Weekly Edition
Return to the Announcements page
| |||||||||||||
15 game-changing Linux moments of the decade (TechRadar)
In a retrospective, TechRadar looks at Linux innovations over the last decade.
"If you were sat at your Linux computer one dark evening in late 1999, things would have been considerably different.
[...]
Your machine would probably be running either Red Hat 6.1 or Mandrake 6.
[...]
Outside your window, the world was going crazy for all things dotcom. Microsoft was prepping both Windows 2000 and its ill-fated Millennium edition, while Apple had just released OS 9 and its Power Mac G4."
(Log in to post comments)
Cluelessness Posted Dec 29, 2009 1:49 UTC (Tue) by ncm (subscriber, #165) [Link] "SCO is currently fighting against bankruptcy": Meaning not, like, already filed, and in receivership? "Mandrake becomes Mandriva": Earthshaking! "Oracle buys Sun (MySQL)": Wow.The standard formula is 10 items. When you're obliged to inflate it out to 15, I suppose you have to reach deeper. There's no way anybody could come up with a list of 15 seriously important events in the Free Software world over the past ten years.
Cluelessness Posted Dec 29, 2009 3:26 UTC (Tue) by PaulWay (✭ supporter ✭, #45600) [Link]
It's pathetic, isn't it? Linux 2.6? Red Hat out-sells Microsoft? Creative Commons license launched? Ten years of Linux.conf.au? Linux gets USBv3 before Microsoft? OpenChange? SaMBa 4 with clustering and full AD support? Hello, anyone home?
Ah well, at least it's something.
Have fun,
Paul
But what about the good drivers? Posted Dec 29, 2009 4:55 UTC (Tue) by dmarti (subscriber, #11625) [Link] Why cover the one problem child NVIDIA driver, when there are thousands of other devices whose manufacturers are doing it right, or at least helping independent driver developers? Compare how much of the computer stuff you could buy 10 years ago and have work out of the box to shopping for hardware to run Linux on today. (Yes, we can thank the server and embedded markets for a lot of it, but hey, `find drivers/ -name *.c | wc -l` -- 5920!
But what about the good drivers? Posted Dec 29, 2009 11:16 UTC (Tue) by hppnq (guest, #14462) [Link] Why cover the one problem child NVIDIA driver, when there are thousands of other devices whose manufacturers are doing it right, or at least helping independent driver developers? For Linux rather then Free Software adoption, I think it is reasonable to mention the NVIDIA driver instead of drivers/frobnicator.c -- especially if you actually read the entry. What I found a bit more disturbing is the claim that NVIDIA is one of the few companies whose drivers "offer similar performance" across platforms. (This is not meant as an invitation for another round of ATI versus NVIDIA, but food for thought for the next decade, in which we *will* see large scale Free Software adoption. ;-)
But what about the good drivers? Posted Dec 29, 2009 17:09 UTC (Tue) by iabervon (subscriber, #722) [Link]
I don't think there are too many companies that actually offer competently-written drivers for any platform, and most companies don't write Linux drivers because they know they would be ignored in favor of better-written drivers developed elsewhere. The main exception I can think of is Intel. Hardware vendors just don't write good maintainable software, in general; they do better to outsource to the Linux kernel community (largely by making hardware that's worth driving and providing access to information about it). It's a different skill, and it's not that worthwhile to get it in-house, so they don't.
For that matter, I wouldn't be surprised if, before long, Nouveau had better performance than nVidia's drivers for either Windows or Linux, in terms of getting data and commands efficiently between applications and the graphics hardware; while nVidia does know how to write drivers for their hardware, they can't streamline the OS graphics pipeline for their convenience like X.org can. If that happens, I bet there will be a major shift in nVidia's strategy as the best performance you can get out of nVidia hardware is not from the newest hardware but from the newest hardware supported by Nouveau.
