Profile cover photo
Profile photo
K. Spoon (spoonix)
cannot be trusted with power tools
cannot be trusted with power tools


FCC granted my application to change my callsign.  The operator formerly known as KF5PMF is now... 

K5PUN     ;)

Post has attachment

Post has attachment
Thinking about the portable radio stuff:  The original plan had been the old "get a deep cycle lead-acid battery and some fuses, maybe try to rig up a way to charge it off solar panels".  That option sucks due to the fact that lead-acid batteries tend to be really fscking heavy, get kind of snippy when they're recharged, don't like being neglected, and the radio is pretty sensitive to voltage differences.

AspectSolar has a LiPo battery+inverter that weighs in around 8lbs and enough features built onto it to make using it pretty simple.  It's a more expensive option than hacking together a DIY power solution, but it seems like the way to go if you want dead-simple, no-muss-no-fuss power.  Add benefit is it's already got the ports to power up other devices, too.

LiPo's downside is it's also snippy about power and likes to get all fiery when it's unhappy... but that's what the inverter and other doo-dads on the box are there for.

They're also selling solar panels, but I haven't done enough research yet to find out if there's a better deal on them elsewhere.  Think this will fit nicely into a Pelican case with the radio, antennae, solar panels, and maybe some extra goodies.

Post has attachment
So, part of the TODO list for 2015 has been to get back into camping, and I'm planning on finally doing stuff with the Yaesu 897.  Currently doing homework on portable and QRP options for power and HF antennae (as well as thinking of maybe getting something lighter).

Also been looking for info for radio networks at/near various state parks, but it sounds like is about it.

It really bothers me when I see someone talk about how they did something "with a single line of code" when it's really just an entry point to some seriously complex, worringly busy, and frighteningly optimistic 20,000 line uber-class.

In hacker culture, we value elegance because not only does it create powerful code, but also reduces the opportunities for bugs to creep in.

Post has attachment

Post has attachment
I find it frustrating that this has to be rediscovered and readdressed every couple of years, the results get ignored, and we start the cycle over again when someone asks "What's taking so long?!?"

This is why as a developer I'm loathe to look for work at large organizations and prefer lurking around in the startup fringes.  It's not the free soda and lax dress code; it's the lack of red tape and TPS reports.

Post has attachment
So we're about halfway through the Docker Challenge (sponsored by the good folks over at Digital Ocean).  About 40% of the people who've taken it have managed to get at least 1 point, but we've only got a single 100% score so far.

I'm really looking forward to posting the solution for the challenge and hoping that it can help give folks a better understanding of how to use (and setup) docker and dockerfiles.  Until then, though... the best way to learn is to on-the-fly, and under pressure....   ;)

Post has attachment
Now this is interesting.  About a decade ago, someone (I think it was Richard Yoo) told me about a rootkit kernel module that hid processes and files by altering the Linux kernel's syscall table.  I spent a few restless nights worrying about how you could possible detect it before I came up with the idea for the "alamo" kernel module and hacked together a quick proof of concept.

Basically, alamo just re-compromised the syscall table except instead of inserting janky code to hide processes I inserted copies of the original functions.  I ultimately did it to prove that it could be done and never had any intentions of maintaining it because I'm not a kernel hacker.

It's been surprising to find that there hasn't been any other solution to the problem presented that's more reliable and that this is such a tough challenge to solve that the guys who specialize in safe'n'secure out of the box are just punting on the concept of LKMs altogether.

Seriously, though.  I'm kind of over this thing with systemd.

Can't we just get back to picking on emacs users and BSD nutcases?
Wait while more posts are being loaded