Slashdot Mirror
ATI Releases New Linux Drivers
dinivin writes "Today, ATI has released all new 2D/3D drivers for Linux/XFree86. The drivers will work on any "Built by ATI" Radeon 8500 or higher card (up to the 9700). Unlike the previous drivers from ATI, these support both the XVideo extension and S3TC (making UT2003 playable with these drivers)."
maybe I wont need to use the Gatos drivers anymore... this would be very nice!
Let's hope they got it right.
Reviews of the stablility and performance of these drivers will probably be a major factor in my decision on whether or not to buy a 9700. I've been hesitating because of all the bad things I hear about their drivers. I use NVidia now and I've never had a problem with the drivers, so I'm a little worried about switching.
Aw crap, ninjas!
Goodbye forever, windows, you won't be missed.
If I ever see a BSOD again, it will be too soon.
Does anyone know if this will work on PPC with Gentoo or Debian such as those Powerbooks that come with the mobility radeon 9000?
I bought a Radeon 9700 Pro a while back, and the only way I could run X with it was to use the VESA driver, which was SLOOOOW! I can finally go back to Gentoo as my primary OS! (Now if they would just release 1.4...)
Spoiled Linux punks.
Back in my day we had a galvanized metal box with a circuit board dangling in it. We had an old VT100 terminal hooked up, and we were happy!. In fact we were so poor we couldn't afford all the serial lines so we had to get by with just both data lines and the ground, but we were happy! None of that Fancy-Pants hardware control stuff that became popular among the Brylcreme'd University people at the time.
Did I mention that to get to this VT100 I had to walk 40 miles uphill kneedeep in snow? Both ways?
bah..
[/curmudgeon]
Trolling is a art,
This is absolutely wonderful for Linux 3D graphics. Depending on how well these drivers perform, gamers and graphics developers alike will have an alternative to NVidia.
The ATI drivers don't even need to outperform NVidia's. An ATI graphics card is almost always cheaper than the corresponding NVidia card. Some of us don't like spending any more of our own money on a computer than we have to.
Am I the only one who's had problems with some games crashing until this last batch of NVidia drivers came out. For that matter, the last batch didn't include an update for my GeForce2Go (stuck in OEM land), and it *still* crashes a lot.
sigs are a waste of space
RPM is nice and such, but please do like Nvidia, and provide a non RPM option ! I can get around this by using RPM and extracting the stuff, then making an ebuild or something, but hey, it is much easier if RPM is complemented by a tgz
life+universe+everything=42
UT 2003!
Linux Games!!
Tux Games!
Neverwinter Nights!
In your face you greasy little "Linux doesn't have any games" troll!
It is all well and good that they are putting out drivers that works "across the board" for their product line, but I have seen, time and again, where a "universal" device driver is not so universal after all. If it was written on a machine sporting an 8500, where does it degrade with the 9700 and so on? If they are not the same card, they won't be 100% compatible.
Another possibility is that the drivers are written to work generically with the chipset. This would have the distinction of having unremarkable drivers that do not push any card to its full potential.
My deep and sincere apologies to ATI if they are successful in making a universal driver for their stuff that actually takes full advantage of each device. I would bet that such a driver would be a real winner.
Be excellent to each other. And... PARTY ON, DUDES!
The original drivers were for the professional FireGL 8800-type cards only. Then people figured out they could be made to work on the regular 8500 as well, and instead of putting the smack down decided to officially support it as well. Now these new drivers have xvideo and s3tc support so that desktop and gaming users will enjoy them a lot more, and work on 8500-9700 ATI cards. Keep up the good work, and don't forget DRI people too :)
Has anyone benchmarked the new drivers vs. NVIDIA yet? I'd be curious to see how well they perform.
Project Steve
nVidia has used a universal driver for years. Doesn't matter if you have a GeForce2 MX or a GeForce 4600, you download the same driver for the OS.
I wonder - is the "installation" package unified, or is the actual driver that gets installed unified?
IE the installation program detects what driver needs to be installed, and then pulls the relevant files out of the installation file and installs them (how many times can one use the word install or it's derivatives in one sentance before you are forced to take a technical writing class?).
