[update] xfce4-wavelan

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

[update] xfce4-wavelan

Landry Breuil-3
Hello,

one new upgrade which needs testing, xfce4-wavelan. wi_bsd.c has been
hacked to work with OpenBSD's ieee80211_* structs, so i'd really
appreciate feedback on this one. It works fine here with my iwi(4), and
i'll try to send the cleaned patch (with #ifdef) upstream for their next
release.

http://ports.gcu.info/doku.php/openbsd/x11/xfce4/xfce4-wavelan

It was really fun to learn how our wireless stack works :)
Thanks for testing/comments.

Landry

Reply | Threaded
Open this post in threaded view
|

Re: [update] xfce4-wavelan

Lars Hansson
Landry Breuil wrote:
> one new upgrade which needs testing, xfce4-wavelan. wi_bsd.c has been
> hacked to work with OpenBSD's ieee80211_* structs, so i'd really
> appreciate feedback on this one. It works fine here with my iwi(4), and
> i'll try to send the cleaned patch (with #ifdef) upstream for their next
> release.
>
> http://ports.gcu.info/doku.php/openbsd/x11/xfce4/xfce4-wavelan

For my ural device it shows a signal strength of 20% instead of 100%
when signal is 20dB.
ural0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
         lladdr 00:02:44:b5:ad:f1
         groups: wlan egress
         media: IEEE802.11 autoselect (OFDM48 mode 11g)
         status: active
         ieee80211: nwid unet-makati chan 6 bssid 00:0f:3d:0d:eb:ac 20dB
nwkey <not displayed> 100dBm
         inet6 fe80::202:44ff:feb5:adf1%ural0 prefixlen 64 scopeid 0x4
         inet 172.16.2.60 netmask 0xffffff00 broadcast 172.16.2.255


---
Lars Hansson

Reply | Threaded
Open this post in threaded view
|

Re: [update] xfce4-wavelan

Landry Breuil-4
2007/6/6, Lars Hansson <[hidden email]>:

>
> Landry Breuil wrote:
> > one new upgrade which needs testing, xfce4-wavelan. wi_bsd.c has been
> > hacked to work with OpenBSD's ieee80211_* structs, so i'd really
> > appreciate feedback on this one. It works fine here with my iwi(4), and
> > i'll try to send the cleaned patch (with #ifdef) upstream for their next
> > release.
> >
> > http://ports.gcu.info/doku.php/openbsd/x11/xfce4/xfce4-wavelan
>
> For my ural device it shows a signal strength of 20% instead of 100%
> when signal is 20dB.
> ural0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
>          lladdr 00:02:44:b5:ad:f1
>          groups: wlan egress
>          media: IEEE802.11 autoselect (OFDM48 mode 11g)
>          status: active
>          ieee80211: nwid unet-makati chan 6 bssid 00:0f:3d:0d:eb:ac 20dB
> nwkey <not displayed> 100dBm
>          inet6 fe80::202:44ff:feb5:adf1%ural0 prefixlen 64 scopeid 0x4
>          inet 172.16.2.60 netmask 0xffffff00 broadcast 172.16.2.255



Mhh, i may have mixed signal strength/SIOCG80211TXPOWER, and rate/RSSI in my
patch (patch-panel-plugin_wi_bsd_c)... i'm not used to wireless keywords :)
I'll have a look at it and try to poke upstream to merge the correct
changes..

Thanks for the report,
Landry
Reply | Threaded
Open this post in threaded view
|

Re: [update] xfce4-wavelan

Kian Mohageri-2
On 6/6/07, Landry Breuil <[hidden email]> wrote:

> 2007/6/6, Lars Hansson <[hidden email]>:
> >
> > Landry Breuil wrote:
> > > one new upgrade which needs testing, xfce4-wavelan. wi_bsd.c has been
> > > hacked to work with OpenBSD's ieee80211_* structs, so i'd really
> > > appreciate feedback on this one. It works fine here with my iwi(4), and
> > > i'll try to send the cleaned patch (with #ifdef) upstream for their next
> > > release.
> > >
> > > http://ports.gcu.info/doku.php/openbsd/x11/xfce4/xfce4-wavelan
> >
> > For my ural device it shows a signal strength of 20% instead of 100%
> > when signal is 20dB.
> > ural0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
> >          lladdr 00:02:44:b5:ad:f1
> >          groups: wlan egress
> >          media: IEEE802.11 autoselect (OFDM48 mode 11g)
> >          status: active
> >          ieee80211: nwid unet-makati chan 6 bssid 00:0f:3d:0d:eb:ac 20dB
> > nwkey <not displayed> 100dBm
> >          inet6 fe80::202:44ff:feb5:adf1%ural0 prefixlen 64 scopeid 0x4
> >          inet 172.16.2.60 netmask 0xffffff00 broadcast 172.16.2.255
>
>
>
> Mhh, i may have mixed signal strength/SIOCG80211TXPOWER, and rate/RSSI in my
> patch (patch-panel-plugin_wi_bsd_c)... i'm not used to wireless keywords :)
> I'll have a look at it and try to poke upstream to merge the correct
> changes..

