[LUGSB] Wireless card

[Chris Wright]

The ndiswrapper potentially allows you to block some functionality
from a driver (such as touching unrelated devices). A native driver
doesn't give you that, as far as I know.

Other than that, I agree with everything you are saying. I have an
annoying habit of arguing with people, especially when they give an
assertion without providing any reasons to believe it.

On 07/02/07, Ilya Sukhanov (dotCOMmie) <lugsb at sukhanov.net> wrote:
> Chris Wright wrote:
> > On 8 Feb 2007 01:40:42 -0000, Ilya Sukhanov (dotCOMmie)
> > <lugsb at sukhanov.net> wrote:
> >>
> >> On 2/7/2007, "William Joseph Lahti" <wlahti at ic.sunysb.edu> wrote:
> >>
> >> >Avishay,
> >> >
> >> >Have you tried working with ndiswrapper to get your current wireless to
> >> >work?
> >>
> >> Ndiswrapper is evil and should be avoided at all costs.
> >
> > What's evil about it? It's a security concern, if a minor one, since
> > any untoward code probably relies on the NT kernel and wouldn't be
> > supported by ndiswrapper (Linux-native binary blobs are a much greater
> > concern); it's a stability concern, though there are enough people
> > using each wireless card that that is probably minimal. As a practical
> > matter, it's a low-risk but uncertain means of achieving important
> > functionality.
>
> Linux wifi drivers provide more functionality than the windows ones; rfmon,
> master.. etc. Sure some people don't need these things, but they are nice to have.
>
> You differentiate between Blobs and the windows drivers. Technically what is the
> difference? and why would Linux blobs be of more concern? Both run in kernel
> space and you have no idea what they do, how is one more secure than other?
> While all blobs sucks, smaller blobs are better (from a functionality
> perspective, security wise they are the same) than larger blobs because it
> decreases the number of bugs which cannot be fixed by 3rd parties.
>
>
> > Politically, if you care about that (not everyone who uses Linux cares
> > about OS politics), ndiswrapper is a step away from corporate hardware
> > support for Linux, but a step toward a universal driver format and a
> > means of increasing the Linux userbase.
>
> I do care for politics of the issue, but I am not radical. I am simply
> discouraging users from settling for options which limit their rights. They
> should be aware that some things are not nice. Beyond that they can do their own
> research and draw their own conclusions.
>
> Step away from corporate hardware support? What do you mean? There are good,
> corporate linux drivers.
>
> Are you suggesting we should use windows as a "standard" for technologies?
> Microsoft is notorious for crappy solutions. In addition, does it make sense to
> tangle ourselves in the microsoft patents?
>
> Its not that there is a lack of linux developers where we need universal drivers
> if companies just released specs we wouldn't need ndiswrapper. Ndiswrapper gives
> companies a weapon, if they don't want to release specs they can say "Why should
> we, our hardware works under linux.. use ndiswrapper"
>
> > Religiously, of course, you are entirely correct. A modern Buddha
> > would (and indeed has!) declared that binary drivers are the work of
> > Angra Mainyu, the destructive spirit. Not to mention the opinions J.R.
> > "Bob" Dobbs. Though according to the prophet of Enlil, the open source
> > movement is a source of moral decay, if not a notable one.
> > Unfortunately, since Emperor Norton I of America is dead, we are
> > unlikely to achieve a definitive answer to this conundrum.
>
> I am not sure I understand, especially the Enlil reference.
> _______________________________________________
> lugsb mailing list
> lugsb at mail.fsl.cs.sunysb.edu
> http://www.fsl.cs.sunysb.edu/mailman/listinfo/lugsb
>


-- 
Some people are like Slinkies: not really useful, but it still brings
a smile to your face to push them down a flight of stairs.