I think will have to wait for the benchmarks to come out to figure out the answer.
Now, I can buy that outrageously expensive alienware laptop with the Radeon 9000 and bring it to my lan party to kick some serious rear in UT200-whatever!
_ _______
I just need more linux games.
Brother, do have another Loki to spare?
One that can run a company this time would be nice.
Ok, now back to serious work.
__________________________________________
ACK
From the release notes:
NOTE: The OpenGL driver can use AMD 3DNow! enhanced opcodes as well
and - due to design - does not need a kernel patch for AMD 3DNow!.
Now that's the kind of thing I like to see.
For Radeon cards (up to the 9000 ATM) there are free software DRI drivers as well. They cannot perform as well as these and the Windows drivers because of restrictions on what can be released as source, but they work well on BSD, which the ATI driver's don't do and NVidia didn't do until very recently.
The FreeBSD porter did a good job with the dri-devel tree - it goes through the tedious process of building and installing a new XFree86 DRI setup for you. I was running my 8500 under FreeBSD the same night I installed it, to my pleasant suprise.
Does anybody know if this driver supports the video input/output features of my All-In-Wonder Radeon 8500DV? I'd love to have xawtv running on my screen, or to watch mplayer on the TV.
Or do I have to run the GATOS driver for that?
I like how their license agreement on the download page is in a text area in a form. I erased all of the text and wrote "ATI will give me one BILLION dollars," and submitted it. And they accepted it! Thanks to UCITA, that's valid, too. (I think. OTOH, who the hell can figure out UCITA?)
Ooh, I submitted it again and now they owe me a monkey. Pay up, ATI!
-Waldo Jaquith
The press release gives more information. These are unified drivers for ATI cards on Linux--COOL.
There were rumours flying around a while ago that open-source Radeon 8500 drivers would be appearing. I'm therefore sad to see that ATi have decided that closed source drivers are the way forward. I don't see any reason to promote this on Slashdot, or to consider this in any way beneficial for the open source community; remember, closed-source Linux drivers are not support, they're marketing. Thanks, ATi, I'll be buying my graphics hardware elsewhere in the future.
Are these drivers open source, or do they include pre-compiled object files that cannot be re-compiled?
This is excellent news. Now all i need is to find the book on "Installing ATI Drivers under linux."
Actually, for what we used it for, we were happy.
Best Slashdot Co
I can finally go back to Gentoo as my primary OS! (Now if they would just release 1.4...)
emerge rsync (update the list of what's available)
emerge -up world (preview what's comming)
emerge -u world (do it!)
Gentoo isn't like other distros, in which you must wait for a release to stay current. With gentoo, the above three commands bring you up to what is current, which is generally close to the leading edge of the state of the art.
Oh, but you don't like the freeze and want all those new ebuilds waiting in the wings for the release? Fine, just set ACCEPT_KEYWORDS="~x86" and you can jump past the pending release and play with all the experimental stuff coming down the pike.
I have one set of partitions for exactly that purpose, and one set for the more formal, stable stuff. And you know what? With this approach, I don't have to even care at all when, or even if, they're going to have a "final" release of 1.4. The only other distros I know which come close to this is Debian unstable and Source Mage. The former suffers from the Curse of Binary Distros (lag behind the state of the art by weeks or, in the case of xfree, months), the latter is quite good, comparable to gentoo in many respects (but a different approach, so like salad vs. steak, the choice is entirely up to your own sensibilities and taste).
The Future of Human Evolution: Autonomy
Now I can buy an ATI card. Good for them and for us.
unfinished: (adj.)
Strange, because Mandrake Has supported my Radeon 7500 since at least 8.2. Probably was in previous version but can not comment as I only had the card since installing MDK8.2 and then upgrading to 9.0 on my computer. It was probably a configuration issue on your part.
Just a Tuna in the Sea of Life
Xi Graphics
I have been less than impressed, but they do work. Maybe you will find it is worth while.
