or she know about all the ocnfigure options... and then, will there be dependency issues (or should the source contain the dependencies too?). Then there are the legal issues of bundling dependancies... and then there will be future commercial Linux apps which won't want to include source code.
In an ideal world, packaged installs will be a compressed single file, containing all source code, configurable on any *nix like normal source code EXCEPT that now there's a graphical interface so that setting compile options, creating desktop shortcuts, and "Make clean, make install, make uninstall" now all work under X with a point-and-click.
PLEASE! Will someone serious about standardizing Linux installs do something about this... or desktop Linux will never take off.
im sooooo gonna get modded to hell for this but, the #1 reason linux sin't going mainstream anytime soon is the community. its greatest strength is also its greatest weakness.
its been said before, and i'll say it again, until my mom and dad can run linux without calling me every day, and they can just install something or simply copy and paste from one app in X to another, linux is just gonna stay a hobbist/server OS.
sorry to say it, but its true, dont give me the "its more stable, its more secure" stuff, you're preaching to the choir here, especiallly at slashdot.
Linux isn't going anywhere for awhile, im sorry, you just have to deal with it.
Everybody denies I am a genius--but nobody ever called me one!
First, the linux installer must be as easy as windows. Looking at the beginnings of the new Debian Installer, that is a definite possibility. They have the autodetection and the automation down. With a spiffy interface and maybe Synaptic in the installer that's about as easy as it gets.
Second the linux desktop has to surpass Windows XP in usability. They have the time to get this done. Longhorn is a long way off. Personally it would be nice to see some INNOVATIVE navigation ideas thrown around in the mainstream such as unified hotkey standards, radial pie menu in the window manager, and/or mouse gestures for launching commonly used applications (gesture down to open web browser, up for email) and common commands (down+left for copy, up+left for paste, for example). Maybe even mousewheel based window navigation instead of alt+tab.
Granted these things can be done now but not without some footwork. These need to be integrated into a "desktop" linux distribution like Lindows, Lycoris, or XandrOS. Somebody just needs to put them together.
Frequent tasks should require less keystrokes or mouse movement to accomplish. It isn't enough for it to be intuitive on where you should start to look for the document that tells you which clock does what. Less applications. Faster access and faster results.
Well Microsoft threw a fit and is one of the biggest lobbiests. They pressured dozens of senators with the phrase "Lost jobs" and "Communist" and they wrote legislation to ban code being release to the GPL and Linux at the NSA. Now are tax dollars are used to buy copies of Windows to help Microsoft.
Gotta love corporate influence.
Other governments its a different story because they are not all whores like ours.
http://saveie6.com/
Why are people so concerned with Linux on the desktop? Linux advocates should be spending all resources on making sure Linux keeps and expands it's adoption with the server. The desktop war is one that is long and hard and really Linux is not in a place right now where it can seriously compete with desktop offerings such as Windows or Mac OS X. What Linux does have going for it however is its fabulous server abilities. However great these abilities are, it cannot be overlooked Microsoft will keep spending more and more money to market their server options. Linux doesn't need some "validation" by being used as a desktop. Linux needs to keep improving as a server to make sure it stays superior to other server options. In time, the desktop may come. Until then, at least at this point, it is not something that is not too important. I only hope Linux keeps its focus and plays to its strengths.
"If you are a dreamer, a wisher, a liar, A hope-er, a pray-er, a magic bean buyer
Here's a question I'd like to throw out:
When, if ever, will there be a clear "winner" between Gnome and KDE for the average desktop?
When Linux takes over the desktop in a few years, will either one of them be the de-facto standard that nearly everyone uses?
Right now there's so much mindshare and development commitment in both, and it's hard to see when that will change. But I don't think it will last forever. Eventually one will have to give. Which one and how long will it take?
I'm afraid that other countries' reluctance to use Microsoft has very little to do with being whores.
I mean, put yourself in their shoes. You can use a foreign OS that you can't see the insides of to connect to the internet made in a country that is a superpower and wants to stay that way. Or, you can pick Linux, which may be made partially in the USA but at least you can look on the insides, modify it, and have your geeks keep an eye for any trojans.
Deciding not to use Windows has little to do with whether or not they're whores and everything to do with nationalism and national security.
In fact if anything, other countries are just whores to their own businesses instead of USA's businesses. Which makes sense, obviously.
Generally corporate interference with and influence on government increases -- out of necessity -- the closer a country is to Socialism. This should be obvious: In a Socialist system, influencing the government is the ONLY way for a business to be successful! For that reason, the US government isn't the big whore it's made out to be compared to the rest of the world.
2004 will be a year when many corporations, especially those who will try to adapt Linux as a primary desktop platform, will recognize Gentoo for several reasons:
Please, explain to me why.
* Portage gives a corporate IT the most fine-grained dependency control protecting the consistency of installations within upgrades;
I don't agree with this one. Corporations that "roll their own" packages have the same advantage. Movifying SRPMS can acheive the same effect.
