The OSS Solution to the Linux Wi-Fi Problem

tobs writes "Matt Hartley of MadPenguin.org fame has published an open source way of solving the Linux Wi-Fi problem. He writes, "For intermediate to advanced users, who are willing to track down WiFi cards based on chipsets, live without WPA in some instances or have opted to stick with Ethernet, buying a new notebook for the sake of improved wireless connectivity may seem a little overkill. When a new user faces problems jumping through the NDISWrapper hoops, tracking down WiFi cards from HCLs and other related activities, the end result is almost always the same — they give up. What so many of us, as Linux users, fail to grasp is that projects like OpenHAL are critical to long-term development. The education on what to expect and what not to expect remains a complete load of hot air when articles claim how easy it is to setup wireless Internet on Linux machines. It's downright misleading."

What's the REAL Solution though?

[eldavojohn]

The OSS Solution to the Linux Wi-Fi Problem

I'm confused, this article did not propose an OSS solution for the Wi-Fi problem at all. In fact, it just told me things I already know from first hand experience that I've posted about before.

What confuses me so much (and I really am ignorant in this department) is why the ethernet chipsets were seemingly conquered right off the bat? I tried my first Linux distro (debian) in 2001 and ever since then no matter what the machine, no matter what the distro, no matter how confused I was the NICs always came up ready to go when I installed Linux. I've done this on a lot of machines, from obscure to well known Dells and used most of the major distributions. They just 'worked' and it was good.

Now, wireless is here and for some reason, there must be a thousand different manufacturers with their own proprietary chipsets with completely different drivers & BIOS data on the flash memory stored in those chips because I've only had Ubuntu work once out of the box on a Linksys PCI WiFi card. Why? Why isn't that standardized? What do the companies gain from that? Is it because of the ever changing standards that the chips are so wacky? Is it because the A, B, G, N, etc. protocols? I don't understand this because I've never coded drivers.

I understand what MadWiFi & OpenHal are trying to do. I now know to look for "Atheros" chipsets when I buy my wireless stuff but they are often more well known brands and more expensive. A reason I switched to Linux was to save money in college, not spend more on the hardware.

Maybe a more helpful article would be detailing the real underlying issue--that these no name brands that get huge rebates at CompUSA or where ever (Hawking Technologies, generic boxes, etc.) are targeting Windows because of the number of users. How do you change their minds or show them a market for an OSS driver? Is there a way to even open up a channel of communication with them to discover how to write drivers for their chispets? How do you convince them it's worth their time/resources?

That would be a solution moving forward.

The next best thing would be to post an article about how to get started making these drivers. I'm a coder (though not the greatest one) with a little bit of free time. How do I start? How do I get access to the BIOS pages on the chipsets? What do I do with that, how does the Linux kernel use it? What books do I read that teach me how to start with a chipset I know nothing about, have no resources on the data or mechanics and then poke it, prod it until I know enough about it that I can set it up for the kernel to use it?

Re:Scientist's viewpoint

[Walpurgiss]

I remember feeling that way around when the internet was gaining traction. It was so hard to find a linux compatible modem in stores since almost everything at the time was a winmodem piece of trash that let windows control everything and had almost no on board processing. I couldn't believe how many hardware vendors wouldn't be bothered to make standalone modems, instead opting for the cheaper windows only idea. Though with my current laptop I got lucky, had an atheros chipset that was supported by madwifi. Took some tooling around to get WPA-PSK to work; but it's ok now.

Re:What's the REAL Solution though?

[jsupreston]

I don't understand why there isn't a "fallback" like we used to have on Ethernet NICs. For many years, it seemed like if you couldn't get a NIC to work, you could always use the old NE2000 drivers. You might not have all the functionality of the proprietary drivers, but it would at least get you on the network. Why can't we do the same with other hardware? Heck, we don't even have that fallback anymore with PCI NICs, so you're screwed if your setting up a machine with a NIC not recognized by the OS out of the box and you don't have drivers for it.

Buying a new laptop? No way!

