Submitted by Here’s an update on the Linux Driver Project. “The Linux Driver Project is alive and well, with over 300 developers wanting to participate, many drivers already written and accepted into the Linux kernel tree, and many more being currently developed. The main problem is a lack of projects. It turns out that there really isn’t much hardware that Linux doesn’t already support. Almost all new hardware produced is coming with a Linux driver already written by the company, or by the community with help from the company. There are two main classes of hardware, video input devices and wireless network cards, that is not well supported by Linux, but large efforts are already underway to resolve this issue, with the wireless driver issue pretty much taken care of already, however there are a few notable exceptions. Because of this, our main effort has turned into one of education. Educating vendors of how to become members of the Linux kernel community, proper coding standards and procedures, and how to get their code into the kernel tree.”
As a longtime linux user, quasi-developer, it looks like he misses the point.
[X] We support hardware foo
That doesn’t mean that we support hardware foo well. There is a lot of hardware that has plain crappy support for linux even though “it works”. Take a look at any broadcom wireless card. After using b43-fwcutter to get the firmware out of a binary driver, the connection isn’t all that great. Yes, yes it works. That does not mean it works well.
Linux should work on improving existing hardware support to be on par or better than other operating systems. We are standing on the shoulders of giants.
Broadcom wireless is listed specifically as one of the hardware devices that is not supported by Linux due to the recalcitrant manufacturer.
FTA: “There are still some wireless vendors that do not provide Linux support directly. Two of these, Atheros and Broadcom have drivers created by the community through reverse engineering efforts. These drivers usually lag the introduction of the hardware by a number of months due to the lack of vendor support. Both of these companies have internal versions of drivers for their new hardware, but efforts on getting them to release them so far has been resisted.”
Therefore … avoid any hardware with a Broadcom chip in it. Shun it.
Maybe then Broadcom will feel the stupidity of their attitude to rejecting custom from a section of the market for no good reason.
“Broadcom wireless is listed specifically as one of the hardware devices that is not supported by Linux”
http://www.linuxant.com/driverloader/
http://bcm43xx.berlios.de/?go=devices
“Therefore … avoid any hardware with a Broadcom chip in it. Shun it.”
Better yet , get to know the Free Software developper who are interested in making it work and help them by testing there solution and reporting the problems and funding them instead of company who are not interested in fixing the problem.
If you really think shunning something as brought on any real solution your seriously delusionnal.
GNU/Linux started with nothing , if it had shun everything that did not work or wanted to support it there would be nothing supported today.
Edited 2008-04-08 16:47 UTC
You mean that developers has abandoned those drivers and doesn’t develop/improve them anymore?
Your reversing reality.
– Mac OS X don’t work on 98% of hardware. There support is done by the hardware vendor in exclusive deals with millions in investments.
– Windows support a limited number of hardware , most of it’s support is done by the hardware maker themself usually 2-5 years after the OS is released do you get a full set of support.
– BSD is a joke that support very limited hardware poorly.
– Not going to bother explaining or naming the rest of OS to you.
The quality of the driver on GNU/Linux is excellent , the problem is always with the one that offer zero support and that GNU/Linux get blamed for not fully supporting them.
You should inform yourself as to why broadcom cards have been a problem in linux land until recently. Broadcom has been downright hostile to linux driver developers.
Nonetheless, my notebook has one of these despicable cards and it works reliably now
If you read the full article it talks specifically about broadcom. Drivers for broadcom hardware are reverse engineered because the company makes no effort to support kernel.org.
“”There are still some wireless vendors that do not provide Linux support directly. Two of these, Atheros and Broadcom have drivers created by the community through reverse engineering efforts. These drivers usually lag the introduction of the hardware by a number of months due to the lack of vendor support.””
Given that most drivers nolonger need to be reverse engineered, (“for almost all modern hardware devices, [reverse engineering] is not necessary.”) if you read between the lines you can infer that quality will obviously lag as well.
“””Category 1 is already being handled very well by the Linux Printing project and the SANE project. Printer and scanner drivers in Linux are userspace programs and libraries and have nothing to do with the kernel at all. If you have any issues with these types of devices, please go ask the developers of those projects about it. They are very knowledgeable, skilled, have vendor contacts, and can do a lot to help resolve your issues. This area is already aptly covered by these people.”””