Just a Tuna in the Sea of Life
I'm typing this on a Gentoo box with two DVI LCD monitors attached to my Ti4600 card. Running one large desktop across both monitors WITH 3D acceleration across both monitors.
I might add that you can't do that with the ATI drivers, nor is there any flavor of ATI card that drives two DVI monitors (not that there's a huge selection of such cards with Nvidia chips, but Gainward does make one).
Nvidia is really the best choice for performance graphics on Linux.
FYI.
jonathan
This topic is timely, because I'm spec'ing out a new computer for the family for Christmas. In these dog days of the economy (national, local, and mine) I'm trying to keep the entry cost down. Besides, it will give me the chance to add parts over the next year or two.
Someday DirectX 9 and OpenGL 2 will be worthy targets for purchase. But today, only the Radeon 9700 is there, and I'm not spending that kind of money. So near term, the target is Doom3.
My price target is around $60, since I plan to replace it in a year or two when R300/NV30 features become affordable. ($150-range) But I don't want to wait until then before playing Doom3.
The Radeon 8500 cards are all above my range, so...
Some Radeon 9000 cards are in my range.
Some Radeon 8500LE cards are in my range.
Will these new drivers work for these cards?
Will these cards (9000, 8500LE) play Doom3?
My backup plan has been a GeForce4MX-440, which is supposed to play Doom3 with reduced features and speed - not a preferred card.
The living have better things to do than to continue hating the dead.
which totally sucks on a laptop. I've got a dell insiron 8200 which has a geforce4. the older drivers could be re-compiled to ignore apm, so I could hobble by. however they crashed a lot. the newer ones are more stable, but will lock up the machine when put to sleep. the recompile/patch doesn't work. nvidia is ignoring the apm issue despite many pleas from the community.
ATI is a Canadian company, are they liable under UCITA?
Dmitry Sklyarov is a Russian guy. Is he liable under the DMCA?
-Waldo Jaquith
Now can Doom III be played on a Linux box, I remember John Carmack saying how only the Geforce series was to be supported?
3000 dead over past 2 years, still no free Palestinians, still
There's nothing wrong with mixing free and closed software. If these drivers enable me to play the likes of UT 2003, then so much the better.
Here on /. I see many posts about driver support for Linux-based Operating Systems lacking - here's one of the market leaders producing drivers for Linux. IMO, we should be congratulating ATI.
Tim
Backup not found: (A)bort (R)etry (P)anic
We used it. RS-485 can go several hundred feet and is highly noise resistant. And much lower cost than interbus-s, fibre, etc. If you need noise resistance, speed isn't a factor (IIRC, after 24kbps it began to Have Issues), and you want to keep the cost down, then RS-485 is a good solution.
Best Slashdot Co
It doesn't matter where you're from when you commit a crime in the US. If your crime occured in the US, you'll get deported, or picked up when you step off your airplane.
The reason that Dmitri shouldn't have been touched is that he didn't violate the DMCA. Someone else in his company did. Whomever distributed his product is the "criminal." Creating the product occured 100% on Russian soil, and was not a violation of the DMCA. Shipping it/wiring it to the USA was a violation. But Dmitri didn't do that. Since this is criminal law we're talking about, you can only go after the individuals that commit the crime, not some random member of their company.
Unless I'm totally misunderstanding the situation. Maybe Elcomsoft is a two person company, and Dmitri really did send the product to the US. Maybe the "crime" was his presentation, and not distributing their product.
Either way, it's the law that's fucked up, not the fact that it was applied to a foreigner. Being from another country doesn't give you diplomatic immunity. And it shouldn't. The US isn't bad in that regard. If you mail a bomb to Italy, and you live in Greece, you'll get deported, or arrested the next time you travel to Italy. Right?
There are no trails. There are no trees out here.
Butterscotch is superior to vanilla when analyzed at a compound level.
In fact, butterscotch is superior in all aspects. Butterscotch tastes better, due to its ingredients it is healthier. Recent studies show that butterscotch *looks* better too. The only thing that vanilla is better than butterscotch at is hit/miss ratio of the trash can. And that is because butterscotch actually gets eaten; recent surveys indicate that butterscotch pudding is preferred 100% to 0% over vanilla pudding.
