Submitted by Ubuntu patron and founder, Mark Shuttleworth, has detailed plans for the distribution’s April 2010 release, codenamed Lucid Lynx. Citing the Lynx deodorant namesake, Shuttleworth said: “This year’s Ubuntu Developer Summit is going to be the sweetest smelling Ubuntu Developer Summit ever, despite the fact that it’s happening in Dallas.”
That is really weird what they’re claiming in the article – tonnes of typos as well. If they did release a distro based on GNOME 3.0 after 2010.4 then it would be out of date by the time the end user received it. 2.30 is 3.0 – so if there is going to be a 3.0 it’ll be in 2010.4.
now microsoft has 3 cats to worry about
1. a Snow Leopard
2. a Lucid Lynx
3. and a Cougar http://www.neowin.net/news/main/09/09/20/big-cat-on-microsoft-campu…
This sounds like the beginings of a FOX tv program from the 90’s. Something along the lines of “slow children, savage cats: wars of the wild”
Actually, it was on Tru TV. It was called “World’s Most Daring Releases”. The episode I saw was about the Windows ME release. It showed hundreds of fans chasing down Geek Squad employees for recommending they install it. They were beating them with leftover floppies and yelling “Abort, Retry, Fail!” I know I was pretty scared when I watched it.
Remember, it’s not Reality, it Actual-ish-ness!
hahaha that just made my day. Thank you for that
I did some quick googling. Apparently, Lynx is the same product as Axe here in the States.
Excellent, we need more women running Linux!
Yeah, so RMS can emacs-deflore then.
Edited 2009-09-21 20:58 UTC
And if that’s not a fate worse than death I’m not sure what is.
Or nano-deflower them: its not the size, its how you use it.
apt-get remove virginity
Unable to lock the administration directory (/var/lib/dflower/), are you rms?
If Axe actually works like the commercials*, there probably isn’t enough in the known universe for it to work for RMS.
My girlfriend “loves” axe. She says it helps her avoid arrogant assholes.
This will probably be the distro to run with if you use Gnome. It’s to Gnome what 8.04 was to KDE, which is excellent news for someone wanting a solid LTS release.
The Xorg driver catastrophe of Jaunty will hopefully also be a thing of the past.
This will also probably be the distro to run with if you use KDE. Kubuntu 10.4 will be the first LTS release which includes KDE 4.x (it will have KDE 4.4).
Given that kernel 2.6.32 is expected to be released in 10 to 12 weeks time, then far as Xorg drivers goes, this will also be the first *buntu release (for any desktop) that will include open source 3D-capable drivers for ATI as well as Intel graphics.
http://www.h-online.com/open/Kernel-Log-Devtmpfs-in-2-6-32-more-dis…
Apart from the quantum leap improvement in xorg, it looks as though in this upcoming LTS release (10.4) there will also be improvements in wireless drivers, in the scheduler, and also in virtualisation support for Windows.
Edited 2009-09-22 01:42 UTC
As you well know, I’m definitely a defender of X and Ubuntu. But “Quantum Leap” and “Xorg” should never be used in the same sentence. Xorg progress is almost always^W^Walways a two steps forward, one step back affair. And often the pace seems glacial.
Fortunately (for me) the core features needed by business desktops are mature and stable. Mainly it’s the “Distracting Eye-Candy” and “Warriors of Goo!” crowds who are affected. (And of course, users of DE’s which bet the farm, and their entire infrastructure, on the distracting eye candy features, and did it too early.)
Expect as much pain from the OSS ATI drivers as with fglrx. At least for a while.
Edited 2009-09-22 02:33 UTC
I’m already running the OSS ATI drivers, and I have been for about a year or so. They are very stable and work just fine, and they are installed out of the box … the only thing they currently lack is 3D graphics hardware acceleration.
Kernel 2.6.32 does not introduce new OSS drivers for ATI, it only expands the functionality of existing stable drivers to include hardware acceleration for 3D graphics features. If you want a stable desktop, you can turn those features off if you wish. As you correctly point out: “the core features needed by business desktops are mature and stable”. This statement is just as true for OSS ATI drivers as it is for any other part of the OSS graphics stack.
