Different task. We were talking about making a firearm not making ammunition
:-) Ammo is easier to procure then firearms 0 of one can make it oneself.
I will take your word for that - I am surprised that ammo is easier to
get, given a gun without ammo is only a crude club, but a cartridge
without a gun is much easier to make an improvised firing mechanism for
(we have all seen at least web pictures of concealed weapons in
cellphones and so forth, yes?)
I could probably build some sort of oneshot gun for a commercial
cartridge without too much effort - particularly a shotgun cartridge,
where accuracy isn't as important as making sure the bloody stuff
doesn't come backwards at you ;)
.... form a closed (at one end) tube,
with or without some sort of breach,
Better to drill out a solid block or rod. This is best done on a lathe but
can be done by hand.
Indeed so. its still a tube closed at one end though, however you make it :)
then make a projectile, and arrange
some sort of ignition system for the propulsive chemical.
Only if you must. But (subject to the usual health warnings) Weedkiller
(NAClO3) and icing sugar in the proportions by weight that the composition
of the molecules requires (or determination by trial and error which mixture
burns fastest and most completely if you don't know the chemistry. Barrel is
bored to 0.25 in plus a couple of thou. Projectile is a steel
ballbearing/lead ball/ metal or stone rubbish between two wads to stop the
projectile falling out of the barrel and to tamp the charge. Most basic
ignition is on the matchlock principle, using a small priming charge of your
propellant and a slow match made by soaking a length of string in a very
strong solution of weed killer and then hanging it up to dry.
Interesting. I wouldn't know this myself, but I do know a few people who
I could ask easily enough for this sort of info.
You get the point I think. There's nothing here that a very average person
cannot do after a little show 'n tell instruction. You could do it now, I
with a little trial and error, yes.
still harder work for me than writing a few lines of code though :)
- Before firing it whilst holding it, fire remotely (twice) with double
your standard load of propellant. If that doesn't destroy it you should be
(H&S disclaimer) OK thereafter if you stick to standard loads of one half of
your proof load.
or hand it to an even sillier 15 year old to fire and then run away? :)
None of this is insurmountable, but would take a bit of practice at
metalwork and a fair amount of research in various parts of a decent
How long did it take you the read this?
How long would it have taken me to find this information (and more
probably, more accurate information on the formation of the propellent)
in a normal library?
Even most Internet texts are designed more to look scary and (by the
looks of things) remove would-be terrorists from the gene pool than to
produce genuine, usable weaponry.
If you want to make a good copy of
something more difficult, such as a Colt .45 semi-auto pistol or an AK-47
rifle complete with magazines, I can take you to villages in Pakistan and
elsewhere where an illiterate 14 year old will give you an excellent six
month show 'n tell course. Another six months of working under supervision
and you'll be more competent than most of the armourers in Western armies
Quite possibly. still wouldn't like to make cartridges from scratch
though (and yes, I know that a lot of professional shooters load their
own ammo, but even if they mold their own bullets, their propellent, and
their cases - with ignition caps - tend to be factory-standard).
Its not about book learning. It's about using your hands, patience,
attention to detail etc..
There is a fair amount of book learning there too. give some guy a fully
equipped tool shed (a la the a-team) and a chemistry lab, and a short
hollywood film with gunfights in it, then tell him to go make something
that can act like one of the guns he just saw....
A crypto library on the other hand requires only a considered-good algo
(such as 3des or arcfour) and a modest amount of programming ability; I
implemented both DES and RSA successfully on a ZX Spectrum in m/c at
around the same age you were making firearms - which didn't work first
time I admit, but worked once I got the bugs out.
Cheat! :-) You didn't write the crypto algorithm and evaluate its strength.
Indeed not. I am simply not stupid enough to do this (I am fully aware
that PRZ decided to roll his own crypto, and got it horribly wrong in
the first version of pgp. I am not going to repeat his mistake :)
I have no problem starting with the product of other people's skill in
mathematics - and even if I were to invent my own crypto scheme, it
would have to be based on math I had learnt, as I am certainly not going
to reinvent math from first principles just so I get *all* the glory ;)
I thought their implementation was fine, but key reuse happened?
There is also the issue of bias in the RNG which isn't a failure in OTP
itself, but in its key material
The (principle of OTP is sound and very simple). Protecting information with
a security equal to the strength of that simple algorithm is hard to do.
Yup. often the crypto is the strongest link anyhow, even if it is
something as underpowered as DES.
This is of course true - but to be honest I still think you could get
better results giving a random programmer the algo to arcfour and saying
"impliment this" than giving a random metalworker a diagram of a gun and
say "build one of these"
Do you still feel the same having read this?
Yup. unless you plan to just have it fire cartridges, you would need
a fairly detailed description (such as you have supplied, but obviously
giving a more exact mix for the chemicals and possibly specifying brand
names) of the propellent, the barrel, the projectile and assembly as a
whole. this is more complex (and requires more skill) from the average
metalworker than implimenting arcfour would for the average programmer.
I would also hesitate to put that rube goldburg gun up against a
machine-shop rifle, but that programmer's implimentation of crypto could
easily be as strong or stronger than many commercial products.