Nonetheless, it would be great if some of those 300 spartans^W kernel developers could help CUPS and SANE anyway.
Edit: since they are underworked and craving for devices to hack on.
Edited 2008-04-08 13:19 UTC
Typical Linux (et al in OSS) thinking. As an end-user I don’t care if the problem is in the kernel, the library, or the application. True that as a technical person I can appreciated the differences. When it comes to being a user, it comes down to the question “Can I get similar functionality I had in Windows?” Heck, even if the library did support my scanner, AFAIK, there is no decent application for KDE 4 for scanning with it! Sorry but people need to realise that hardware support means the full stack. Until it is complete, I’ll keep XP installed on my machine and I actually hesitate about booting into Linux* since I might need to scan something and I hate rebooting.
* Though I much prefer it.
“Typical Linux (et al in OSS) thinking.”
Ridiculous. This is a typical arrogant “all developers are demons that must be killed” kind of arrogant comment that seems to plague the Linux world these days.
Suppose you are a car radio manufacturer. I come to you to complain that my car lights are broken.
Me: “My car is broken. Fix it.”
You: “The problem is not in the radio. We’re only specialized in car radios. You should see a car mechanic if you want your lights fixed.”
Me: “Typical ‘car people’ (et all) thinking! As a driver I don’t care whether the problem is in the radio or the lights! Fix it!”
Riiiight. Your comment is like flaming the plumber that he doesn’t fix your television.
There is a difference between kernel developers and userland developers, whether you like it or not. It doesn’t matter what you say about end users “not caring”, it doesn’t make the difference go away, nor should the difference go away because if it does, chaos will ensue.
These people are kernel developers who try to help you with their expertise. Their expertise do not lie in userland. Just because you don’t understand the difference doesn’t mean you have any right to blame them for not being the people responsible for fixing your problems.
I am a developer, I hear these kinds of things from users too. It’s not just Linux, but users in general and a practical truth. Have you never received an issue filed against the UI for not showing certain class of products when they haven’t even been entered into the database? The correct business* response is to solve the problem not reply “Oh that’s not my job.”
Users don’t care where the issue is, they care that it exists at all. A user really doesn’t care where the issue is.
E.G. If there was a bug in Windows that made Firefox ( unusable even if it was coded against the proper of the behaviour of the API users won’t use it. Sure the Firefox developers can say that its not their job… but in that case people won’t switch from IE.
The fact that the user said “driver” is immaterial, what they mean is that their hardware is not supported, nor does a user want to shop around for specific hardware that is supported.
* While Linux is not a business per se, if it wants desktop share it needs to behave as one.
Edited 2008-04-08 22:21 UTC
It is necessary to caveat this heavily.
You wouldn’t try to run Vista on a Sony Playstation … the Playstation is not designed to run Vista.
Therefore, you only run Vista on hardware that is designed to work with Vista.
Likewise, you only can run OSX on Apple Macs.
So apply the same rules to Linux … only attempt to run Linux on hardware that supports Linux.
Therefore, shun hardware that is not designed to support Linux … such as Broadcom wireless chips. Do not buy.
Similarly Lexmark printers. Steer well clear of them.
Complaining at Linux for not supporting (say) Broadcom wireless chips (or this or that closed hardware device) is most decidedly the wrong way around … compared to the way that one evaluates the hardware support of the other two major desktop OSes anyway.
BTW … even counting such no-go hardware such as Broadcom wireless chips and Lexmark printers … there is still far more hardware that works with Linux than there is hardware that works with either Windows Vista or Mac OSX. Just to bring this discussion back to the ideal of “apples-with-apples” comparisons.
If it were a metter of complaining you’d be right.
Since many of us have the goal of Linux destroying Windows and coming to dominate the desktop market, we naturally want to do anything that furthers this goal. We advocate the production of (and write the software for) friendly end-user applications, installer routines, etc. so that Windows users may be converted with ease to Linux.
Understand please that outside the enthusiast and advocate circles users are efectively stupid. I mean no disrespect, but few purchasers of a Dell are going to be able to figure out what chip drives its wireless card, even if they knew enough to try and find out. You cannot tell these people “Shun Broadcom.” The only viable solution is to make sure that *no matter what they buy*, it works. The same goes doubly for those who have existing hardware.