Edited 2009-09-22 02:55 UTC
As we keep speaking of “video drivers” and “kernel” in the same sentence, it’s probably worth explicitly mentioning that only the bits that absolutely need to be in the kernel are in the kernel.
While certain kernel devs have no problem with belittling Xorg devs and their code quality, the kernel devs have consistently balked at the idea of taking on the formidable tasks which the Xorg team has tackled.
While I’m happy enough with that division of labor, it is probably best not to imply that the Linux Kernel project deserves the credit for any video drivers.
Of course, most people here already know that. But I feel better making it explicit.
Edited 2009-09-22 03:05 UTC
Fair enough too. The kernel developers did not write this driver, it was developed out-of-tree.
However, to be clear, the kernel developers have now accepted the entire ATI xorg-server-video-ati driver into the kernel staging area (not just the kms parts). This code will be released along with the rest of kernel 2.6.32 in about 10 to 12 weeks time on the current schedule.
And furthermore (correct me if I’m wrong) most of it lives in user space. So not only was it developed “out of tree”, but it actually only has a limited association with the kernel at all. The kernel bits needed to facilitate the far more substantial 3D user space code are scheduled to be included in kernel 2.6.32.
I think that is all correct.
It is just that “drivers” are generally considered part of what you get when you run a “kernel”. Exactly where the “driver” stops and the “userspace applications” begins is a bit of a fuzzy line, really (well at least it is fuzzy to me anyway).
Maybe I’m just old. But to me, this whole business of X drivers having any sort of kernel component at all seems a bit new-fangled. Traditionally, everything in X has been user-space. And the KGI guys received not just a cold shoulder… but active hostility from the kernel guys,
But then came DRI… which needed DRM support in the kernel. And now KMS.
I don’t really see it as being all that fuzzy. Those bits that so clearly need to be in the kernel to retain any degree of design sanity, the Xorg guys have mostly been able to convince the kernel guys to include. (Mode switching? Hello? How many years did this take?)
But I’m pretty sure that if you did a line count of kernel and userspace portions of Xorg drivers, the kernel portion would be tiny compared to the rest.
I am not sure on which planet you live in, but that is definitely not earth. The claim of the stability of “OSS graphics stack” is ridiculous, given the overwhelming amount of bug reports and frustrated users. While the core X.Org and associated utilities are stable, most of the drivers can not be in any way associated with the word “stable”.
XAA, EXA, UXA… Each new acronym developed for “rotating the cube” has took several years to stabilize.
ABC, BCA, CBA… So typical for Linux: rewrite and expect something better.
Yes, they are stable … right here on planet earth.
‘XAA, EXA, UXA… and each new acronym developed for “rotating the cube” ‘
… none of this is required for a stable business desktop. The functionality which is required for a stable business desktop has been stable in OSS ATI drivers for over a year now.
Do you always have so much trouble keeping up with a discussion?
Edited 2009-09-22 04:43 UTC
Given that most “business desktops” come with Intel graphics, answer to a simple question with a simple answer: are the Intel graphic drivers stable in your opinion?
Just so that readers have a clue what kind of stability to expect from the stable “OSS graphics stack”.
Edited 2009-09-22 04:48 UTC
Just lately at the cutting edge of development … no. There have been some perfomance regressions in recent versions of the Intel drivers.
However, one can easily run with older versions of the drivers. Fill your boots. Stable as you want.
If you want “stable” then you should be running Debian Lenny or something similar. That has no performance regressions AFAIK. Stable and supported. This (or something like it) is what you want to use for “a stable business desktop”.
http://www.debian.org/releases/stable/
See the URL? It says in effect “debian releases stable”.
Edited 2009-09-22 05:11 UTC
Yes. They’re the most stable of the lot, given that you generally don’t need to do anything when you install a distribution. No binary drivers, no trying out of the open source drivers and then trying fglrx to see if it makes any difference and then finding neither are ideal. I think most people would be surprised what gets used on a ‘business desktop’ and how it isn’t wise to cut things out and say “Oh, we don’t need accelerated video!” Yes, you need a sane graphics stack for that, and no, unfortunately desktop Linux systems just don’t have it.
