Discussion in 'Armory - Q&A' started by rudd, Jul 28, 2015.
Encryption key changed every newly established link?
>Proprietary encryption algorithms are not to be trusted as they typically rely on ‘security through obscurity’ and not sound mathematics. These algorithms should be avoided if possible.
This is a fairly important thing in foil in a way that it is not in epee, that has come up in-club when testing the LP system.
Firstly, the critical difference between foil and epee is that if a wire breaks in an epee it will register no light, rather than a white light as a foil does. So in foil, if you have something loose or dodgy in the socket or the tip, and someone beats your blade, you'll get a white light. We've had moments when simply beating the blade, has registered a coloured light.
In epee, a failure on the weapon results as a disadvantage to the weapons owner, so they are incentivised to have working weapons. By failing in a way that can cause coloured lights, there is some incentive to have barely working equipment.
To be honest, it's a really great product, and if the price point that I've hear rumoured is correct, I would likely buy one, but I still am not comfortable fencing on the wireless system for bouts that I think will be close, even in-club. If both fencers are sweaty, the wireless system is still noticeably different from even an older wired system.
However, the benefit of being able to keep a wireless system in your fencing bag is pretty hard to beat.
What you're describing is a problem with foil itself, not the scoring machines. Nothing LP can do about white lights from bad wires and connectors.
However I don't see how a wire break due to a beat could cause a colored light, with either wired or wireless systems. Unless of course the tip of the weapon was on the opponent's lame at that time. But again, this could only be solved by redesigning the way the weapons themselves work. And nobody seems interested in that.
I am at one of the clubs that is testing the LP wireless box - it *did* cause a coloured light that was replicable. The fencer jiggled their body cord, and where you would normally get a white light due to a loose connection, they got a coloured light without being anywhere near the opponent.
Hi bonehead can you please drop me a line on alex at domainnameofourbusiness.com i have heard of this problem with sockets once before but have never managed to repeat it. If you have kit at your club that when you waggle a faulty connector can generate coloured rather than white lights we can probably filter it out. Please e mail me.
It's probably an artifact of the portion of the design where the systems aren't actually connecting wires but detecting other situations that come along with a touch. Can't comment on exactly what may be causing the problem because Leon Paul isn't releasing the theory behind the system, it's probably due to the nature of having to circumvent the way the system was intended.
It's actually a pretty neat example of hacking in its rudimentary form. We have a system intended to operate in a certain way, and to introduce an unintended effect they're searching for an utilizing unexpected or ignored behavior.
Not hacking- that's a fundamental part of any decent testing regime that will be a part of any decent development process; don't just check that it behaves correctly when you follow the intended use cases perfectly, but develop a battery of negative test cases to ensure that it doesn't misbehave when used incorrectly, try to anticipate what unintended conditions might occur to include those in the test plan, and once you've got it to a largely functional state get it out into the hands of real world users with the partial intent of turning up the unintended conditions that you did not anticipate.
Sorry I think I didn't get what I was saying across clearly! I meant the idea of creating a wireless fencing system, utilizing a wired system as the base. It's a kind of hacking in a way that we're taking an existing system (ie the wired fencing scoring system) and trying to implement a different function (ie wireless) through unintended or ignored effects (ie capacitance and whatever Leon Paul is using). I'm familiar with testing a product in development.
Many of the difficulties could be solved with changes to fencers personal kit. Our vision for this is mass market rather than just the few fencers at the top so we decided that we had to make it work with standard fencing kit.
I would love to know how close we are to being able to purchase this product. That would help me to determine if I will teach my class at the school itself, or if we will have to go downtown to fence at the club.
And once you achieve market dominance, you can take a page from Apple's book and change the required bodycord every couple of generations .
My advice, don't hold your breath.
Just curious, what would you propose? As far as I can tell, the fundamental difficulty is with having to detect electrical conduction between two sensors which are basically isolated. And you'd have to make some very drastic modifications to the equipment to eliminate that requirement, especially in sabre.
Maybe we can finally go with kinetic solutions instead of solely electrical ones in the future. Or lasers. Lasers are cool
I can't imagine a "purely kinetic" system capable of distinguishing targets.
"Lasers" are actually an idea worth looking at, imo. My ideal system would be to make the tips of the weapons color-detecting, though I think it would be pretty challenging to run a fiber optic line up a blade instead of a copper wire. And I don't know if that could be applied to sabre since it needs to sense contact along its length.
As long as the wire doesn't break, it should last the lifetime of the blade.
It would bring new meaning to the term "dirty sensor" though.
While I would agree that it is more likely for a "hacker" to attempt a spoof on a wireless system because it's what they do, I don't think that wireless is more easily spoofed than a wired system. Clearly you do not have to actually touch the system to spoof it with wireless, but how hard is it at say an NAC, to add device or switch on the current wired system. Even a small electronic device that would "break" the valid wire in foil or complete a circuit in epee inserted into a connecting cord plug or reel would be easier and require much less technical savvy and be more difficult to detect.
Picture someone (looking like an armorer) going to a strip, wearing jeans with a tool-belt and vest and hat and replacing a connecting cord with a "spoofed" cable. Evil coach/fencer later simply presses a button that breaks the valid circuit in foil or completes the valid circuit in epee at the right moment. Bwahahaha!
Until we are talking about FIE approved, I think that for club or domestic use, it easy to get overzealous with the attempt to make the system "foolproof" from the hacking perspective. I would be most happy with a system that just functions as it should with a relatively minimal encryption protocol(s).
As far as the fencer end of the system is concerned, I think that it makes sense for every fencer to have to buy their own potted device. This has the added benefit for the manufacturer of selling a vast number (maybe even improving the price point due to volume) of units and of course eliminating the "walking off" scenario so many have expressed concern about. I would like to see all the fencer's personal info (including photo) incorporated into the device (added by the manufacturer in the ordering process) for various purposes including display on the scoring apparatus. They are unlikely to easily break, and considering I break several FIE blades every year would likely be a minimal cost in comparison. Unlike blades and replacing various items of clothing, it would likely be a one-time purchase. The fencers personal information stored on the unit would also make the item less useful to others and make recovering a lost unit more likely.
Fiber optic cable is quite a different animal from metal wire though. But if you used plastic optical fiber and hermetically sealed both ends really well, it might last a long time. I've always wanted to try developing such a scoring system, but I only have expertise in electronics, not machining or metalworking.
All the wireless and sensing protocols are going to be proprietary, that means that even if another company develops a similar wireless system, there will be no cross compatibility between them. So you either 1) Grant a guaranteed monopoly on scoring equipment to one manufacturer, or 2) Require all fencers to own sets from all manufacturers. Obviously both options are terrible, so unless someone thinks of another way, I think we can write off personal scoring equipment.
Separate names with a comma.