I wrote two emails today, one dealing with the subject of the use of
ARM platforms for app_rpt/ chan_voter software, and one dealing with
the audio capabilities of modern laptops, particularly Macbook Airs.
The intent follows:
--------------------------------------------
So after alot of wrangling, I've found out something:
DAHDI and Zaptel support on ARM Linux kernels version 3.0 and above
just doesn't happen. I personally believe this is because most ARM
systems lack a PCI slot of any description, and pretty much will
forever lack PCI slots (the last ARM I heard of that had PCI was the
Ionynx PC from the UK, running RiSC OS - oops?). Now this doesn't give
Digium an incentive to code up some support for non-existant hardware,
under the pretense that app_rpt and chan_voter needs tonezone and a
handful of other elements from Zaptel (or DAHDI), does it? Which means
you effectively cannot, for the foreseeable future, use an ARM
platform as the controller for a PMR network.
That means that, for the first time in five years, I'm considering
buying an x86 platform. Huh, you say, this is Shane we're talking
about here, he don't do x86! Unless a certain set of conditions are
met, no, I don't...
Meet the Fit PC-3. You might know it, or the Fit PC-2. I admit to
knowing the Fit PC-2 quite well. A nice piece of gear. What I didn't
know was they updated to an AMD APU based system, and qualified the
PC-3 for use with Linux Mint. Uh huh, headless Ubuntu, here I come!
And this little, fanless, 17Watt total draw bad boy will run all of
the usual suspects in a PMR network that you may so desire...
Which is what I desire. ARM can be such a pain in the ass sometimes...
This system, with 4GB of RAM, no HDD, and postage from Israel (I don't
think it'll get here before Christmas, but thanks for asking), is the
princely sum of $485, max, including Paypal "tax." We pay the nobles a
damn stiff fee for civilisation, don't we? I have a 40GB SSD here,
just about ready to go. I also have some 12VDC gear, ready for this
little thing to strut its stuff.
As I want it to run headless, I'll probably get it to talk SSH
("Shhhhhh!") to the outside world, but I do have a 12VDC compatible 15
inch DVI monitor, with HDMi to DVI cable. Yes, yes, it uses cold
cathodes... gah, its so hard to get good help these days. I don't
actually have to have the monitor on for an entire weekend, only when
things hit fans in interesting, intersecting ways they're not supposed
to. If that happens, and I can't log into my PMR controller, I think
the network is pretty much toast by that point.
I don't believe I'd need a failover, I remember reading something
along the lines of underclocking will occur until such time as a heat
issue can resolve itself, the thing, with SSD, has no moving parts (no
pesky fans to fail!) and as long as you don't hook the power up
backwards, this thing is pretty much unkillable. Dual core APU at
1.0GHz, as mentioned, 4GB RAM, 40GB SSD. Running a nice, recent
revision of something that can operate in console/ SSH mode. I don't
want an X server to slow me down. This thing is a PMR controller, not
a Facebookin' machine.
So, I've solved all my problems with ARM platforms in one fell swoop.
This is good you say, I'm finally making progress. Lucky I found out
with the RPi about the DAHDI/ Zaptel driver situation, and of course,
the existence of the Radio Thin Client Module as well (thanks Jim! I
owe you a tall, cold beer one day!). The RTCM is the basis for
VoterCard, the PMR full transmit simulcast controller for Simoco
radios.
--------------------------------------------
For use of a laptop system with IAXRpt, one finds that most laptops
audio systems are inadequate, especially in the high noise, and high
tension area of a command centre. The Macbook Air is further
disadvantaged by the fact it has no microphone port, simply an inbuilt
microphone which is prone to picking up alot of noise, and in a
command situation, broadcasting four operators voices over one channel
is a recipe for failure, and yes, even danger...
However after all the work I had invested into getting IAXRpt running
on my MBA, I felt it appropriate to address what I see as Apples total
shortcomings in the audio department. Before I go on to explain how I
addressed these issues, I will elaborate - the running of the Windows
application IAXRpt under Mac OS X is simple, using a version of WiNE
for Mac OS X. The particular wrapper software I use is called
WineBottler. However, I could not use the stock "for OS X 10.6"
revision software, as it did not support audio in any way, shape or
form. I actually had to download WineBottler, with 10.6 core, and
replace the core with 10.7 audio enabled core. As far as I know, the
10.7 core WILL NOT run on anything less than 10.7. Of course, I don't
have anything to try now, do I?
So, a quick look on eBay located a USB sound card, with advertised Mac
OS X capabilities - have you ever tried putting a Winmodem into a Mac
G4? =P This was mine for under the $5 mark. Similarly, another quick
search turned up a headset appropriate for the USB sound card with
singular headphone and boom mic, as well as separate plugs for the
audio paths. Compatibility was assured - but I just had to TEST it!
Today I got the sound card after about a month of waiting. After
inserting the USB card, I looked through the sound settings - "Generic
USB Sound Device" came up, and the connection type was "USB" (versus
"Built In"). After setting the appropriate settings, I took a deep
breath, and started "Above the Storm" by Alpine Fault in iTunes...
Success!
I was listening to those strums, explained to me as being similar to
those of Nightwish, through my headphone! Noise and auditory
interference would now be minimised. But what about sending audio?
Would the mic work too?
Opening GarageBand, I quickly selected "Voice" and made a recording.
My nasally voice popped out of the headphone. I was satisfied that the
Mac would work. Now the real test: would WiNE (and thus, IAXRpt) work?
I am unable to test sending audio, although the error tone came out
loud and clear from IAXRpt. It appeared I wouldn't have to change any
settings on the IAXRpt (Windows) side like I was expecting - it must
be transparent to the Windows software where the audio is actually
COMING from or GOING to... gotta love that old Mac reliability to just
bloody well work...!
--------------------------------------------
Hopefully this helps anyone in my position, and updates you all on the
progress of the RPi project. It is a little unfortunate, but I have
found out alot along the way - Jim, I do owe you that beer for
pointing out the RTCM to me. And if you have any news concerning the
status of the ARM drivers, I'm all ears...