But what about the good drivers? Posted Dec 29, 2009 19:03 UTC (Tue) by drag (subscriber, #31333) [Link] The best thing that Nouveau can offer is a high level of consistency for the Linux platform. Having a predictable 'lowest common denominator' is critical if you want to make it possible for application and game developers (either open source or proprietary) to be able to support Linux in a effective manner. Having the ability to offer a constant level of OpenGL support (and other APIs) is just massively important.It terms of performance Nouveau has no hope of matching what Nvidia can offer, unfortunately. It is certainly possible to exceed Nvidia, but it's just extremely unlikely. Linux and the OSS graphic stack is just much too far behind what is offered by Nvidia proprietary drivers or what is available on Windows to really be competitive. To get a idea, in just pure 3D graphics, were we stand is the following: * OpenGL 3.2 is the current standard for OGL graphics. API standard released Dec 2009. * DirectX 10 is roughly equivalent in terms of features and hardware support with OpenGL 3.2. DX10 was released for production in Nov 2006. * OpenGL 2.1 was released in July 2006. This is what Mesa 7.x OpenGL stack is able to offer. (but only in software rendering) Mesa with OpenGL 2.1 support was released in 2007. * OpenGL 1.4 was released in July 2002 * OpenGL 1.3 was released in Aug 2001. Linux OSS graphic stack is able to offer hardware rendering support for OpenGL 1.3 or 1.4 or so. This the current state of the art in Linux. Now with Nvidia's proprietary drivers they not only shovel huge amounts of code into the kernel, they also replace large parts of Xorg's functionality and have their own proprietary OpenGL stack. Nvidia is able to provide Linux users with the full OpenGL 3.2 support; plus whatever extensions Nvidia offers. Now in terms of the level of efficiency that Nouveau may offer in getting OGL1.3 commands relayed to the hardware it may or may not be able to perform better then the proprietary drivers... I do not have a clue. But the problem is that as soon as a application requests operations not supported by the driver then you'll, at best, run into massive slowdowns, or worst, simply lose all compatibility with the application. HOWEVER if OSS graphic support is able to provide a modernized graphic stack capable of supporting OpenGL 2.1 in a _consistent_ and highly reliable manner regardless of the hardware then that would be a HUGE win. It still would not be up to what is offered by Windows, but it will still be a huge step forward. Meaning you can expect a application to be compatible and perform acceptably with Intel, ATI, AND Nvidia hardware. Currently developing 3D graphics intensive applications for Linux is nearly impossible to do correctly unless you are willing to ignore OSS stuff and require people to use proprietary drivers. Simply because support is so spotty otherwise. As I understand it, one of the major reasons application developers and game designers began favoring DirectX over OpenGL early on was the ability for Microsoft to provide a reliable and consistent 3D platform. In comparison each hardware vendor had to have it's own seperate proprietary OpenGL stack. This meant that if you were programming OpenGL you had to deal with the overhead of supporting multiple OpenGL implementations.. each with their own bugs, limitations, and advantages. Were as with DirectX Microsoft provided only one platform they needed to worry about. The quality of API and features that DirectX offered over OpenGL (and visa versa) mattered much less then that sort of thing. So that is the most important thing.. if the OSS OpenGL protocol stack can provide a 3D platform that is highly reliable and consistent across all different sorts of hardware then that would be a something that would be a huge win. It still would not be really competitive until it's up to feature parity with proprietary stuff, but it will still be a hell of a lot better off then what we have now. If the Xorg/Mesa/Linux folks can get up to that level then it will probably get the attention of Nvidia and get them to turn around somewhat. But I doubt they will abandon their primary focus on proprietary graphics in any sort of foreseeable future.
But what about the good drivers? Posted Dec 30, 2009 12:11 UTC (Wed) by tdz (subscriber, #58733) [Link]
Hi
> Linux OSS graphic stack is able to offer hardware rendering support for
In my notebook, I have an Intel 945 IGP, which only supports OGL 1.4! Hardware shaders are not available, so OpenGL 2.1 will never be possible on this chipset. Not even with proprietary drivers on Windows. And I guess that any software emulation of shaders is so slow that it would be almost useless.
I recently took a detailed look at the kernel's DRM subsystem. I wouldn't consider the code to be of really high quality, but this seems only a matter of time. Then I guess its only a question of adding support for new chipsets. I also own a recent AMD graphics card with an R700 chip. I guess this will be supported with OpenGL 3.2 within the next two years or so.
> It still would not be up to what is offered by Windows, but it will
No. Microsoft simply solves this problem by defining what each version of DirectX provides, similar to what Khronos does for OpenGL. The gamers then go and buy graphics cards that support the DirectX version needed by their games.
Regards, Thomas
Cluelessness Posted Dec 29, 2009 5:00 UTC (Tue) by maro (subscriber, #34315) [Link]
> SAFE SUN: Linux can freely take from any of Sun's open source
> projects, including its OpenSolaris operating system
Heh, and earlier he wrote that the kernel is the only component officially
Cluelessness Posted Dec 29, 2009 5:22 UTC (Tue) by ncm (subscriber, #165) [Link] Well, yeah. The rest is GNU. At least he got that part right.Right now Graham is calling us a bunch of whiny losers. He didn't want that assignment anyway. He texted it from his iPhone while he was waiting for Avatar to start. TechRadar pay stinks.
Cluelessness Posted Dec 29, 2009 8:29 UTC (Tue) by maro (subscriber, #34315) [Link]
I was questioning the credit given to Sun for allowing Linux (which he
already defined as being the kernel) to "freely take from any of Sun's open source projects, including its OpenSolaris operating system"
Cluelessness Posted Dec 29, 2009 10:41 UTC (Tue) by djzort (guest, #57189) [Link]
especially how much solaris has benefited from the mountain of open source software that actually makes it useful
gnu tools
just to name a few that spring to mind.