Butterscotch also contains a larger feature set than vanilla. When distilled, butterscotch makes a great, long-lasting chew-candy. When frozen, it makes a fantastic jawbreaker, when heated, it results in a glorious milkshake.
In conclusion, your must see what is obvious: Butterscotch is Better Because it Begins with a B, and because they don't make Apple pudding. If you weren't so closed minded to new ideas, you would have seen this a long time ago. I hope that this simple explanation corrects your longstanding error in judgement.
Video meliora proboque deteriora sequor - Ovidius
Well, I own an All-in-Wonder Radeon. It's not _that_ old (300USD a couple years ago), and it's unsupported by their unified driver! And I don't even talk about the multimedia features, TV in-out, which are mostly broken in Gatos tools/drivers and non existent in their own driver.
I'm back on Win2k for the time being, partly because of this. And I wonder if my next purchase will be ATI, based on my current experience. Sad, because the hardware is rock-solid!
have you been defaced today?
Likewise, my 8500 128 sits in the gaming (ie, XP) machine. A 7200 sits in my Linux workstation. Plays Quake 3 like a charm, but without the S3TC support, no UT2003.
I went to the download page and discovered that the
." This allows me to build the NVIDIA drivers for any distro I'm using OR any tweeked kernel I'm using.
rpms were ONLY in i386 packages, no re-linkable source distro.
In the past I've always downloaded the NVIDIA src RPMS and just done a "rpm --rebuild . .
Restricting the users to the distro's stock kernel kinda sucks.
But it doesn't suck nearly as bad as having NO support whatsoever.
Thanks ATI, you just made the decision for my next notebook considerably more difficult.
Well, my Radeon 64 DDR VIVO works on my EPoX VIA KX-133 based motherboard.
Using Mandrake 9.0 (worked in Mandrake 8.0 -> 8.2 as well), I can play any games based on the Q3A engine (such as RTCW), and it works fine. The only problem with the open source drivers for the Radeon 7200 (aka: 64 DDR VIVO),is lack of S3TC support.
This also seems to be causing a problem with WineX as well. I just want to play Counter-Strike and Natural Selection under Linux with a half decent frame rate, and NOT have to buy a new vid card.
By the time this driver if functioning to the performance level I want, I'll probably have bought a new card anyways. But this is only a problem for games played thru WineX.
...waiting for my new Voodoo 5 5500 drivers, damn it! %$#!@
:P
C'mon, ATI, throw me a bone here!
Oh my god, have we slashdotted ati.com? I can't reach their site anymore.
:-)
Probably running their new drivers on the linux-powered webserver.
The 4200s are probably close to $100 even right now (Were $125 2-3 months ago). They will probably drop down even more by Xmas.
I would strongly suggest waiting for reports of driver quality before jumping to ATi because of this release - Their track record as far as drivers go is not very good... (I've been burned by ATi products not living up to their claims and crashing my machine due to bad drivers too many times to ever touch them again.)
retrorocket.o not found, launch anyway?
Should've thought of this before hitting submit.
Doom 3 won't be out for another few months. Recycle your old video card and don't buy a video card for Doom 3 until the game comes out.
retrorocket.o not found, launch anyway?
RS-485 protocol is the same as RS-232. The electronics are different, but that should be handled by the hardware.
Best Slashdot Co
Never heard of alien?
Converts from RPM to debs, or tar.gz's, etc. apt-get it if your debian, urpmi it if you want it on Mandrake for any reason (converting those goddamn debs!), I guess if your slackware you already know how to find it...
Usually support for newer cards is built into the driver sets before the actual hardware is released. That's the case for the Windows version, so hopefully there will be a build for Linux available which supports NV30 by the time it comes out.
-- Jim
RPM is the standard Linux package file format. If your distro aims to be Linux Standards Base compliant, it must have a mechanism of installing such files.
Preferably a full RPM implementation, but systems like alien or even (I guess) rpm2cpio are acceptable.