Mind you, I just laugh that after all these years people still have the mental disease that makes them talk about about ‘business desktops’ in relation to desktop Linux. Linux is on a very small minority of ‘business desktops’, as useful as it might be, and someone needs to work out why that is before they start spouting the ‘business’ and (shudder) ‘corporate’ desktop nonsense.
I’m not quite sure I am following you. I think we actually agree, but I’m not sure. In my experience running about 70-80 thin clients on widely varying hardware, ranging from old to new, I have found that for basic 2D work needed by general business desktop users, both the core Xorg infrastructure and the drivers, both OSS and proprietary, have been stable for a very long time. But the 3D and composited stuff is still dicey for some drivers.
That said, I do have the occasional issue of having to fiddle to get the desired resolution. Though as often as not the problem is that the monitor is lying about its capabilities, rather than any actual problem with drivers. Manufacturers don’t really seem to care about getting their EDIDs right, since they ship a “driver” (for Windows, of course) with the monitor.
About OSS ATI drivers: it’s encouraging to hear that 3D acceleration is the only thing they lack, especially seeing as it’s only 2009.
The specifications for R6xx/R7xx ATI chips (which enabled the 3D graphics capability to be developed) were only released to OSS developers in January of 2009.
Having OSS ATI drivers for R6xx/R7xx before 2009 would have required either some cunning tricks with time travel, or industrial espionage.
Edited 2009-09-22 04:35 UTC
Time travel is not an option, so you’re forcing me to consider “industrial espionage”. Does it include reverse engineering?
If so, are you implying that it was impossible, too difficult, or undesirable? (Which one(s)?)
Just wondering. I am aware that a perfect reverse-engineered 3D driver is less feasible than a perfect BSD networking driver.
Note: Previews don’t show quotes formatted correctly.
Edited 2009-09-22 11:01 UTC
“Industrial espionage” is spying … breaking into a building and stealing plans or specifications would be an example. Clearly not an ethical option either.
“Reverse engineering” is perfectly ethical. It is a perfectly legal way that one may use to uncover a trade secret.
I don’t know which one for sure. I suspect “too difficult”.
For a long time, the OSS “radeon” driver supported hardware-accelerated 3D graphics for ATI cards using the R500 GPU or earlier. This was reverse-engineered to start with, and about 18 months ago AMD released the specifications, and the driver was improved.
For R6xx and R7xx GPUs however, the architecture was changed, and there was no successful 3d hardware-accelerated driver ever developed for those GPUs using reverse engineering techniques. It wasn’t until the specifications for R6xx/R7xx chips were released in January this year that any headway at all was made to making a 3D hardware-accelerated driver for these chips.
Edited 2009-09-22 12:02 UTC
I’ve never used open source drivers for ATI. I don’t use ATI. I’m not even interested in ATI. I only considered what you wrote:
“[T]he OSS ATI drivers (…) are very stable and work just fine, (…) … the only thing they currently lack is 3D graphics hardware acceleration.”
Of course I don’t necessarily believe what you said in this quote (just as I don’t necessarily believe what anybody posts on this forum just because they posted it). I only watch opinion flow.
And, particularly when information is scarce, I simply take language to mean what language itself means and try the least: to fix potential contradictions or obvious anomalies so as to make one’s (your) message look more credible to me.
This is what I did with your original statement. And here is how I’m fixing it (by using your latest messages): it seems it should have been:
“[T]he OSS ATI drivers (…) are very stable and work just fine, (…) … the only thing they currently lack is 3D graphics hardware acceleration [for the latest ATI cards].”
Or something like that. And now the world makes sense again.
This really misses the important point, going forward. When new chipsets are released, you can bet that the Windows drivers are going to be ready by the day of release. How long, if at all, are Linux users going to have to wait? Are FOSS devs going to get the advance info they will need to have drivers available in a timely fashion? And if so, are they going to be able to develop the drivers as fast as AMD’s Windows driver devs? Or are we going to continue to be forever behind?