Cluelessness Posted Dec 29, 2009 14:40 UTC (Tue) by armijn (subscriber, #3653) [Link]
There are many Linux systems (mostly embedded) where the rest is *not* GNU. Suer, it's often GPL or LGPL licensed, but that does not make it GNU.
Cluelessness Posted Dec 29, 2009 19:32 UTC (Tue) by dennisk (guest, #12308) [Link]
I bet we can come up with 15 major events since 2000. My criteria is that the event will have (or has already had) a long term effect on the adoption of Free Software.
I'll start the list:
1. GPLv3
Dennisk
My game-changing Linux moment of the decade Posted Dec 29, 2009 11:04 UTC (Tue) by NAR (subscriber, #1313) [Link]
The game-changing Linux moment of the decade came for me two days ago when I finally removed Linux from my home computer. I've tried to install Fedora 12 to see if it finally supports my two year old videocard. In my 13 years of Linux experience, this was the first installation that managed to break my Windows XP, I couldn't boot that OS after. There was no sound (thanks to pulseaudio), xawtv was missing some fonts from its dependency list (according to Google search results, this is not a new bug) and couldn't handle my TV card anyway. Interestingly I was able to use the very same TV card in the last decade with each Linux distribution I tried. I haven't got that far to actually check if 3D works with my ATI card. Good bye, Linux.
My game-changing Linux moment of the decade Posted Dec 29, 2009 11:15 UTC (Tue) by Linux. (guest, #62725) [Link]
Good bye, NAR.
My game-changing Linux moment of the decade Posted Dec 29, 2009 13:08 UTC (Tue) by JoeF (guest, #4486) [Link]
So, you blame your experience with one distribution on all of Linux?
It seems you haven't learned much about Linux in the 13 years of experience that you claim to have... As for the distro killing XP, all I can say is PEBKAC.
My game-changing Linux moment of the decade Posted Dec 29, 2009 14:25 UTC (Tue) by shalem (subscriber, #4062) [Link]
Actually, not a PEBKAC, but a problem with Fedora's installer anaconda,
we had a bug where we set the active flag on /boot, which causes the active flag for the XP partition to be removed, which it does not like.
This has been fixed now, but in F-12 we have this issue.
Regards,
Hans de Goede <anaconda developer>
My game-changing Linux moment of the decade Posted Dec 29, 2009 20:57 UTC (Tue) by JoeF (guest, #4486) [Link]
XP doesn't need the active flag to be set for it's partition.
I am using my own multi-OS boot loader, written some 20 years ago, which has the active flag set for it's own partition. I've thrown a bunch of operating systems at it over the years, and the only one I've found that expects it's own partition to have the active bit set is SCO Unix. Of course, if you remove the active partition, and no other partition is set to active, nothing would boot. But that has nothing to do with XP, the same would happen for any other OS (in fact, it did when I replaced XP with Linux on my netbook, and forgot to set the active flag.)
My game-changing Linux moment of the decade Posted Dec 30, 2009 6:45 UTC (Wed) by shalem (subscriber, #4062) [Link]
XP does need the active flag set when people choose to use the XP bootloader as the primary one and chain grub from it.
My game-changing Linux moment of the decade Posted Dec 30, 2009 10:48 UTC (Wed) by JoeF (guest, #4486) [Link]
The master boot loader needs the active flag set to know which partition to boot.
But again, that has nothing to do with XP per se. Using lilo as master boot loader works the same way. If grub ignores the active flag, that is a feature of grub, then.
Active flag Posted Jan 9, 2010 19:49 UTC (Sat) by anton (guest, #25547) [Link] Why should lilo need to have the active flag set to know which partition to boot? My current lilo setup has the active flag set on a partition other than the boot partition, so it's obviously not needed for that. And I dimly remember running a system with no active flag set; that worked on some box, but on another box the BIOS did not want to boot unless the active flag was set on some partition (no matter which one).
Critical views of the decade Posted Dec 29, 2009 13:22 UTC (Tue) by man_ls (subscriber, #15091) [Link] Your insightful comments in the past show that you are not a troll, although your message has a certain ring to it. To me it is clear that we (as in the community of Free software users and developers) have to learn to accept criticism even in its less constructive forms. Critical views make us learn new things and adapt to changing circumstances.However, I have to say I fail to see your point. So Fedora 12 does not work for you. Have you tried other distros? In the past Linux users were used to having to wrestle with the hardware for a little while before everything was working perfectly. Especially when using one of the more dev-oriented distros, like Fedora. In short, we accepted a little inconvenience for having something that we could modify and even rewrite to our liking. It is true that most of us have less time to tinker, having got (more demanding) jobs and family in the meantime. That is probably why we have gladly welcomed more user-friendly distros. But sometimes it still takes a little patience to get things working properly. In fact, even die-hard Windows users have to ocassionally download one or two drivers to make hardware work properly; only Apple users have become spoiled enough to demand that everything works perfectly, or return the computer, and they are willing to pay for the privilege. Demanding the same from Windows or Linux, while being unreasonably sensible, is an endless road to frustration IMHO.