I just had a 1-hour confrontation with those drivers. There are several things:
Well, after installing a fresh X 4.2.1 from debian unstable, fixing about thirty parser errors in a source file and wreaking general havoc, I was at least able to start X. 3D seems to work, but I was not inclined to do much testing beyond fgl_glxgears and glxinfo after realizing that I was unable to use a text console without snapping back to the X console every second.
All this slowly leads to a heartfelt "fuck ATI" feeling and I'll have plenty time to ponder this while I restore my X config that mysteriously lost all 3D acceleration and Xvideo capabilities after switching back to the DRI driver.
Fight hunger. Filet a politician and send him to a 3rd world country of your choice.
Sorry, but when I couldn't get my Radeon 8500 to drive the DVI output at 1600x1200 under Linux, I voted with my wallet and went out and bought a GeForce 4 card for my new (Linux) machine. The ATI is left in my Windows machine, which is in the process of being shut down, while the Nvidia card drives my TFT at 1600 x 1200 very nicely with SuSE.
3D I don't need, but I was surprised ATI hadn't figured that high-end cards are also bought to drive high-end displays (as well as for playing games) and so cross-platform support does count for sales (see also the ATI workstation cards).
I spent a lot of money on booze, birds and fast cars. The rest I just squandered. - George Best
Um, I didn't mean to imply that deporting someone was easy. Is that the only part that you felt was really incorrect?
Dmitri did something in Russia that would have been a crime, had he committed it in the US. So he should not have been arrested. All I was trying to say was that being Russian isn't should have made him safe. It was that he didn't violate the DMCA, someone else in his company did. If *that* person had flown to the US, all the same things would have happened to that person, and there'd be no jurisdictional question at all. Right?
There are no trails. There are no trees out here.
source for that would be nice, to port it to NetBSD eventually. Not all the world is Linux!
- Hubert
Nope. Unfortunately you're wrong. PowerColor's 7500 does not work with any of the available drivers. Buying "Powered by ATI" instead of "Built by ATI" is a complete crapshoot and I'm willing to bet that a lot of the complaints that one hears about ATI are because of these shitty OEM clones. The XFree86 people only work with Built-By-ATI cards and their drivers work _beautifully_ with them. ATI is shooting itself in the foot with the "powered by" stuff.
Oh, yeah and I forgot to mention, if you look at the response time for the memory on a lot of "powered by" boards, it's not the same as the official "built by ATI" boards.
My strong advice to anyone that is thinking of getting an ATI is to spend the extra bucks and get yourself the real thing.
RPM is the standard Linux package file format. If your distro aims to be Linux Standards Base compliant, it must have a mechanism of installing such files.
No. RPM is not the standard Linux package file format. The standard Linux package file format is the tarball, either gzipped (.tar.gz) or bzip2ed (.tar.bz2), or uncompressed (.tar).
RPM is a part of the LSB standard, which is just one of several Linux standards that are NOT universally accepted, nor should it be. RPM was placed in the LSB because of Red Hat politicking and in an IMHO very illegetimate effort to give them an edge over other distributions. Indeed, RPM's inclusion in the LSB is the main reason why the LSB should, IMHO, either be rectified to exclude it, ignored altogether, or (ideally) adhered to in other respects, with the RPM provision sumarilly ignored.
The pointlessness of including RPM in the LSB standard is underscored by the incompatability between Suse RPMs, Red Hat RPMs, and Mandrake RPMs (to name just three), and by the success of many products which have been packaged in proper, distribution-agnostic form (nvidia drivers being one such example, but by no means the only one).
Yes, superior distributions such as Debian and Gentoo can extract the necessary data from the cumbersome RPM format, but forcing them to jump through that particular Red Hat hoop is neither justified, nor desirable.
The Future of Human Evolution: Autonomy
Oh. Of course, sometimes it's very hard to get someone deported. Hence, "You'll get deported, or arrested the next time you travel to Italy."
But I really don't understand the case you describe. There was a case where a Canadian committed murder in the US, got arrested in the US, and they tried to get him deported to Canada before he was tried in the US? I've definitely never heard of anything like that before. I've heard of Canada refusing extradition due to our (braindead) capital punishment, but what you describe is pretty bizarre.
