Un-mothballing attempt

[Charles Yarnold]

In an attempt to un-moth ball this project I grabbed out my kit and put it together! I started to look at code and realised we haven't yet solidified the more basic concepts about how we go about making this thing, so thought it would be useful to talk some of them out with you lot, the main questions I think we need to decide on are:

1. Are we thinking people build their own gear to bring to play, or just buy kits?

I would say both, and with that in mind:

2. Do we trust what others gear is doing?

I would say no, but as Tom says "If they do that I will punch them in the cock" but to do that we need to know what they are doing and record it for easy automatic verification at the end of a game. This could take the form of "oi, we said your BFG could only do up to 100 damage and shoot only once per second"

What I propose is that the gear accepts what its getting from other players during the game, but at the end when data is dumped to a main repository we have automatic checks of the data to know if anyone has been taking the piss. This makes it easy on the gear, and lets you know if someone is being a dick at the end of you game (with telling them to bugger off for the next one)

3. What stats do we need for bullets?

Off the top of my head: (example)

• Damage done - (25 points)
• Type of damage - (Ballistic, explosion, fire, radiation etc) (This could be useful for giving the game items like "hazmat" suits that would apply, say, a 90% reduction of damage dome from radiation type "bullets") 
• Armor piercing - (i.e. to negate shields)

anything else?

[Nigel Worsley]

> 1. Are we thinking people build their own gear to bring to play, or just
> buy kits?

I think you have to assume both. Unless you took their gear apart you wouldn't
know if it was a kit or not anyway.

> 2. Do we trust what others gear is doing?

Unless there is evidence to the contrary them yes. Which is why evidence gathering
capability needs to be part of the system!

> What I propose is that the gear accepts what its getting from other players
> during the game, but at the end when data is dumped to a main repository we
> have automatic checks of the data to know if anyone has been taking the
> piss. This makes it easy on the gear, and lets you know if someone is being
> a dick at the end of you game (with telling them to bugger off for the next
> one)

Sounds good, and most of that data will be wanted anyway for stats.

> 3. What stats do we need for bullets?

> <snip existing stuff>
> anything else?

Range? Useful for grenade style weapons, but would only really be useful if there
was some sort of microlocation system in operation. Decawave have some cool
chips for this, claimed resolution of 10cm, but they are currently only available
in tiny BGA packages to customers buying them by the truckload :(

I had some thoughts about the whole IR bullet approach a while ago, and have an
alternative idea to suggest:

The vests have lots of IR LEDs that constantly output the player ID, the guns are
actually receivers not emitters. When the gun receives this information it sends a
'you have been hit' packet back via a short range RF link which can include the
bullet type info, this is then processed locally to decide the damage etc.

A big advantage of this is that the vest can be covered in a large number of relatively
closely spaced emitters instead of a much smaller number of expensive receivers. It
may also eliminate the need for the vest and the gun to be connected, which seems
to me to be a bit of a weak point durability wise.

The gun to gun transmissions could also be picked up by the main computer,
enabling real time diisplay of statistics. A remote commander could also use this to
send tactical commands back to the troops, which would be rather cool.

This may make it easier to spot cheating too, a player that switches off most of
his IR receivers would be difficult to catch, but an IR camera would quickly reveal
disabling of the emitters.

There are probably lots of disadvantages to this that haven't occurred to me yet though!

Nigle

[Charles Yarnold]

On 9 May 2012 22:35, Nigel Worsley <nig...@googlemail.com> wrote:

Range? Useful for grenade style weapons, but would only really be useful if there
was some sort of microlocation system in operation. Decawave have some cool
chips for this, claimed resolution of 10cm, but they are currently only available
in tiny BGA packages to customers buying them by the truckload  :(

Range would be an interesting addition, I suppose the easiest way to do range is the ir power you put behind your weapon, if your gun is too weak your not going to hit someone across the other side of a field, but if you put a few high brightness leds and put a good lens on it you can have your self a sniper rifle!

I had some thoughts about the whole IR bullet approach a while ago, and have an
alternative idea to suggest:

The vests have lots of IR LEDs that constantly output the player ID, the guns are
actually receivers not emitters. When the gun receives this information it sends a 'you have been hit' packet back via a short range RF link which can include the
bullet type info, this is then processed locally to decide the damage etc.

A big advantage of this is that the vest can be covered in a large number of relatively
closely spaced emitters instead of a much smaller number of expensive receivers. It
may also eliminate the need for the vest and the gun to be connected, which seems
to me to be a bit of a weak point durability wise.

The gun to gun transmissions could also be picked up by the main computer,
enabling real time diisplay of statistics. A remote commander could also use this to
send tactical commands back to the troops, which would be rather cool.

This may make it easier to spot cheating too, a player that switches off most of
his IR receivers would be difficult to catch, but an IR camera would quickly reveal
disabling of the emitters.

There are probably lots of disadvantages to this that haven't occurred to me yet though!

Some interesting "turning it on its head" ideas!

It may be worth sticking to the gun emitting model for the moment as we have the dev gear made for that mode of detection.

That said I love the gun to gun to computer transmissions bit, this could be done with either the phones connected over bluetooth to the gun, or sticking one of the low power and very cheap nordic rf modules on the gun (about 20m range).

[Nigel Worsley]

> That said I love the gun to gun to computer transmissions bit, this could be done with either the phones
> connected over bluetooth to the gun, or sticking one of the low power and very cheap nordic rf modules
> on the gun (about 20m range).

I had the Nordic modules in mind, they have the advantage over most other low power RF devices of having
built in ack/retry mechanisms to deal with message collisions - which would be rare due to the very short
packet durations.

There are also higher power versions available that include transmit and receive amplifiers, we could
probably get away with only using these for the control computer, the guns are likely to be closer
together.

The features and protocol for this don't really depend on the choice of RF link, or even if the gun is
an IR sender or receiver, so the spec can be worked on without committing to any of these.

Nigle

[tom wyatt]

well the range thing can become a physics thing, for example we say "all guns mustnt have over 5 whatever-wattage IR LEDS in them".

If you want a short range shotgun then you arrange the LEDS appropriately, if you want a sniper then you bung some focussing optics in

--
Johnnysausage.com : people hurting themselves! woop!

[Charles Yarnold]

Yarg, would be easy to make a testing unit that records the power and spread of a shot to calibrate the guns before a game.