Intel Linux Driver Version 1.0 For Centrino WLAN

Werner Heuser writes "Intel has just released Linux driver version 1.0.0 for the Centrino PRO/Wireless 2100 chipset. Almost two years, since the first Centrino generation with Pentium-M Banias has reached the market, a stable issue of a native Open Source driver has become available. The Wireless LAN driver for the current Centrino generation with Pentium-M Dothan and PRO/Wireless 2200BG chipset is still at version 0.13. This driver is intented to support also the third generation of the Intel miniPCI WLAN adapter named PRO/Wireless 2915ABG. Though Intel intended these projects to be community efforts, there are some possible working constraints. Mainly, no hardware documentation is available."

Comments

So put the documentation

[Rares Marian]

task in a job on rentacoder.com. er so I can bid on it.

Excellent job

[kompiluj]

Well, Intel is doing excellent job, unlike nVidia and ATI. And as to the docs, well I think that Intel will have nothing against looking at them, ehm, over the shoulder, you know.

Re:Excellent job

[Stevyn]

How many times does this have to be brought up? NVIDIA can't release all the information because they don't own it.

Re:Excellent job

[nathanh]

How many times does this have to be brought up? NVIDIA can't release all the information because they don't own it.

Do you mean, how many times does that nonsense argument have to be brought up?

They could release the information they do own. We don't need "all the information". We just need the basics like opcodes and register addresses. That information is not patentable (they're just numbers) and is not copyrightable (not creative). That information is simply a trade secret that nvidia CHOOSES to not give us, presumably because they're *ssholes. Nvidia can keep their precious licensed code; that's not the stuff we need to write our own driver.

Instead, nvidia chooses to release NOTHING and hide behind this furphy of "we don't own the information". What's worse is that people such as yourself are now repeating nvidia's furphy, saving them the effort of having to lie directly to us! Goebbels spirit lives.

Re:Excellent job

[TiggsPanther]

They could release the information they do own. We don't need "all the information". We just need the basics like opcodes and register addresses. That information is not patentable (they're just numbers) and is not copyrightable (not creative). That information is simply a trade secret that nvidia CHOOSES to not give us, presumably because they're *ssholes. Nvidia can keep their precious licensed code; that's not the stuff we need to write our own driver.

Well in the updates to the interview that was previously mentioned here they do state that it's to do with the IP involved.
Plus the real issue probably isn't with the driver developers, probably more to do with management or even with IP lawyers.

Also it matters very little as to whether the trade secrets really need to be kept secret. I suspect that someone in their legal (or financial) department is saying that stuff "like opcodes and register addresses" are things that they can't release for free - even though that's information that they do own. And, sadly, they're the ones whose decisions tend to stick.

Yes, it sucks. But unfortunately big businesses still like to play the "Intellectual Property" game. Until that changes (Soon, please God, soon) then we're stuck with things happening this way.

Re:Excellent job

[aminorex]

No, there's a simple solution: Make all your RFPs specify criteria that exclude NVidia products. That's what I do. When enough companies do this, NVidia will fold.

Re:Excellent job

[c0p0n]

We just need the basics like opcodes and register addresses. That information is not patentable (they're just numbers)

Yeah, and chemical formula is no more than several letters and digits...

Re:Excellent job

[LWATCDR]

So if all your RFPs exclude NVidia and ATI what graphics cards are you left with? Matrox?
Don't get me wrong. I would love to see open drivers from NVidia but just excluding them from bid will do very little to help. Atleast they are providing free as in beer drivers for Linux instead of tell Linux to go suck wind.

Guess I'll have to upgrade

[swillden]

... when I have a few minutes with nothing better to do.

I've been using the 0.55 version of the ipw2100 drivers very happily for quite some time now. And I only upgraded to them from the 0.44 version because of an issue related to ACPI suspend/resume.

My hat's off to Intel for doing a great job on these drivers. Even the beta versions have been extremely usable for quite some time.

I'm posting this with my IBM Thinkpad T40, via WiFi using the ipw2100 version 0.55 driver on Linux 2.6.9. It's how I do all of my work (or non-work, as the case may be) these days.

Re:Guess I'll have to upgrade

[arodland]

Yeah, the drivers aren't bad, but it's been my experience that if they work for you, and nothing in the changelog looks that interesting, you might as well stick with what you've got. They've done their share of accidentally breaking things :)

Intel is really a friend to Open Source

[infonography]

And I am not saying that as a joke. However I wished they has done this three months ago when I went out and bought a new card because my internal 2100 wasn't Suse friendly. Oh Well still got it.

Re:Intel is really a friend to Open Source

[Stachel]

Right on time. I messed up my Centrino-based SUSE laptop yesterday evening. The previous version of ipw2100 was instable and kept restarting (PAYLOAD PACKAGE TO LARGE messages in system log). Tried ndiswrapper and that caused networking to fail altogether.
I hope this version '1.0' fixes the payload issue.

Re:Intel is really a friend to Open Source

[infonography]

I kept getting PC Load Letter error. WTF is PC Load Letter?

No Don't answer that, I am kidding you would know if you saw Office Space. BTW, ndiswrapper Blows.