Critical views of the decade Posted Dec 29, 2009 16:36 UTC (Tue) by NAR (subscriber, #1313) [Link] Earlier this year I've tried a newer Ubuntu, but there were still no sound and no 3D. My point is that after 10 years, Linux on the desktop hasn't improved, it's still "let's hope that this stuff works on this particular combination of hardware" - and unfortunately the problematic hardware has about 50% market share (if we add Intel's problems, let's make that 97%). I've forgot to mention that I've also got three kernel oops-es during the two hours I was using Fedora, which isn't that reassuring either, but at least I could send the bug report with a single click. I've tried to report the xawtv dependency bug, but Fedora required to register to their bugzilla (apparently they haven't read about the The $300 Million Button), so I bailed out.In short, we accepted a little inconvenience for having something that we could modify and even rewrite to our liking. - the thing is, if you target this audience, you'll get only this audience. That won't lead to world dominance. And this is not related to Fedora - Ubuntu didn't break the XP installation, but didn't support the hardware either, so that's useless too.
Critical views of the decade Posted Dec 29, 2009 18:06 UTC (Tue) by robla (subscriber, #424) [Link] man_ls: In short, we accepted a little inconvenience for having something that we could modify and even rewrite to our liking.NAR: the thing is, if you target this audience, you'll get only this audience. That won't lead to world dominance. I don't think it's a question of targeting a particular audience. There has been a ton of work on fire-and-forget installations, and things have generally gotten better for me over the past decade (your mileage has obviously varied). Good 3D support and good sound support are two areas that have been severely hampered by indifferent vendors with key technology (e.g. video drivers, codecs, other binary widgets), along with some self-inflicted damage on the part of Linux desktop developer community. The vendors haven't been forced to care by the consumer market, thanks to Linux's traditionally anemic desktop marketshare. The token effort they have made has been squandered by ever changing APIs, ABIs, which further reinforces the view that desktop Linux isn't ready for primetime. It's a vicious cycle, because these problems need to be solved before many vendors stop being indifferent. It's taken way longer than I would have thought to (still not) solve these problems. In spite of the sluggish progress, I have a hard time believing this situation is going to persist. I'm pretty optimistic that there's going to be some effort out there (e.g. Chrome, Moblin, Maemo or something) that busts into the mainstream (forcing vendors to care) and provides a more focused target for vendors to target. Additionally, I'm even enough of a dreamer to think that much of the proprietary widgetry from indifferent vendors is headed for the dustbin of history. However, I'm enough of a realist to realize that many people don't share that view, and won't believe it until they see it. So, I can totally understand NAR's lack of patience with the process. I'm sticking with it for the foreseeable future, but we can't expect that sort of foolish persistence from everyone.
Critical views of the decade Posted Dec 29, 2009 18:07 UTC (Tue) by man_ls (subscriber, #15091) [Link] I don't have the perspective of a decade because I am practically a newcomer, but since 2003 the situation has improved considerably. Now it is entirely possible to buy a computer where Linux just works, even a laptop, and (seen from the other side) most hardware just works too -- apart from a few high-profile fiascos. Hardware vendors are aware of Linux and many do not just offer binary afterthought compromises but good source support. For those fiascos we can take the reasonable position and buy our way out of the mess (with [god forbid] Windows or Mac OS X), or we can be unreasonable and try to change the world.I prefer the easy, cheap way out and buy supported hardware, something orders of magnitude easier now than 7 years ago. I even have reasonable 3D on my Asus EeePC 1000H!
Critical views of the decade Posted Dec 29, 2009 19:40 UTC (Tue) by drag (subscriber, #31333) [Link]
> I've tried to report the xawtv dependency bug, but Fedora required to
register to their bugzilla (apparently they haven't read about the The $300 Million Button), so I bailed out. <br><br> Don't worry. They probably would of just ignored you anyways. Fedora is a playground operating system. I don't think it's really meant to be used by mere mortals or people that have not been using it for years and years already. <br><br>
-------------------
Distros served a important service in the past, but nowadays, at least with
But even though they are all using the same software somehow they manage to
There are small advantages. Like I like how Debian's apt-get and network
The old Linux user advice of "Have you tried a different distro?" is so
That is all kinda of insane if you think about it. A example:
Look... If NAR was using Debian and he was complaining that a old ATI card
But if he is using Fedora and Ubuntu then what do I tell him? I'd tell him
I know the probable answer in Debian because I have already gone though
Now it's not bad that the software is not perfect. Well it's bad, but it's
Choice is generally good, but it's only good if there are really really
I like Linux, but after a while these sorts of issues DO get tiresome.