Sure, if the US asked Russia to extradite Dmitri, we'd get laughed out of Moscow. But that's not what happened. Again. The problem is that we have an unjust law. Dmitri shouldn't have been arrested because he did not violate that law. Not because he's Russian. Being a citizen of one country doesn't mean you can violate the laws of another, and then expect to travel there. Again, if the *crime* occurs in the US, and then the criminal is in the US, arrest the criminal. This is not complicated, and it works the same way if you... do drugs in Singapore... steal fruit in Qatar... whatever.
The problem is the unjust law.
The problem is the unjust law.
The problem is the unjust law.
Right?
There are no trails. There are no trees out here.
Did you generate a new XF86Config file or did you just use one from the previous FireGL/8500 drivers?
I'm using the drivers here and XVideo works fine.
Dinivin
Because ATI obviously is having problems doing this. ATI's not a small company. They can obviously make some good hardware, so they've got engineering talent. If they've been unable to get a good solid set of drivers together after this long, there is obviously a problem in their driver development team, and studying binary code doesn't seem to be working. If I was NVIDIA, I wouldn't take the risk and give ATI any more help than was necessary.
A deep unwavering belief is a sure sign you're missing something...
Don't get me wrong, I'd love to see open source drivers for NVIDIA hardware.
As far as instability, I'm guessing you have an AMD system with a VIA chipset? There are a lot of hardware bugs with VIA chipsets, especially the earlier Athlon chipsets. There are a bunch of workarounds, you should try them all before giving up.
Cheers.
Sticking feathers up your butt does not make you a chicken - Tyler Durden
"Built by ATI" or "Powered by ATI" ?
The website mentions the gatos project several times for support of this card.
My Radeon 7500 works great right now. I'm using the latest gatos drivers, and xawtv works great, and 3D acceleration is good enough to play RtCW and Tribes2 (albeit in pretty low res on my 1Ghz Duron). The only shortcoming is that XVideo mode won't let me use the framegrabber, though I think this is working in CVS versions of gatos at the moment... after that, we should be able to record as well as view TV inputs.
I had some instability earlier, but it turned out to be problems with my motherboard BIOS instead. Playing around with different BIOS releases seems to have fixed the problems.
Let me know if you need any help... it looks like the Radeon 7500 is a dead end though (but I wasn't about to pay twice as much for an 8500 AiW). Support from the gatos project has been superb!
Basically, all parser errors occur on lines that use the __KE_DEBUG (or something similar) macro in fglxr_public.c. The macros are defined as __KE_DEBUG(fmt,args...) and it seems gcc <2.96 can't handle that when called with just one parameter. All I did was rewriting each call to that macro to have at least NULL as second parameter.
There are also errors that are caused by the patched drmP.h. I got around those by disabling the patch contained in make.sh.
Fight hunger. Filet a politician and send him to a 3rd world country of your choice.
Do you know why NVIDIA backtracked on their promise to deliver open source drivers? It was a couple years ago, at the time they released some initial drivers that worked with XFree86 3.x.
Sticking feathers up your butt does not make you a chicken - Tyler Durden
It makes no sense to me why people hate RPM so much. A packaging format of some kind seems like an absolute necessity to me. Tarballs are most certainly not a packaging format.
A proper packaging format will keep track of what's installed on your system, where it's installed, and what depends on it, and what it depends on. This is so that addition and removal of packages is easy.
The only advantage people have ever given for .deb is apt-get. But, it seems to me that the same functionality is replicated in RedHat network and Red Carpet. The usefulness of apt-get has more to do with infrastructure support than it has to do with the .deb packaging format.
So, please, tell me why RPMs are bad, other than that RedHat created them.
Need a Python, C++, Unix, Linux develop
That's why I mentioned stealing fruit in Qatar and doing drugs in Singapore. Both of those actions carry incredibly unjust consequences, and the US Gov't will do nothing to protect its citizens from those consequences.
Even when there was that huge outcry 'cause some dumbass American was going to get caned in Singapore, the *Gov't* didn't do anything. Pols might have lectured about how Singapore shouldn't cane the kid, but there was no official action.