Thus it is not *complaining* when people say that Linux sucks. What a user might mean is “I could not get my broadcom chip to connect to a WPA network” but what they *say* is “Linux doesn’t work.” If they complain to Canonical because they found this to be the case in Ubuntu, an answer of “We don’t write the kernel” would not be appropriate. The user is stupid, remember? he isn’t going to know how to post a bug report to the LKML, he doesn’t know that it’s really Broadcom’s fault, and he wont try to find out. He’ll just switch back to Windows.
Maybe you’re okay with that, I know a lot of people say that if you can’t make Linux work for you you should just stick with Windows (with a subtext of “You useless ignorant lunk.”) I, as a Linux advocate who would like to see it come to dominate Windows in the desktop market, do not want to see users switch back to Windows. I want to make sure that whatever a user tries to do he succeeds in doing. I don’t want to shun Broadcom, I want to get Broadcom wireless chipset support fixed.
Sure, but people from this driver project are not a business. They’re volunteers who want to help the world by writing kernel drivers – for free even! Complaining to them – as opposed to complaining about RedHat or whatever commercial vendor – about non-related issues, is nothing but rude and pointless. If you pay them to do their work, then you’d have something to complain about.
It’s like complaining to your neighbor that the cake he gave you yesterday doesn’t suit your taste. He’s *not* a business. The cake is a gift.
These people have also never claimed that they want desktop domination. They’re just trying to help via the best way they can, namely writing kernel drivers. The only thing they expect is some gratitude, but instead they’d have to deal with ungrateful people who apparently can’t stand them trying to help the community for free.
Why not take your complaints to your commercial vendor? You’re a software developer. Suppose some commercial vendor (say, “FooCompany, inc.”) takes your (free) software and sell it with commercial support. Would you feel that it’s justified if people came complaining to you about issues that are related to the commercial addons that FooCompany added to your software? Would you take the time to stay polite, and forward any and all complaints to FooCompany for the complainers, even though you don’t get paid for it? Frankly I have a hard time believing that you wouldn’t think “fsck these ungrateful freeloaders, show me some money, THEN we’ll talk”.
Edited 2008-04-09 07:46 UTC
You are confused. Linux is nothing but a kernel. And a user, as you describe him — not understanding and/or not caring about the underlying structure of his OS — has as much business knowing about Linux as he has knowing about the NTOS. He doesn’t even have to know that he runs GNU/Linux! He has, however, a business knowing his distributor, his vendor, and that’s the one he should talk to.
If you think you are talking about an OS when you say “Linux” you are wrong and you are doing other users a disservice in doing so, they get as confused as you already are. If you think Linux is something it is not, it is no wonder you are frustrated. You need to understand that you, as a user not understanding and/or not caring about the underlying structure of your OS, are talking to the wrong people if you talk to anybody else but your vendor/distributor.
“Linux” therefore does not want “desktop shares”, not even GNU/Linux wants “desktop shares”. Ubuntu wants “desktop shares”, SUSE wants “desktop shares”, any number of distributors want “desktop shares” and for the time being their mutual goal may be to gain “desktop shares” for any GNU/Linux based operating system. I won’t bet on it, though.
Ironically this kind of confusion is something Richard Stallman predicted years ago, albeit from a different angle, and now it comes apparently around to bite not only GNU but also Linux in the ass. Funny.
No. It’s just as unimportant and as much of a distraction as it always was. Like Jan 1, 2001 being the “real” first day of the millennium. It simply does not matter outside of casual dinner conversation. Richard’s penchant for obsessing on minutia (and obvious minutia, at that) has always seemed odd to me. Though not as odd as the way some people herald it as an uncanny ability to see into the future.
Let’s focus on real problems and challenges… and not end up off in the weeds, arguing trivia, and wasting time.
and yet my four year old usb webcam doesnt work at all, great project that is
Well then report it to the site listed in the article.
Um, video-input devices are specifically mentioned in the article as having poor driver support in Linux.
Care to name the brand and model of this four year old webcam of your’s ?
That Linux is one of the best with regards to the number of drivers is one thing, but just counting supported devices isn’t a quality measurement…especially when the drivers that are most needed (like video and network) are lacking.
Its getting there, but for it to be perfect the important hardware must be supported.
Delete
Edited 2008-04-08 15:59 UTC
The problem is individual experience. If your Printer doesn’t work when you tried, you don’t care that there are hundreds of working ones.