* Gentoo makes possible to compile everything from sources on a reference hardware, adapting by that to the last bit of any available performance optimization, and then distribute the compiled binares to compatible hardware cross the enterprise (using GRP for fresh installations and just shared /usr/portage/packages for already installed systems);
Normally I would respond to this one saying that most people who use CFLAGS to optimize binaries actually hurt themselves, but corporations would have people that actually know how to use them best (i.e. -Os over -O3 or even -O2). However, I don't think that this is really an issue for corporations.
* Gentoo (mostly thanks to Portage) represents really the next generation design of Linux distro;
How so, specifically? There is something to be said for having a dedicated box to building binaries for the whole infrastructure, but the idea that Gentoo can do this and no other distro can is rather ignorant.
Gentoo is a really cool distribution (no joke), but I fail to see any technical advantages it has over other distributions. It's real strengths are in how it brings a lot of advanced administration techniques down to the level of an intermediate-level user. Plus the forums are cool, and portage is really well maintained.
Trust me on this one, though, there's no actual technical superiority over other distributions.
By the way, can you do reverse dependency checking yet? Like uninstalling gtk, and having every app that builds against gtk also unistall? I'm not "knocking" it if it can't (this isn't too important to corporations anyways), I'm just curious.
Slashdot: Where people pretend to be twice as smart as they really are by behaving like children.
Create a pool of Linux trained students, and they won't need 'conversion' to handle it in their workplace.
Breaks out tinfoil hat.
No sooner did the towers fall than China had us flags and bumber stickers shipping in at alarmingly sudden rates...
Patriotism for anyone in the free world should be the celebration of liberty, the glorification of the decision to exersize that liberty for the benefit of humanity, and the denouncement of taking advantage of said liberty.
Slashdot: Where people pretend to be twice as smart as they really are by behaving like children.
"I think everyone agrees that rpms suck."
Really? Who? Have you looked at the feature-list for RPMs? It has a *LOT* of good features that makes it a joy for packaging software programs. It's very well defined and easy to write spec file format(yes its nit-picky -- but that's good), package signing, package integrity checking(i.e. missing files), package querying, dependent lib checking, SRPMS format, idea of prestine sources, package roll-back(very cool), etc.
What sucks about RPMs is that rpm *installation* utilities are stupid. RPM was designed by Redhat as a format to install/upgrade distro-packages - so all dependencies would already be satisfied. Meaning you have to explicitly provide all dependent RPMs during installation. This is the part that sucks. The higher-level utilities are not smart enough to satisfy dependencies by themselves, and we experience dependency hell.
Using a tool like apt will solve this problem, but it doesn't know if a particular RPM is 'pure' or has been 'tainted'. A 'pure' RPM only uses libs/packages provided by the distro-vendor, while a 'tainted' RPM contains custom "external" dependencies. In the latter case, apt will not be able to figure out dependent RPMs without the user providing an additional repository. However 'tainted' RPMs are the fault of the packager - and it's the packager's responsibility for dependency checking -- not RPM.
Finally, many rpms cannot work interchangibly on different distros(i.e. SuSE and Redhat) or even across multiple vendor versions(i.e. RH9.0 rpm --> RH7.2). Who's fault is this: packager and rpmbuild for making package building too damn easy.
- A desktop offering hardware acceleration, scaling, blending, composition effects. There are promising extensions for X for this and they should be leapt on. But if means ditching X / a WM then so be it - QT / GTK are meant to be abstraction layers after all and X can run rootless on top of whatever-it-is if need be.
- A desktop where KDE / GNOME / dist homegrown tools are blended into a single cohesive entity. Not one with a generic KDE / GNOME slapped together with some weird tools (i.e. Mandrake). The desktop will be referred to as "the desktop" in all dialogs / apps and not KDE / GNOME / Drak / Yast etc. except in advanced documentation.
- A unified help system - one that offers one stop access to all man pages, info, html, READMEs, GNOME / KDE / dist help, all ordered in a task centric way with full search facilities.
- A desktop that offers to install additional apps (especially DVD, MP3 player etc.) in a user friendly manner click N run style during installation and at any time after. Even if there are legal reasons for not shipping MP3 on the CDs for example, the dist could still make it easy to find them remotely.
- A unified distribution neutral driver model with detection, installation / removal architecture. The situation at the moment with
getting a driver (or the hell of writing and supporting one) on Linux is a joke. Even a popular driver like NVidia involves screwing around at the command prompt and having a toolchain and kernel source if your dist is not directly supported.
- A unified theme engine. A single engine that any app, toolkit or WM can use to render buttons and decorations.
- Identification of every day operations and a UI to support them completely with no overlapping functionality. There should be no need for mere mortals to drop to the shell. Not even once. If OS X users can control a BSD derivative with a (single button) mouse then so can Linux. It doesn't mean Linux must be 'dumbed down', just that needless complication should be identified and removed / hidden from those who don't want to see it.