Critical views of the decade Posted Dec 29, 2009 19:40 UTC (Tue) by drag (subscriber, #31333) [Link]
Sorry. I accidentally hit 'post' instead of 'preview'. If anybody can fix
that I would be happy.
Critical views of the decade Posted Dec 30, 2009 6:11 UTC (Wed) by eru (subscriber, #2753) [Link] The old Linux user advice of "Have you tried a different distro?" is so horrible it's pathetic. I mean it's just really really really bad. How does it make sense that in order to get a old BTTV video capture device working with a old application like Xawtv is it necessary to try out a different operating system with slightly different configuration that is built using almost all the same source code "just to see if it works"?I see your point, but in this case, If a person is unhappy with a bleeding-edge distribution like Fedora breaking things, suggesting a slower-moving distribution is appropriate. Most of the pain occurs during upgrades, which Fedora more or less forces you to do about every year, if you want to stay current with security. One can install Linux and forget about it for years, if one uses CentOS or similar and makes no major hardware changes. OS upgrades are an annoying thing, whether Linux or Windows. Thinking back, it is actually very rare that they bring big improvements. Usually the machine just feels more sluggish after the upgrade, and random things break until subsequent patches fix things. The main reason one does it is that some new software that one really wants to use is not compatible with the old OS version (or would require retrofitting tons of updated libraries). (When depressed, I often start thinking that one half of the software developers just create problems for the other half to solve, and vice versa, without much genuinely useful work coming out, and that the world might actually be better off without all this computer stuff...)
Critical views of the decade Posted Dec 30, 2009 17:22 UTC (Wed) by tjc (subscriber, #137) [Link] What makes it hellish in Linux is that different distros manage to take the same software and make it broken and hard to use in _different_ways_.Good point. Apple is in the best position: They control both the OS and the hardware it runs on. Their list of supported hardware gets pretty long at times, but they periodically discontinue support for both older hardware and operating systems, so they remain bounded on both sides. Microsoft has to deal with a much wider range of hardware, which is one of the reasons Windows is more trouble than OS X. Linux is more or less unbounded on both the hardware and software sides, which is more trouble still. Ten years ago I was running Red Hat 6.1 with the mid-release "October Gnome" updates, and things had become fairly stable, although still troublesome at times. I'm currently running Ubuntu 9.10, and I can't really say that's it's any better than what I was using 10 years ago from a stability standpoint. Weird things still happen, and I'm less likely to be able to fix them, since the entire system is now an order of magnitude more complex than it was 10 years ago. The problems I can fix seem to take longer. It could be that this is as good as it gets using an unbounded software/hardware model. We may have passed the peak and be moving downhill. It might be beneficial for someone to create a hardware-specific Linux distro, using the Apple model of supporting limited hardware instead of the Microsoft model of supporting "everything." This has been done with systems like the Asus Eee PC, but as far as I know it's never been done with the goal of creating a distro that "just works" on a limited range of popular hardware.
Critical views of the decade Posted Dec 31, 2009 9:16 UTC (Thu) by drag (subscriber, #31333) [Link]
The hardware support is relatively easy compared to the software issues
different distros introduce higher up. Really, getting the hardware working is minor compared to providing a maintainable system that allows for very good application compatibility!
It's just that most people that hit problems in Linux never get to that
That is sort of what happened with the EEEPC. Asus was able to remove all
Now Ubuntu does a lot better job on the userland side then Asus/Xandros
Which is why now most netbooks are sold with windows on them despite the
This is sad, but unfortunately it is the reality we live in.