[rg3]

My laptop came with a Broadcom 4318 chipset. The support for it is flacky and it only seems to work properly using ndiswrapper. Some days ago I decided I was going to try to buy a USB wifi device that was compatible with Linux. If possible, its drivers had to be already part of the vanilla kernel. To my surprise, those devices exist! They are the ones that have the ZyDas zd1211(b) chipset (the "b" one is better). I thought it was going to be hard to find one of those specific devices, but no. They are present in a wide range of USB wifi devices. I went to the two main malls in my country they had one of those devices each. Piece of cake. Furthermore, a USB dongle can be used in future computers very easily, and don't take power unless they're plugged in.

http://linuxwireless.org/en/users/Drivers/zd1211rw/devices [linuxwireless.org]

You only need the device, a vanilla kernel and firmware, which can be downloaded from SourceForge.net, and it's also probably available for your distribution as an official package.

http://sourceforge.net/project/showfiles.php?group_id=129083&package_id=187875 [sourceforge.net]

Re:The all OS solution

[Technician]

Since i have a Medion Mim with some proprietry medion chipset, i'm stuck with no wireless for ubuntu.. :(
the driver doesnt work with NDIS wrapper either...
of course...
that doesn't stop me using debian stable on my desktop!

It's not pretty, but I found a solution which even works with old Windows 95 (for testing) and enables full WBA encryption. It works on any OS that can use the wired NIC in a machine. Are you ready...

Use an access point which is capable of Client Mode operation. I use a D-Link AP in client mode. I configure it with my browser. It requires no software install of any kind. Testing was done on the D-link AP and now a Linksys 54G router has been added to my travel pack because it cost less (lucky find, a version 4 for $12 at Goodwill).

I have been running wireless with an AP in client mode since Breezy Badger. Upgrading the firmware to DD-WRT has added the client mode. As a bonus, you get to use high gain antennas with much better range than a stock laptop provides, and the power is adjustable for use in poor signal locations. The router does the site survey for you internally, so you don't even need to know the SSID ahead of time. It is as simple as switching to either client or client bridge mode, scanning, choosing an AP, and picking the encryption and entering the key. After that it's net, nothing but net.

There are hardware solutions out there. The package may be a little big and bulky and not run on self contained batteries, but it provides excellent connections in hotels in marginal reception areas. With the external box, it can be positioned in a window where the neighbors open AP may provide better bandwidth than the hotel provides. I went to a Starbucks once not knowing the wireless wasn't free (T-mobile). I was able to find 2 unsecured APs from inside Starbucks to use instead. Nobody at Starbucks was the wiser. It beats getting busted for sitting in a car leaching on some residential street.

Re:What's the REAL Solution though?

[ncc74656]

So, in short, wireless drivers are hard because wireless cards are really complicated.

Here's an idea: on one card, combine a wired network controller (RTL8139 or whatever) with a wireless bridge. It'd be like plugging a bridge into the network jack, but everything's already in one place. Conceptually, it's not much different than the way internal modems used to be built: combine a serial interface and the guts of an external modem on one board. Just as the internal modem appeared to your computer as just another serial port (that happened to have a modem hanging off of it), this would appear to your computer as just another wired NIC (that happens to have a wireless bridge hanging off of it). Configuration can be done with a web interface.

It'd cost a few bucks more for the wired network controller, but compatibility with Linux would be guaranteed.

I Call BS!

[Perl-Pusher]

Wireless on windows isn't exactly flawless either. Sometimes you get windows wanting to control it over the manufacturers software and it doesn't work. Until SP-2 on XP doing 128bit WEP was a hit or miss proposition. And being a systems admin, I've seen a lot of people who are unable to connect to a wireless network using windows. I end up doing it for them. Also, windows can for some reason forget a network profile that it has previously had no problems with. OS X seems to be better at it but it too can suffer from dumb user syndrome. You also have a problem with windows blocking its own connections. And then there is the 3rd party pre-requisite anti-virus / network security suites that block outgoing traffic and want to examine every packet and then ask you if you want to allow. And last but not least. When you buy network cards sometimes they conflict with other chipsets. I have had cards that conflicted with the sound system. Easily fixed by changing settings in device manager but windows will change everything back to the setting that doesn't work next reboot.