[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

modified wmwave for Aironet driver



When I used to have a laptop with an Orinoco card, one of my most treasured
dockapps was wmwave.  Since moving on to a Cisco 350 MPI card, I have hunted
high and low for something similar to no avail.  I tried wmWiFi but all it
did was spit out ioctl() failures and I really did prefer the look of
wmwave.  So after looking at the differences between the driver header
files, I decided last night to just hack wmwave.

I must admit that I was lazy and didn't bother to formally convert this
port.  I just simply hacked the main file.  Perhaps when I have more time
I'll clean things up and make it nicer.  Of course if someone wants to do
that for me, I'm not going to stop them. |8^)  The file can be found at:

                http://www.neebu.net/~khuon/codebits/wmwave.c


Here's a summary of the list of changes I made:

[0] I grabbed a copy of the net/wmwave ports and extracted it into /tmp to
    do the work while not contaminating my own ports tree.

	[a] I just performed a make (not make install) to get the files into
	    a state for hacking.

	[b] After I finished, I renamed wmwave to wmairo and copied it
	    manually to /usr/local/bin instead of doing a make install.  As
	    I said above, if I happen to find some copious free time, I'll
	    make this a nice and clean port.


[1] Most of the definitions were similar so I started out doing a global
    replace on a lot of variable and structure names.


[2] For some reason the signal cache (ANCACHE) for the aironet drivers never
    get properly populated.  Either that or I'm interrogating the structure
    wrong.  However "ancontrol -i an0 -Q" always returns a blank for me. 
    After playing around with it for a while, I ditched the idea and just
    grabbed the information out of the an_ltv_status structure instead.

	[a] The signal quality seemed to be reversed.  Even ancontrol
	    reports what I think is opposite.  In other words, as I move
	    closer to the AP, I saw the value of an_cur_signal_quality going
	    down.  I decided to do a quick hack on the value to reverse it. 
	    It now _seems_ correct.

	[b] The scaling seemed wrong with the original code.  Perhaps it was
	    correct given values as reported in wi_sigcache.  I removed all
	    the scaling modifiers and the values seem to line up better.


[3] I changed the name of the labels so that they make a little more sense.

	[a] The general status label used to show either "Quality" (card
	    present) or "No Card" (card not available) followed by a
	    coloured dot to indicate general signal quality (red = bad,
	    yellow = mediocre, green = good).  The actual "Link" bargraph
	    showed the quality in a more graduated manner.

	    I decided I could make better use of the label area so I added
	    some code to display the channel number (an_cur_channel)
	    instead.  It will still display "No Card" when a card can't be
	    found.  My intent was to give me a better sense of which AP I
	    was associated with.  I thought about using the value of
	    an_ap_name but my experi4ence is that it's unreliable.  Many APs
	    don't have it properly set and actually my config tools for my
	    APs don't seem to have a way to set them.  Additionally, I'm not
	    sure they're ever transmitted.  Then again I'm not an 802.11
	    guru.  I also thought about using the MAC address off the AP but
	    I didn't think it would fit nicely.

	[b] I renamed "Level" to say "Strength" since I'm grabbing the value
	    of an_normalized_strength and thus seemed to make more sense.  I
	    also fixed scaling.

	[c] I left the "Noise" label alone but I'm not sure I'm retrieving
	    the correct element.  I'm looking at an_noise_prev_sec_pc which
	    should be giving me the noise in percentage but I've never seen
	    it anything but 1% although every once in a while I will see a
	    slight flash to around 20% but that may just be an anomaly in
	    the polling.  Perhaps I should try firing up one of my cordless
	    phones and sticking it next to my laptop.  I also added the
	    ability to report SNR instead of noise.  Compile with -DSNR to
	    enable this.


Well, I think that's about it... I hope it's useful to someone other than me.


--
/*===================[ Jake Khuon <khuon_(_at_)_NEEBU_(_dot_)_Net> ]======================+
 | Packet Plumber, Network Engineers     /| / [~ [~ |) | | --------------- |
 | for Effective Bandwidth Utilisation  / |/  [_ [_ |) |_| N E T W O R K S |
 +=========================================================================*/