[conspire] "madwifi" is proprietary sludge (was: driver)

[Rick Moen]

Quoting Daniel Gimpelevich (daniel at gimpelevich.san-francisco.ca.us):

> No, it's really more dire than that.

Since I'm having a busy day, I'll have to get back to you on this post.
I suspect we're having a communications gap.

> I specifically said "among 802.11g chipsets" here, and you pretty much
> confirm that by having to go outside that category to find a tolerable
> alternative. 

A bunch of the Intersil chips supported by open-source Prism54-family
drivers (what I meant when I said "Prism" for short, before) are
802.11g.  As is Intel 2200.

Not that I expecially care, personally -- as 802.11a/g strike me as
solutions in search of a problem, and I'm perfectly happy for now with
what I have.  Which also means that I don't bother to remain fully
current on 802.11a/g hardware and drivre developments:  They're not of
near or medium-term relevance to me.  

Kind of like Flash interpreters.  ;->
 
As to Intel 2200 being incapable of some operations at a hardware level:
OK, if so, don't use them for that role.  This has no visible relevance
to our preceding conversation, your (erroneous) claims that I was
conflating hardware-based and software-based limitations
notwidthstanding.


> Although that is also true, what I was trying to say was that given more
> time, any specific stage 1 hardware gets more opportunity to reach stage
> 2, at which point it ceases to be a comparative advantage over that same
> specific hardware.

You know, I don't buy it.  Reverse-engineering isn't materially aided in
any way by the presence of Linux drivers that may not be lawfully
examined or probed, and manufacturer end-of-life schedules are pretty
much determined by other considerations entirely.

I'm not sure whatever other than that you could mean.  If I'm not
following your point, my apologies -- but, at this point, I need to move
on.


> At this time, the most tolerable 802.11g chipset is Atheros....

Sorry, no.  Atheros would be the Broadcom of wireless, i.e., a sad
example of how to screw up by the numbers with the Linux community, if
Broadcom hadn't already usurped that spot.

And, no, your model of people flocking to proprietary sludge and thereby
building a groundswell of support for open-source alternatives is just
not how it works.

 
> It may not have been much of an itch, but the itch did exist, and largely
> disappeared with Macromedia's initial release. 

Revisionism.  Hardly anyone in open source has _ever_ given a rat's ass
about Flash.


> ATI, as supported by the open-source drivers, is more tolerable than
> either Matrox or Intel.

Not unless you're a framerate freak, e.g., a gamer, which I'm not.
For one thing, Matrox has much better image quality (which the gamer
kiddies disregard).  Intel 945G has decent open-source 3D, and more than
acceptable other characteristics.

> The Intel [945G) driver isn't quite open-source,

My information suggests you are in error.  Do you have evidence to
support your claim?

> and neither are the drivers for anything comparable from Matrox.

If by "comparable" you mean with fast-framerate 3D, as I've abundantly
noted in many places, I really don't give a damn about that.  All the 
patent-encumbered proprietary solutions in this area can go hang, as far
as I'm concerned.

I _know_ Matrox cards.  They rock, for all purposes I care about.