Critical views of the decade Posted Jan 2, 2010 18:20 UTC (Sat) by RobSeace (subscriber, #4435) [Link] That is sort of what happened with the EEEPC. Asus was able to remove all the hardware issues and then create a system that just worked for a lot of people. However 3-4 clicks into actually using any portion of the operating system everything turned to shit for a lot of people. OpenOffice.org is slow and it was not a really good version. Online videos and people's favorite web games failed to work and all sorts of bad things.I bought a EeePC (one of the Linux ones, of course) for my grandmother this summer, and she hasn't run into any such problems with it, or found anything to complain about (other than the cable Internet connection occassionally going down, which has nothing to do with the EeePC)... And, I know she watches online videos on various web sites, and people send her videos and games and junk via E-mail, and she never has any problems with any of them... (Other than learning how to drag and drop with the trackpad, because she strangely resists using an actual mouse for some reason beyond my comprehension...) And, she regularly uses the included OpenOffice (or is really StarOffice?) without any complaints, as well... So, I don't think it's quite as horrible an environment for everyone as you make it out to be... If I had to use it, well yes I'd absolutely hate it, and would install something much more powerful on it instead... But, for her, and others like her, it seems just about perfect to me... *shrug* Which is why now most netbooks are sold with windows on them despite the 100-200 dollars increase in premiums that go along with the increase in hardware and licensing costs. This is sad, but unfortunately it is the reality we live in. It may not be as sad or real as MS would like you to believe: Linux owns 32 percent of netbook market, says study
Critical views of the decade Posted Jan 4, 2010 13:40 UTC (Mon) by NAR (subscriber, #1313) [Link]
A friend of mine bought some netbook with Ubuntu 8.04 preinstalled. It worked fine except one small problem: the pidgin version in that Ubuntu couldn't connect to MSN (as far as I know, there was a protocol change in MSN). The "Linux-way" of fixing this is to either compile a newer version of pidgin from source (for a non-technical newbie this is a no-go, he didn't even now where to find the terminal on the UI) or to upgrade the whole distribution, newer Ubuntus have newer pidgin, but this idea was met with a "you can't be serious" exclamation.
Actually, on the pidgin site there were some instructions on how to install a new binary package for Ubuntu. It involved the copying and pasting of two longish command lines, but then pidgin worked mostly fine (this time connecting to gtalk didn't work, but that could be a configuration problem). So maybe there's hope for Linux, if application developers take the time to create such instructions and create and maintain separate packages for the supported distributions.
IM is a weak point Posted Jan 4, 2010 16:33 UTC (Mon) by Cato (subscriber, #7643) [Link]
I had a similar issue with Kopete in Ubuntu 8.04 (KDE 3.5.10) and Yahoo instant messaging - unfortunately there was no fixed version except for later Ubuntu versions and KDE 4.x, and building Kopete from KDE 4.3.x for Ubuntu 8.04 took (literally) several days including learning how to use kdebuild and Cmake, and building Qt4. I finally gave up when I realised that the new Kopete didn't work with Yahoo emoticons...
So I switched to Google Chat (based on Jabber) still using Kopete, which was OK for me but wouldn't have worked for most people.
Until IM becomes really standards based, Linux will continue to have this problem - unlike hardware support, there don't seem to be paid developers working on IM clients, so support for this sort of issue is fairly slow (months not days) or non-existent.
Critical views of the decade Posted Jan 3, 2010 19:58 UTC (Sun) by nicku (subscriber, #777) [Link] I would tell him that all he probably has to do is install a package that contained 'non-free' firmware for the kernel and that would get the ATI 3D acceleration working.To my surprise, with Fedora 12, I found 3D support by the free radeon driver for my cheap, new $AU75 RV710 [Radeon HD 4550] to be excellent (I installed mesa-dri-drivers-experimental-7.6-0.13.fc12.x86_64). mesa-dri-drivers-7.6-0.13.fc12.i686 supports 3D on the RV350 AS [Radeon 9550] on my son's computer very well.
My thoughts for Linux at the end of this decade include thankfulness for the great efforts to bring 3D support to free software.
Critical views of the decade Posted Dec 29, 2009 20:45 UTC (Tue) by cmccabe (guest, #60281) [Link]
> Earlier this year I've tried a newer Ubuntu, but there were still
> no sound and no 3D
*sigh*. Periodically someone posts an indignant message about how installing Linux is hard. This is considered Deep Wisdom and we're all expected to agree. If only Linux were easy to install, it is implied, it would take over the world.
This is BS. No operating system is easy to install. You have to hunt down driver disks (or open source support websites), fool with upgrade packs, and deal with software reinstallation. Installing a new version of Windows, or a copy of Linux, is not a task you will attempt unless you are a Computer Person.
Microsoft has been struggling with this problem for years. People are not enthusiastic about reinstalling their operating system. People rarely buy a new copy of Windows for a machine that already has it, no matter how many shiny new features the company dangles.
The key to the consumer market is two words: preinstalled software. Linux will never be the majority operating system until it's preinstalled.
If you look at the parts of the consumer market where Linux has been successful so far, like cell phones and set-top boxes, they are also cases where it has been preinstalled. (Linux has also been successful in the server market, but that is a different market.) Luckily, there are some projects on the horizon, like Android and Chromium, that will help Linux get traction in the consumer market.
P.S. It's always worthwhile to try to make Linux installs easier. Canonical has done a lot of good work in that direction. (although I don't agree with all of their policies.) I just think it's unrealistic to expect a one-click install.
Critical views of the decade Posted Dec 29, 2009 22:49 UTC (Tue) by rahvin (subscriber, #16953) [Link]
The only way a one click install works is if you are Apple. You can do the one click install if you can control all the hardware used. The Modern PC is a hodgepodge of millions of different peripheral hardware devices and finding the software and drivers for those millions of items is the problem. The solution is the Apple way, produce and license the only hardware that runs on the system that way every OS install has every driver built in. Otherwise the answer is that some stuff doesn't work with certain configurations.
