(Note that I'm not on orinoco-users, so if you don't reply to me
directly, I'm not going to see it on list.)
> RG> Dragorn is actually quite scathing of the 0.15 driver
> RG> chain:
> RG> http://www.kismetwireless.net/Forum/General/Messages/1137431697.4090331
> RG> But I agree with Tomas, 0.15v4 seems to run ok with
> RG> firmware 6.16.
>
> Dragorn don't like 0.15 driver - he has got his own unconfirmed
> objection without any reason. 0.15 drivers works with Kismet much
> better than 0.13 ones. If dragorn will drop 0.15 support in next
> version of Kismet, we will simply use older version of Kismet.
I'm not dropping support of anything. If you're talking about the fact
that it isn't in newcore yet, I'll point out that there hasn't been a
RELEASE of newcore yet, and that there are LOTS of cards which aren't in
newcore yet. Primarily, every card I don't use a day to day basis or
which didn't have interesting conversion issues when I moved the capture
code over. And, hey, guess what -- It's open source! Want support in
newcore faster? Write a patch and send it to me.
As far as my opinions of 0.15 -- For me, they're unusable in rfmon, and
anyone who wanted to talk to me about it would have gotten a more
explicit list of why. They appear to drop dead about 1 minute into
capture. The if is up, it's still configured, blah blah blah, it just
stops sending packets and the drivers deadlock themselves. It happens
with every firmware I've tried, it's happened to other people, it's
been reported, and no-one has been able to track it down as far as I
know.
At this point, I consider it fairly known behavior, and me whining about
it on the list all the time wouldn't do any good, though it'd probably
annoy people a fair bit. I personally haven't gotten any time to try to
dig into the drivers myself and set some debugging and see where it thinks
it gets stuck, but, then, neither has anyone else with the same
problems, so it's the tragedy of the commons.
Sometimes it works fine. Works fine for you? Congrats. Be happy.
I *have* dropped support for patching in per-packet stats to the 0.15
chain, and I *have* dropped trying to up-port the 0.13 drivers which
just wasn't worth my time for a card I didn't use as a primary. No-one
else has been interested in taking over up-porting them, so I figure
no-one cares.
Once either a) the radiotap/ieee80211 stuff goes fully mainline, or
b) the new direction for the kernel stack is decided, I'll write a
radiotap patch for the 0.15 drivers and send it along, since the RT
headers address all my per-packet needs and the previous arguments about
including or not including the FCS bytes.
So:
* Yes, in my experience, the 0.15 drivers don't work.
* AFAIK it's been reported a long time ago.
* They work for you, congrats. They don't work for lots of people, so
I don't recommend them in general.
* Want it in newcore now? Send me a patch. Otherwise it'll be added
sometime before the actual release.
* What was the point of all this?
-m
--
Mike Kershaw/Dragorn <dragorn-8I8a9W+rkIX1ZkIrOfZKgqxOck334EZe@xxxxxxxxxxxxxxxx>
GPG Fingerprint: 3546 89DF 3C9D ED80 3381 A661 D7B2 8822 738B BDB1
"Please, let me sleep. I'm very tired, and I think one of those needles
broke off in my headmeats."
-- Filler Bunny, the genetically engineered bunny of happiness
pgps3ED6RvlgP.pgp
Description: PGP signature
|