As a Radeon 7000 and 9100 owner, I’ve been around this particular block a couple of times before, when ATI was supposedly being Foss-friendly.
Edited 2009-09-22 20:36 UTC
It would depend on a major architecture change, I think.
If a new chip introduces more RAM, more pipes, more shaders, faster thingummies and whatnot, surely the driver needs only a “definition” of the new resources in order to work. I would see this as a similar (but of course different) thing to “ppd – poscript printer definition” files for the CUPs driver, to describe new printers to it. It would be insane to require a whole new driver for each new chip/printer model that comes out.
The arguement above applies equally well to Windows as it does to Linux.
As far as the timing of supporting new chips on Linux compared to Windows, of late it has been “same day”:
http://www.phoronix.com/scan.php?page=article&item=amd_evolution&nu…
“AMD also continues to publish new programming guides and register information on a routine basis for their latest and greatest hardware.”
“AMD has not only provided same-day support for their just-announced Radeon HD 4800 “RV770″ series, but they’re now beginning to ship the Linux drivers on the retail CDs included with these newest graphics cards. In addition, AMD is very close to reaching feature parity between their Windows and Linux drivers.”
The OSS ATI driver has included 3D acceleration for about a year now, for certain chipsets where ATI has released the specifications. I remember because I upgraded a friend’s machine to Intrepid and had Compiz out-of-the-box. Integrated (recent) ATI graphics, no fglrx.
Your statements contradict Lemur’s or at least humiliate Lemur’s sarcasm. But I’m sure Google can help show who’s wrong.
Reediting note: Previews don’t show quotes formatted correctly.
Edited 2009-09-22 12:07 UTC
Your statements contradict Lemur’s or at least humiliate Lemur’s sarcasm. But I’m sure Google can help. [/q]
There is no contradiction … there is just a major change in architecture between R500 and earlier and R600 and later GPUs.
The earlier 3D code worked only for R500 and earlier chips. Specifications for those GPUs were released by ATI about 18 months ago.
PPS: {edit – correction} about 22 months ago.
Specifications for the later (and radically different, and apparently far more obscure) R6xx and R7xx GPUs was not released by ATI until January this year, as I stated.
PS: And here they are, all nicely dated for your inspection:
http://www.x.org/docs/AMD/
Ergo, until January this year, no work could be started on a driver that would work in 3D for all ATI chips.
Vindicated !
Edited 2009-09-22 12:19 UTC
As I just explained, his reply does contradict your *original* statement, the one I originally addressed: “[T]he OSS ATI drivers (…) are very stable and work just fine, (…) … the only thing they currently lack is 3D graphics hardware acceleration.”
Please read my full explanation in my other recent post.
Not that it matters, though. It’s just for the sake of accuracy.
This is exactly my thought too. Actually, the core features have been stable for decades, but the development seems to be pushed forward by the 3D crowd.
As a user of Intel graphics, during the past couple of years the Xorg progress has unfortunately seemed like “one step forward, two steps backward”.
You should learn the original meaning of that phrase. You’d be pleased to learn that it means a very small, but discrete change. Very appropriate when speaking of X.org dev.
http://en.wikipedia.org/wiki/Quantum_leap
“””In the popular sense, the term is usually applied to mean a large or significant change, which is thus not strictly correct.”””
I was just musing a bit about the programming specifications that were released by ATI for their GPU chips:
http://www.x.org/docs/AMD/
Readers (at least those who have some passing acquaintance with programming) will note that this type of information is not easily discoverable about the chips, it is not self-evident or intuitive, and it is absolutely required in order to be able to write a driver. It isn’t really innovative either, it is basically just reams and reams of mundane information, it is not a “creative work” in the copyright sense.
Astute readers will also note that in publishing this information, ATI are actually “giving away” very little information about how to actually design and make such GPUs to competitor firms.
So exactly why is this same type of information held so super secret by Nvidia for their chips, one has to wonder? What purpose is thereby served?
Hmmmmm?
Edited 2009-09-22 12:51 UTC
Lugubrious Lycan – the logo would have been awesome