The original complaint (especially the oops's) indicates to me there is a hardware problem, maybe memory with a few bad pages that does ok in windows but when confronted with the Linux kernel that is intolerant of lost or corrupt memory pages it does the right thing and oops's instead of pretending to work and corrupting data. Hey I'll admit I'm primarily a windows desktop user, but when I'm troubleshooting hardware it's on linux because I get real information back, and a Kernel oops, outside blatent misconfiguration, is frequently tied to broken hardware that's still mostly functional so that it fools windows.
Critical views of the decade Posted Dec 30, 2009 5:58 UTC (Wed) by alankila (subscriber, #47141) [Link]
I think NAR was complaining about regressions more than missing support for this or that. Notice that in his long list of issues were things that used to work, but no longer do. As I read it, the only thing that apparently never worked was 3D, thanks to going with ATI hardware.
Incidentally, I've had the reverse experience with ATI: at one point it was driven by fglrx and worked well enough, but then fglrx dropped support and I got stuck with the open-source driver, which was crap for a year and now it again shows signs of working. Awesome, I guess, but the laptop is now 2 years old and for half of the time I've had to suffer glitchy video.
Regressions feel really bad. They shake the confidence in the mantra that open source is improving and will get there eventually, if it isn't quite there yet. (And it isn't.)
Critical views of the decade Posted Jan 6, 2010 21:17 UTC (Wed) by nix (subscriber, #2304) [Link]
Yeah, but at the same time as your two-year-old laptop showed signs of
working, my then-leading-edge 4870 *also* showed signs of working (I'm using KMS with it now, with 3D working fine for most things I try it with, including things like scorched3d which failed with *every* previous 3D-capable card I've ever owned). All of ATIs non-grossly-obsolete cards seem to be emerging into the free software light at the same time, thanks to airlie and daenzer and the rest of the fine graphics-hacking crew.
Critical views of the decade Posted Dec 30, 2009 10:14 UTC (Wed) by djao (guest, #4263) [Link] In short, we accepted a little inconvenience for having something that we could modify and even rewrite to our liking. - the thing is, if you target this audience, you'll get only this audience. That won't lead to world dominance I'd like to offer a different response to this often-cited truism. Linux (and, to a lesser extent, BSD) is the only OS in the world that can thrive without achieving world dominance. This fact is politically incorrect, but it's true. It's also nearly impossible for a non-Linux user to understand, because non-Linux users have no experience with anything other than commercially developed platforms, and Linux is nothing like a commercial platform. On a commercially developed platform such as Windows or Mac OS, the original manufacturer is the only source of further development. The financial health of the original manufacturer is directly proportional to the size of its userbase. Without users, the platform has no money. Without money, the platform dies. That's why Windows and Macintosh users focus so heavily on achieving world domination -- the financial survival of their computing platform literally depends on the number of users that it has. In short, Windows and Mac OS need to achieve popularity, upon pain of death. Linux is very very different. As indicated in the original quote, Linux (and BSD) is distinguished by the fact that any user can improve the code. Linux is not dependent on any single corporate patron for its development. Popularity affects Linux and BSD only to the extent that it helps attract additional volunteer contributions. In addition to this general principle, which holds for any open source project, the minimum level of popularity required for Linux's survival is even lower than for open source projects aimed at a wider audience (such as web browsers), because the nature of an OS is that the more skilled you are as a developer, the more attractive the Linux platform becomes. Since Linux users are, on average, highly skilled developers compared to average Windows or Mac users, even a few of these highly skilled Linux users is enough to ensure that the platform stays technically competitive The combination of open-source development and a highly skilled userbase means that, for the small proportion of computer users who are highly skilled, Linux will always be able to stay in front of proprietary platforms such as Windows. No world domination is required.