I'm not very familiar with wireless either, but the 80211TXPOWER is
the 100dBm shown, while the 20dB is the RSSI (I thought).  Could
anybody explain to me why it is incorrect to show 20% given the output
Lars has provided?

Kian

Reply | Threaded
Open this post in threaded view
|

Re: [update] xfce4-wavelan

Stuart Henderson
On 2007/06/07 14:46, Kian Mohageri wrote:

> On 6/6/07, Landry Breuil <[hidden email]> wrote:
>> 2007/6/6, Lars Hansson <[hidden email]>:
>> >
>> > Landry Breuil wrote:
>> > > one new upgrade which needs testing, xfce4-wavelan. wi_bsd.c has been
>> > > hacked to work with OpenBSD's ieee80211_* structs, so i'd really
>> > > appreciate feedback on this one. It works fine here with my iwi(4), and
>> > > i'll try to send the cleaned patch (with #ifdef) upstream for their
>> next
>> > > release.
>> > >
>> > > http://ports.gcu.info/doku.php/openbsd/x11/xfce4/xfce4-wavelan
>> >
>> > For my ural device it shows a signal strength of 20% instead of 100%
>> > when signal is 20dB.
>> > ural0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
>> >          lladdr 00:02:44:b5:ad:f1
>> >          groups: wlan egress
>> >          media: IEEE802.11 autoselect (OFDM48 mode 11g)
>> >          status: active
>> >          ieee80211: nwid unet-makati chan 6 bssid 00:0f:3d:0d:eb:ac 20dB
>> > nwkey <not displayed> 100dBm
>> >          inet6 fe80::202:44ff:feb5:adf1%ural0 prefixlen 64 scopeid 0x4
>> >          inet 172.16.2.60 netmask 0xffffff00 broadcast 172.16.2.255
>>
>> Mhh, i may have mixed signal strength/SIOCG80211TXPOWER, and rate/RSSI in
>> my patch (patch-panel-plugin_wi_bsd_c)... i'm not used to wireless keywords :)
>> I'll have a look at it and try to poke upstream to merge the correct
>> changes..
>
> I'm not very familiar with wireless either, but the 80211TXPOWER is
> the 100dBm shown, while the 20dB is the RSSI (I thought).

if that figure was genuine you would most likely not be posting
to misc@ right now

> Could anybody explain to me why it is incorrect to show 20% given
> the output Lars has provided?

dB != %

dB is unitless, it's a measure of one power level in relation
to another power level. dBm is in relation to 1mW (0dB=1mW, 3dB=2mW,
6dB=4mW etc).

The units used by various wireless drivers are... inconsistent.
if you tcpdump radiotap (-yIEEE802_11_RADIO) ath reports like
'RSSI 33/64' which ifconfig shows as %; ral reports 'signal 48dB'
which ifconfig shows directly; wi doesn't do radiotap but the
one I tried shows the network in ifconfig around 20dB. This is
all in the same location, just switching cards. fwiw wispy-tools
peaks around -45dBm.

Given all this I think the only thing you can do is come up
with a number and print it, trying to convert to some % is not
likely to give a meaningful result. For a given card, numbers
changing in a certain direction indicate a stronger signal,
but I think that's about all you can really count on.

Reply | Threaded
Open this post in threaded view
|

Re: [update] xfce4-wavelan

Lars Hansson
In reply to this post by Kian Mohageri-2
Kian Mohageri wrote:
> I'm not very familiar with wireless either, but the 80211TXPOWER is
> the 100dBm shown, while the 20dB is the RSSI (I thought).  Could
> anybody explain to me why it is incorrect to show 20% given the output
> Lars has provided?

http://marc.info/?l=openbsd-misc&m=117390393004111&w=2

The 20dB *is* 100%.

---
Lars Hansson

Reply | Threaded
Open this post in threaded view
|

Re: [update] xfce4-wavelan

Kian Mohageri-2
On 6/7/07, Lars Hansson <[hidden email]> wrote:
> Kian Mohageri wrote:
> > I'm not very familiar with wireless either, but the 80211TXPOWER is
> > the 100dBm shown, while the 20dB is the RSSI (I thought).  Could
> > anybody explain to me why it is incorrect to show 20% given the output
> > Lars has provided?
>
> http://marc.info/?l=openbsd-misc&m=117390393004111&w=2
>
> The 20dB *is* 100%.


Thank you both for the explanations.  I'm not really sure what I was
thinking but I've learned something anyway :)

Time to read more about wireless...

Kian