Also, some very important drivers work but suck. Graphic drivers so far are mostly pain. Even nVidias drivers are. Partly this is to blame on the binary interface catch up game they have to play. Luckily, this might change with Atis published specs.
We will see, but have patience with people annoyed because of driver problems. Try to convince them to research compatibility before buy. This is much more productive then being arrogant about them.
The real arrogance come from people who say it don’t work because they tried it.
If printer where easy there would not be a billion dollar industry that serve to sale and maintain them from many vendors …
Important driver don’t suck , Graphic driver are also excellent for the direct investment put in them.
When people are going to stop buying windows system and install GNU/Linux on it , then the vendor will start allocating more funds to there GNU/Linux solution.
It’s funny to see people try and make there old pentium 1 work with GNU/Linux and then fail , don’t ask for help and go buy a 1000$ – 2000$ Macbook or Macbook pro and say it’s perfect for them.
Or try there printer on Ubuntu witch they got for free and cry Ubuntu is crap and when they buy vista go and buy a new printer after the one they got don’t work with Vista.
Actually, your creating a myth. The real arrogance are people who hate testers .. aka people who tried drivers, and the drivers do not work.
That’s a personnal reviewer , a tester will go into long detail about what they did and where it failed and what materials they where using. They also report there finding directly to the developpers.
2 things if you don’t have enough projects to work on:
– suspend/hibernate ;
– ATI drivers. The specs are out, there’s enough already to keep a few devs busy
KernelTrap: http://kerneltrap.org/Linux/Separating_Suspend_and_Hibernation“>…
Edited 2008-04-08 19:45 UTC
Or you could co-ordinate with this new project:
http://www.via.com.tw/en/resources/pressroom/pressrelease.jsp?press…
… just announced 8th April.
I wish the same could be done on XGI cards. They released some source code but it seems it has not found its way into the trident driver ( I own a V3). Their new Z7,8,9 products are really open source but I could not find a discrete solution with them.
How about helping out the guys writing the sony-laptop driver? My laptop (VGN-FZ180E) bought summer 2007 is NOT well supported. No ways to control screen backlight. No power in the expresscard slot. Getting anything else working (like audio-out or general ACPI-stuff) requires hacks and workarounds.
Once again, just to be clear … your issue is that the Sony Vaio laptop you purchased does not support Linux (and not the other way around).
If you wanted to run Linux on a laptop, why did you buy a Sony Vaio which has such poor support for Linux?
http://www.zareason.com/shop/home.php
http://system76.com/index.php?cPath=28
http://www.linuxcertified.com/linux_laptops.html
http://www.linux-laptop.net
Many models from HP, Acer, Lenovo or Dell look good.
Why are you blaming the user? This attitude is very damaging.
The fact that his hardware does not work under Linux is not his fault, nor is it the fault of Linux developers. Refusing to improve the situation, however, is not a reasonable answer. If a Sony Vaio laptop back light does not work, why not make it work? Then more users will use Linux. Is it wrong to try and add support for a piece of hardware just because the manufacturer doesn’t care?
Incidentally, your argument is semantic only. “Does not work” and “Is not supported” are synonymous and often used interchangeably. Exactly who is not supporting what is entirely immaterial.
I apologise for this being a slight tangent to the effort described.
I bought a Logitech RumblePad 2 (USB) yesterday. I plugged it in and it was recognised as a joypad, all the buttons on it worked in games that supported joysticks (which was a fair few). Wow! Things have moved on a long way from when I started on Linux. Part of this is probably due to more standardised interfaces (what with USB providing standard specs for various devices including game controllers). Part of this is also due to Linux driver support (and distro handling of new devices!) being improved tremendously.
It’s not been a 100% perfect experience; rumble support doesn’t work (not that many Linux games support that, as far as I can tell). I don’t think the kernel recognises this device as being force feedback capable (and Ubuntu’s kernel doesn’t compile in the – experimental – force feedback support). I can’t do much about Ubuntu not supporting this by default, but since I’m a kernel hacker by trade I intend to see if I can get the device supported. I’m hoping it uses the same ff protocol as other Logitech devices and that I can just add it to a list of supported devices…
Anyhow, bottom line: Linux hardware support is impressive these days but not perfect. Given I have the skills, it’s satisfying to be able to dip in and have a go at fixing it. If it’s a simple tweak, I imagine it’ll be coming to the user community in a kernel release soon. That’s actually kind of awesome