Iran *does* have "crazy religious laws" but it's still a particularly bad example. Since the US and the Iranian gov'ts do not have any relations, an American woman in Iran would have less than no diplomatic sway. The only protection she might get would be due to internal popular pro-US sentiment. But that probably wouldn't do anything anyway.
Iran would be wrong for doing it, but for the same reason they'd be wrong for doing it to their own population. As long as Iran is a sovereign nation, they can make whatever inane law they please. I guess it can't violate internationally accepted human rights, or they might face war, but that's really the only threat.
There are no trails. There are no trees out here.
1. A package format is expected to provide more than a mere compressed archive of files. Tar is an _archive_ format, not a _package_ format. I'd ask you kindly not to post a response to this comment until you understand the difference.
.debs, or what have you).
... are you going to take issue and declare that none of those developers release packages, merely archives?) and its availability to the distribution user, generally much quicker than binary equivelents (though Mandrake Cooker, as another pointed out, is fairly good at keeping current, albeit at a stability cost the source based distros don't suffer from). Other advantages include a 20-30% speed improvement from compiling the system optimized for the hardware it will run on, added stability by eliminating the sort of subtle incompatabilities binaries often suffer from when compiled against a slightly different library revision, and so on.
Translation: "I disagree with you, therefor you are an ignorant fuck. Please shut up until such a time as you agree, both with my definitions and my conclusions."
The difference between an archive and a package is one of semantics. A tarball can easilly contain all of the information necessary for a package to be built, indeed, most source tarballs these days do exactly that.
You want to define a standard for that, go right ahead. But do not in the process favor one distribution over another, or chuck out a perfectly good archival approach that is a widely adopted, cross-platform standard for one that is obfuscated, inferior, less widely adopted, and less generic.
Such as standard could be as easy as a parsable, human readable text file "dependencies.txt" in the top tarball directory, for example. There are any number of solutions to that requirement, almost all of which are more elegant than RPMs (or
Furthermore, not all distributions use binaries. Any standard that makes an assumption that all od (as LSB to some extent does, by adopting RPM) is inherently inadequate and broken. Indeed, acceptance of such a standard would likely inhibit a fair degree of development and progress in the GNU/Linux community, particularly when it comes to exploring less traditional methods of package organization, handling, and distribution.
For example, many distributions, such as Gentoo and Source Mage, use source and build the installation dynamically. Debian apt-get source is another such example. All of these have the advantage of having a very short time-to-market between the developers release of a package (generally in tar format
In short, there are compelling reasons why adding a binary package format, particularly one such as RPM, will not have a beneficial affect for Linux as a whole (though it certainly does benefit Red Hat).
Using incompatability between rpm's produced by different distros as an argument against rpm as the LSB standard package format is really back-asswards, given that the one of the main points of the LSB is to _ensure_ distribution interoperability.
If that is indeed its purpose (and I don't dispute that), then it is already a miserable failure. Suse and Redhat RPMs that are LSB compliant still break from time to time when used on the other platform, so clearly LSB compliance alone isn't enough to guarantee compatability anyway.
An rpm made in adherence to the LSB spec will work on any LSB-compliant distribution.
That may be the claim, but as noted above, it simply isn't true. LSB compliant RPMs still fail to be compatible across distributions. By including RPM in the standard they've created a Red Hat specific loop everyone is expected to jump through, yet the dubious benefits it purports to offer remain unrealized, indeed, are quite possibly unattainable without hamstringing diversity to the point where all distros are required to be One Distro for compliance, a la the woefully misguided "United Linux" initiative.
RPMs are ugly, unwieldly, distribution specific, and an unnecessary complication that has no place in the LSB standard. Were it not for politics and certain entities wielding undue influence on the standards body in question, it never would have been included, and the LSB standard would have been better for it (and much more widely adopted).
The Future of Human Evolution: Autonomy
Gentoos portage of the Xfree doesnt support the 9700 yet. (Thou I saw 4.2.99-3 was out last week, which might) I know the CVS version of Xfree did find my 9700 with --configure.