Hardly Free

[cjsnell]

Use of this driver requires that you download Intel firmware which is covered by a very non-free and restrictive license.

I urge you to write to Intel and let them know that you are dissatisfied with their license and that you want the ability to freely distribute their firmware.

Please note that nobody is asking them to open the source of their firmware--they just need to make it so that free operating systems can distribute their firmware without having to force their users to agree to this licensing.

Re:Hardly Free

[bconway]

I, as a home and corporate user, have no reason to redistribute their firmware. I will not be mailing them, sorry.

Re:Hardly Free

[iCEBaLM]

It is however in your best interest. If you write them in support of packagers who want to freely distribute the firmware then that allows you the ability to install a Free OS (Linux, OpenBSD, etc) and have your wifi work "out of the box" instead of having to download the firmware from intels website by some other means separately.

Re:Hardly Free

[arodland]

The folks at intel have said that their license makes it perfectly fine for distribution, but it's not realistically true, unless your distro's hardware-detect script pops up a window when it notices the card telling you that you're using firmware under a license, that you only have permission to use it in conjunction with Intel hardware, and that you won't distribute, sell, copy, disassemble, or yadda it.

Apparently they think that would work just fine. But really it's a load of crap. I'm using an ipw2200, and I live with the personal-use license, but the distributor license is an absolute beautiful load of bullshit.

Re:Hardly Free

[arodland]

Actually, this might be the best solution. Do all the work to integrate the whole ugly license-management system into distributions' package-management systems, tie it into the hardware support, and have the license-verify wizard be a great technicolor flashing skull-and-crossbones effect, MovieOS style, with 3D surround-sound sirens, and a digitized voice saying "WARNING, WARNING, RESTRICTED ACCESS SOFTWARE. ACCEPT LICENSE TERMS OR YOUR COMPUTER WILL REBOOT IN 30 SECONDS.... 20 SECONDS..."

Sorry, but Intel Sucks

[DesScorp]

Go one way or the other. Either get us a proprietary driver that works RIGHT AWAY (as the article said, the Centrino chipset has been around for two years already. Microsoft wouldn't have stood for this kind of stonewalling), or make the drivers REALLY open source, firmware too. Intel has dicked us around for two years, and we're supposed to be grateful now? This is why I won't buy a Centrino book.

Re:Sorry, but Intel Sucks

[Txiasaeia]

What's your alternative, an Athlon notebook? My IBM X31 1.6GHz (Banias) notebook gets around 5 hours of battery life, 6 if I push it. It can do this running cold. Compare this with an AMD notebook that gets what, an hour and a half worth of battery life?

Rant and rave at Intel all you want - the Pentium-M is the best mobile processor ever made, full stop. Don't hate the hardware because you don't like the maker's policy towards OSS.

Re:Sorry, but Intel Sucks

[DAldredge]

No, chips derived from the ARM line of CPU's are some of the best mobile CPUs every made, it just that the Pentium-M is, currently, the best performing, in W/per unit of work, mobile CPU that runs windows.

Different CPU's are made for different things and blanket statements are almost always wrong in regards to CPU's

Re:Sorry, but Intel Sucks

[empaler]

This was actually the only reason I never kept a *nix on my portable (which I bought january) - no wifi support, and I didn't want to reboot all the time...

Re:Excellent

[Anonymous Coward]

The Windows file isn't Free. But neither is Intel's solution for Linux. Read the terms. It is proprietary, unFree software.

Marginally better than Windows software... mayyyybe. Maybe.

Sometimes I wonder

[agraupe]

Why don't these people just use ndiswrapper? I used it with this precise chipset. Don't worry that the Linux police will take your computer away if it has non-free software on it! If free software causes me undue hardship, I just use non-free. Why does this concept trouble so many people?

Re:Sometimes I wonder

[ssimontis]

Ndiswrapper is great for a lot of people. But the average Joe Linux user might not want to play around with his system to get it to work. A lot of people forget about the average Joe Linux user these days, but he is out there. Drivers with Windows can be, in some cases, simpler. You put the hardware in your computer if it isn't already, insert a CD, copy the drivers, and your ready to go. Not all drivers work like that under Windows, but some do. I want to see that on Linux. We need to see some other companies release drivers for Linux. Providing a fair license agreement is important, but I would gladly accept a somewhat restrictive license agreement for my WMP11 if I could switch to Linux and get the Microsoft crap off my computer for good.

Re:Sometimes I wonder

[Anonymous Coward]

yes but the average joe users neighborhood geek can do it.

taht is the person installing the windows drivers anyway. joe isnt

Re:Sometimes I wonder

[safecracker]

I agree some Linux users just want a out of the box working system but the fact is Linux has always been a OS for ppl who know what their doing when it comes to software (SuSE and the like have been trying to change this). ndiswrapper (and DriverLoader)although great solutions for many people are not for everyone. ndiswraper is still fairly new and some people feel weary about using it.

Linux driver wireless feature table

[183771]

Check out how compare Intel wireless driver with other Linux wireless drivers. Linux Wireless Drivers