Critical views of the decade Posted Dec 30, 2009 22:53 UTC (Wed) by anselm (subscriber, #2796) [Link] »World domination« hasn't been in Apple's dictionary ever since the Apple II stopped being popular. (Disregarding the iPod division, anyway.) But even Apple doesn't need »world domination« to keep MacOS going, for two reasons:
The other thing to consider is that Apple, unlike Microsoft or the Linux community, is really a hardware company. MacOS exists because Apple boxes need an operating system, and because Apple needs a reason for people to buy their overpriced (if sleek-looking) PCs. If Apple boxes just came with Windows, or if MacOS ran on generic PCs rather than only Apple-branded ones, then Apple would basically be just another Dell or HP. So essentially everything Apple does as far as software is concerned, they do for a single purpose: selling more Apple boxes, which is how Apple makes money. So, the market share of MacOS is directly tied to the number of Apple boxes sold. Every Apple box leaves the factory with MacOS already on it, and except for a minuscule fraction of machines it never gets replaced by something completely different, because in that case why pay for overpriced Apple hardware in the first place (it is good but not quite that good)? Microsoft and the Linux community (including Red Hat, Novell, etc.), on the other hand, compete for machines manufactured by all the other PC makers, who don't bother rolling their own operating system. Hence every PC more preloaded with Linux (or indeed not preloaded with anything) means one PC less preloaded with Windows, an idea that is thoroughly distasteful to Microsoft. We have already established that Microsoft can't really do without Apple, and MacOS doesn't eat into Microsoft's sales on non-Apple boxes, anyway, so the best thing for them right now is to just let Apple do their thing, which at 5% or so market share isn't that big a problem as far as Microsoft is concerned. Linux, on the other hand, has the potential (even if it is rather theoretical today) to replace Windows on a sizeable fraction of generic PCs, and that for Microsoft is a big problem. Thus, Apple doesn't need »world domination« any more than Linux does. They just need enough people who think that Apple computers are great so they can turn a profit (which as long as they keep selling lots of iPods and iPhones isn't all that many, anyway), and as long as they don't tread on Microsoft's toes too badly everything is just peachy for them. They probably wouldn't mind a 10% market share instead of 5%, but I don't think they would really know what to do with anything more than 20% or so, and Microsoft wouldn't be happy either.
Critical views of the decade Posted Dec 31, 2009 9:47 UTC (Thu) by djao (guest, #4263) [Link] They've been chugging along at 5% market share for two decades or soApple's market share varies depending on who you ask, but the consensus seems to be that in the late 90's, when Microsoft had to bail them out, the Macintosh market share was in the high 2%'s, whereas lately Apple's PC market share has been close to 10%. In no way has the Macintosh market share been a constant 5% for the past two decades. The historical record indicates that Apple needs a minimum of 5% market share to survive and maybe 8-10% to thrive. These numbers are perhaps short of "world domination", but they do suggest that popularity is a lot more important for Apple's survival than for Linux. I would not count on the Microsoft bailout being repeated if Apple's market share were to fall back to low levels. At the time of the bailout, Microsoft was facing an antitrust lawsuit and an aggressively anti-monopoly presidential administration. Yes, it's possible that the same confluence of factors will repeat itself, but Apple's shareholders and customers would be remiss if they were to rely on it, as you suggest.
Critical views of the decade Posted Dec 31, 2009 9:52 UTC (Thu) by drag (subscriber, #31333) [Link]
All Microsoft would have to do to kill off Apple is to pull MS Office for
Apple out of the marketplace. Any sort of positive growth that Apple has in the PC market would be completely gone in a matter of a couple months.
The whole PC vs Apple thing is just complete marketing bullshit. Microsoft is
Apple survives because they are not a threat to Microsoft. Linux survives
Critical views of the decade Posted Dec 31, 2009 13:45 UTC (Thu) by magnus (subscriber, #34778) [Link] The combination of open-source development and a highly skilled userbase means that, for the small proportion of computer users who are highly skilled, Linux will always be able to stay in front of proprietary platforms such as Windows. No world domination is required.True, but you also have to consider interoperability. Without a large enough user base, closed or semi-closed formats/protocols can prevent us from using our OS of choice for everyday use. MS Office and Flash can still today cause problems. In a way, having a skilled userbase can be negative, because the skilled users silently find ways around these problems (Wine, dual-booting, various hacks) so they never get solved properly. What is needed is a non-skilled (preferably rich) userbase that can complain loudly and force companies to find solutions to avoid losing revenue. For that reason, I really think Firefox has been very important. Because they got a large enough market share on Windows, web site owners actually care about compatibility. If it hadn't existed 4-5 years ago, I probably would have been forced to dual-boot to do online shopping etc even if it had existed as it is today.
My game-changing Linux moment of the decade Posted Dec 29, 2009 16:28 UTC (Tue) by tjc (subscriber, #137) [Link] I understand your frustration, but unfortunately there aren't any real alternatives. OS X can match Linux in it's ability to confound users, and Windows can easily exceed it. I find it best to stay with the trouble I know. The problem in all these cases is software complexity. I could go and on about that, but I won't, since it usually falls on deaf ears. The fact that Snow Leopard and Windows 7 are both we-have-to-clean-up-this-mess-because-people-won't-pay-for-it-anymore releases gives me hope. Hopefully Linux distributors will take note of this trend.
My game-changing Linux moment of the decade Posted Dec 29, 2009 20:25 UTC (Tue) by Los__D (guest, #15263) [Link] In my 13 years of Linux experience, this was the first installation that managed to break my Windows XP, I couldn't boot that OS after. Featurenotbug ;)
|
|||||||||||||