I'm considering a 9700-pro.
Do you find cvs-xfree to provide adeqaute 2d performance / support?
Have you tried ATI's new binary drivers (for good 3d support) and if so, how did you find them?
TIA
The Future of Human Evolution: Autonomy
OK, after fiddling with the config, it actually got somewhat better. If you create a XF86Config using the fglrxconfig tool and then copy some stuff from there to your real XF86Config (omitting the BusID and Screen entries), XVideo works and the overlay problems seem to be gone. My system still restarts X when switching to a text console though and 2D feels a little slower than with the DRI drivers.
On an unrealted note: does anyone here know how to get the "two screens, one framebuffer" behaviour under Windows?
Fight hunger. Filet a politician and send him to a 3rd world country of your choice.
Sticking feathers up your butt does not make you a chicken - Tyler Durden
Have you tried LSB-compliant RPM packages and found them incompatible with one or more LSB-compliant distribution, or is this pure FUD based on your previous experiences / hearsay about non-LSB-compliant RPM packages and/or non-LSB-compliant distributions?
I realise that last sentence was a bit long and unwieldy, so I'll rephrase: for the failure(s) you are reporting, are you sure the RPMs and the OSes were actually labeled as LSB-compliant? It's not like all RPMs or all releases of Red Hat are LSB-compliant, obviously.
Forgive my scepticism, but it sounds like you are confusing LSB-compliant RPM with Red Hat RPM and SuSE RPM. Either that or you are confusing LSB-compliant Linux distribution with any RPM-using Linux distribution. Or you could simply be blaming the LSB for what is actually an issue of poor, non-LSB-compliant packaging work mistakenly labeled as LSB-compliant.
Now, I'm not much of a fan of RPM myself, being a Debian user and all. But please, try to keep the FUD down a little.
Where in your conspiracy theory do you explain why the LSB standardised on RPM v3 rather than v4? The party line is that RPMv4 format is too new and would be unfair to all the non-Red-Hat players, who would have to play catch-up. This would seem to put a serious dent in your 800-pound-Red-Hat-gorilla postulate.
Where in your conspiracy theory do you refute the claim that RPM was chosen because a vast majority of existing Linux installations already use it as a standard package format, and a majority of the remainder have decent alien support for converting RPM packages to their respective native package formats? Remember, the job of the LSB is largely to codify existing common practice wherever practical, so as to cause minimal disruption to the Linux user base while providing useful guarantees to software vendors / packagers. RPM would appear to fit this bill quite well.
Where in your conspiracy theory do you provide the alternative package format which would satisfy the above goals? "tar.gz" is not much of an answer: there is no standard way to handle processing before or after installation, configuration or removal, or quite a few other useful tidbits. Anything you might add to provide such a mechanism would negate the value of using this "well-known" tarball format - your new format would become "tarball with certain magic pixie dust added". --Basically a home-grown RPM with fewer features and no existing buy-in. (But hey, at least it uses tar instead of that eeeeevil cpio!) Would that be sufficient to satisfy your NIH instincts?
"How can you claim that you are anti-crack, while still writing a window manager?" — Metacity README
All there?
If this is the whole story, it's a significant departure from ATI's previous R200 driver. Which is why I suspect it's not the whole story.
With the R200 driver, there were four components:
So 2 of 4 components were binary-only, a third was binary-mostly with a small stub for multi-kernel compatibility, and the last was a set of trivial modifications to existing GPL code and was therefore open source.
If this driver is like their previous one, which I suspect it is, there's no way it could be considered "open source".
"How can you claim that you are anti-crack, while still writing a window manager?" — Metacity README
Why is this marked redundant? I just searched through the comments and couldn't find an answer to whether it works with notebook chips.
Does it?
ATI formerly released specs for their older cards. I don't think specs are available for the latest cards and their vertex/fragment program featuers. Also, ATI would not give out specs for their video in/out hardware (I asked them and was refused). As far as "openness" is concerned I consider them friendlier than NVIDIA, but not by much. The difference is outweighed by NVIDIA's large lead in driver quality (esp for Linux) and general support.