GUN Linux: On the range with TrackingPoint’s new AR-15s
arstechnica.com
arstechnica.com
(half joking on that last question)
I did skim through the previous article on the gun: http://arstechnica.com/gadgets/2013/03/bullseye-from-1000-ya...
My instinctive question was...if the software is for the scope, then why does TrackingPoint have to be involved in the manufacturing of the carbine? To a layperson, it would seem that a sophisticated scope could be recalibrated to any version of the AR-15...but in reading the 2013 article linked above, TrackingPoint feels that it must control every aspect and characteristic of the shooting experience, including the ammunition (which it sells for $5-7 a cartridge):
> The PGF is able to function at range because the system is preprogrammed with the ballistic characteristics of not just the three types of rifles, but also of the ammunition. TrackingPoint has partnered with ammunition manufacturer Barnes Bullets to turn out ammunition with very specific characteristics, effectively selling both the razor and the razor blades. Our original coverage of TrackingPoint from CES drew lots of questions about reloading or hand loading—that is, assembling one's own ammunition from casings and powder and bullets—so I brought this up with Bret Boyd while we were at the range.
"Gun people are very passionate, and if you're into reloading and hand loading you're in the one percentile." We talked between shots, wearing identical Peltor electronic ear protectors. "The problem with hand loading is that you just have inconsistent results. Some people do it very good, and some don't. What I really want to avoid is the situation where someone says, 'Look, your gun doesn't work. I'm missing, and it's your fault.' And the real issue is, well, you're missing because you screwed up the ammunition—you have it loaded too hot and it's firing too high. But you're never going to believe me!" He laughed. "I'm really not trying to make a ton of money on ammo, but I want to control the outcome and I want people to have a good experience."
To be geekish about this, it seems this gun system is more akin to the Apple ecosystem than to Linux. What are the possibilities of "hacking" this gun/scope ("hacking" in the non-black-hat sense)
Re: the ammunition, I think this approach would be interesting combined with some form of learning algorithm that could adapt to different ammunition (assuming it was different but reasonably consistent ammunition), but, if I recall correctly, this all seemed to be very much "hard coded" into the rifles software, based on the proprietary ammunition. Given the $20k price tag and the possibility of expecting 1000 yard shots, the $5 - $7 a shot seems what I'd expect (match grade .308 ammo can go for $1 - $3 anyway).
Still reading through this article; interested to see if the reliance on proprietary ammunition holds on something that's not meant to give 1000 yard shots to relative novices
Edit: ARs are used in NRA high power matches out to 600 yards.
This rifle 'system' doesn't accomplish anything that a well-trained human can't accomplish.
That's sort of the point to me. Tech always gets cheaper, people usually don't. If you're wealthy enough to buy a $20,000 rifle for fun (I'm assuming the kind of person that can buy a $20,000 rifle probably doesn't need it) then the time spent to learn to consistently hit the shots might be worth more than the money you spend on the rifle.
Other than that, "This x doesn't accomplish anything that a well-trained human can't accomplish" could have been and likely was said about almost any technology that's existed. In a similar somewhat less historically accurate vein, I'm reminded of John Henry.
Well, not in a match in which you plan to be competitive, anyway.
NRA High Power, sure. Palma, certainly (the rules require an unmodified .223 or .308 case), but you're not shooting commercial "match" loads.
I'm more interested in the application of learning algorithms and how useful they would or wouldn't be in being able to get a somewhat optimum statistical pattern out of them (optimum within the tolerances of the particular loads quality). Basically, what's the maximum accuracy you could get out of standard off-the-shelf ammo if your rifle could self-adjust, is a question I'd be interested in looking into.
If you've got any particularly relevant links on the various types of ammunition you'd use for competition, I'd be interested in reading them, though.
It's important to note that TrackingPoint isn't just the scope, it's the entire weapon system. Everything about the gun (right down to the ammunition used) needs to be 'known' to the system to make an accurate shot.
Things are pretty hackable, in a very different way from computers. It is amazing how far hobbyists can take their passion for the various technologies and physics – few of which ever have any need for electricity, and many of which evolved from the more gruesome acts in the annals of human history.
But like – you know how you can get rid of KDE and go Gnome? Similarly, you can also get a milled lower receiver for your AK-47 (exotic, as it was designed from stamped metal components), and if you do it right you can still use your favorite wooden furniture.
To extend the metaphor•, that's like being able to reuse all of your conf files and compiled binaries on a new processor architecture based on a different distro – as long as you're extra careful about `rm -rf` and you understand that mishandling powerful tools will kill you and your friends.
I don't have any interest in killing people, but I find guns fascinating in the same geeky way that compels me to like having opinions about various other dramatically complex systems whose histories are steeped in the bizarre slow-brewed lore that forms around the sort of stories geeks occasionally offhandedly tell in their few and far-between hours of not being furvorous (like e.g. when github is down).
• - quite beyond its utility, I am afraid, my apologies