And most importantly:*rolls eyes*
No WONDER everyone seems to have this thing against Gentoo users. I think a lot of us get too caught up in our own distribution's "superiority" without remembering that the cool part of multiple distributions appeal to certain people. However tons of people seem to wish their particular distro will catch on and take the mainstream.
Frankly, as a Gentoo user, I don't ever want it to "take over" (and being source based I don't think it will). I like its niche and I like its community. Mass usage is going to kill that.
I don't think Gentoo is ever going to appeal to "big" corps or businesses. Small shops perhaps (I use it for all my operations) but the big guys? Nope. Corporations like dealing with corporations, it's that simple.
Conventional, but probably true answer: servers. There are still many companies running standard services like mail, web etc. on proprietary operating systems (Sun, Microsoft) in a time where it makes no whatsoever sense anymore. With kernel 2.6, Linux will gain acceptance as a high-end Unix replacement and be deployed wherever older server installations need to be replaced.
No. The desktop UI is still too inconsistent across KDE/Qt, Gnome/GTK, Mozilla/XUL and Openoffice and still offers no viable alternative to the commandline when it comes to system administration/configuration.
I predict that in 2004, attention will move away from KDE and Gnome as all-in-one-solutions. Instead, it will be finally accepted as reality among developers and users that different GUI APIs will continue to coexist, and that efforts should be made to standardize the protocols and user interfaces across the APIs. For the future of GNU/Linux and *BSD on the desktop, freedesktop.org will be much more important than kde.org and gnome.org, but it could take five-ten years until the difference between a KDE/Qt, GTK/Gnome, Tcl/Tk, Fltk program will be as irrelevant to users as the difference between a Carbon and a Cocoa app on MacOS X or as that between a Microsoft MFC program written in C++ or an OWL-based program written in Borland Delphi for a Windows user.
Once this level of standardization is reached, the importance of all-in-one desktops like KDE and Gnome could dramatically decrease, since users instead could combine components like taskbars, window managers, file managers and system menus at will. (Which, thanks to freedesktop.org, is already possible: fspanel / fbpanel / suxpanel / the xfce4 panel can be used as drop-in replacements for the Gnome panel, rox / xffm4 as drop-in replacements for Nautlius, and the list of freedesktop.org-compliant window managers suitable as replacements of metacity / kwin is endless.)
However, it will take yet another five years until 2013 or 2014 that a standardized Unix/GNU/Linux/BSD desktop will allow developers of system components (like sendmail/exim/Postfix/qmail, lpr/cups, Samba, Grub/Lilo...) to write GUI configuration panels for their own software. At the moment, desktop projects like KDE, Gnome/Ximian and Webmin can only provide insufficient configuration wrappers around low-level system tool; the only sane solution is that such GUI configuration panels are provided by the original component developers in sync with their release schedules, and will work consistently on any GUI configuration (as opposed to the present situation where a configuration panel would have to be provided in separate versions for KDE, Gnome, XFCE, webmin and what-have-you). Only at this point, GNU/Linux will be able to replace commercial end-user GUI operating systems on a large scale and be accessible to home users.
Contrary to what Eric S. Raymond says: The unclear situations of RedHat/Fedora and SuSE (after it has been bought by Novell) could create a strong push towards Debian as the standard binary (GNU/)Linux distribution. The Debian core distribution could become a de-facto-replacement of the disappointing "Linux Standards Base (LSB)", as more and more (commercial and community) distributions will be based Debian. Knoppix, Lindows and, in the near future, User Linux are prime examples. Debian itself will gain more acceptance in the mainstream and among new users as soon as it will ship with the new installer.
Given the record of Netscape/Mozilla's, StarOffice/OpenOffice's and Apple Darwin's transformation of corporate into public development projects, I doubt that RedHat/Fedora will ever become a true community project. It is also being overlooked that the equation RedHat=Linux is specific to the U.S. only
gopher://cramer.plaintext.cc http://cramer.plaintext.cc:70
A famous quote comes to mind:
i don't think it's a good idea to take out HOWTO's that are not considered neccesary for *most* people.
maybe things will autoconfigure in the (near) future, but for example the ethernet HOWTO you mention was very instrumental in getting two nics to work on an old box i transformed into a router for my DSL connection some years ago.
so they might not be of use to you, but they sure as hell are to thers. and it's not as if they're using up a lot of space as compared to your average modern linux desktop environment...
i do think some HOWTO's need serious updating...
I would think most of the theoretical advantages with compiling packages yourself is not from different -O settings, but because #if statements inside the code may completely eliminate parts of it at compile time. For instance compiling out support for some back-compatability thing you don't have or use may speed up a program, and will certainly make it smaller.
This is all in theory, however. I am unsure if there is any reason for it. If machines get fast enough that people don't mind a compile as part of install then I see no reason not to distribute source-only. But the only definate advantage is the fact that the thing you download is far more likely to work, I am unsure if there would be performance advantages.
As with everyone before me, I'm not saying Gentoo's bad, just that it is unlikely to catch on in large corporate scenarios.