|S.NSA incoming and outgoing|
|Intangibles and goodwill||7.98496216||0.33331337||8.31827553|
S.NSA has a total of 4`737`075 authorised shares outstanding. The total assets per share implied value is thus 0.00009980 BTC. The cash+tangible assets per share implied value is thus 0.00009804 BTC.
S.NSA realised no operating revenue this period. The P/E implied value per share is so far 0 BTC.
S.NSA has Special Stock Warrants outstanding issued as per the IPO agreement, as follows :
Excerpts from a letter :
1) Routed mainboard. Introduced a small detail: transistor to let CPU cut power to micro-sd card during 'red' (crypto) operations - to frustrate a hypothetical Satanic sd card (which could, in principle, carry out differential power analysis.)i Board gerbers will be sent off to golden turd when item #3 is complete, see below...
2) Eliminated /all/ dependencyii on our microcontroller manufacturer's (Microchip Inc.) build chain and libraries: by writing all of the support routines normally supplied by the latter from scratch, in asm. Based only on the supplied chip docs! We now use exclusively traditional gcc/gas/ld (MIPS little-endian ver.) This not only abolishes the closed-source turd from the build chain, but enables a quick and painless future port to any MIPS based core (there are quite a few of them, e.g. Loongson, or even my own MIPS-compatible Verilog core for use in FPGA or ASIC design.)  Presently the firmware is such that a completed board, when it appears, can be meaningfully tested. ('Pre-alpha' ver., if you like.)
3) Why didn't order boards fabricate yet? Answer: these particular boards (mainboards) will be a substantial expenditure. Can't afford the inevitable duds  if assembling by hand (tweezers OK for passive 0805 components, not so much for TQFP-64 flat-pack ICs.) Paste application by hand for a large SMT component is also out of the question. But the expense of professional assembly run for a miniscule qty. of boards is ruinous. So I finally bit the bullet, and came up with a sharp, if arduous, cut for the Gordian knot:
Assembling a semi-automatic  pick-n-place and paste layer apparatus in jungle conditions!iii
Loony? nope. See figure 1. That's most of it... All that's really needed: a gadget for moving a small object (paste extruder or vacuum nozzle) in Cartesian coordinate grid (xyz) in a repeatable way. No artificial intelligence is necessary for the prototype assembly - what I want is essentially a 'waldo' micro-manipulator. Fortunate find a few weeks ago: a carcass from laboratory robot, with three functioning lead screw rails and motors. Z axis is not shown in the photo, I have removed it for rework 
Presently building simple driver apparatus (pc interface) for the motors. For the actual targeting of z-plunge, I obtained: a small crosshair laser. (cheap and angry, works.)
When #3 is complete (see note 4) it is to be tested by way of assembling a number of RNG boards, for which the PCBs, complete component sets, and test rig are already at hand.
 this was a first-class bitch - essentially wrote an elementary OS (init, interrupt handlers, drivers for peripherals: i2c and spi eeproms, sd interface, the whole shebang.) Not because of MIPS itself - it's perhaps the world's simplest cpu arch, instruction set fits on an index card. Mainly on account of the gnarliness of the peripherals. But now I understand this chip very, very well...
 it is possible to assemble virtually anything by hand. but to do so accurately, and never have to wonder 'doesn't work; is it a defective pcb? firmware bug? or 0.5mm^2 of splashed solder paste?' is another matter.
 it only needs to obey carriage motion and vacuum release commands from pc keyboard, and return to particular, saved locations at a later time when so prompted. if I had the necessary cnc tooling, I could fabricate feeders for the SMT component tapes (if you've never seen these tapes, they look like movie reel. most parts ship lodged in such tapes.) but I don't presently have the tooling. Instead, the tapes will be peeled and clamped to the base of the gantry - just as they are clamped to my desk during a tweezer assembly. Dirty.
 removed boom which holds the z-axis rail and its motor. (the original aluminum boom that came with the carcass) - need a rather different shape here. was originally planning to machine the replacement from Al, but presently lacking the tooling. Instead: laser-cutting the part from acrylic. This is arguably a kludge, but will suffice for the actual short-term purpose of this instrument (rapid prototype assembly.)
P.S. The components of the instrument described in Item #3, just like the rest of my workshop tooling, are purchased by myself in my capacity as a generic human and will not be charged to the S.NSA piggy. Total cost of it so far is ~200 usd. Won't starve on account of this.
As you can see, there is a charge on the S.NSA account, so he's not getting away with that PS.———
- This is not quite as theoretical an avenue of attack as it may seem. Suppose one's opsec is casual enough to allow control of inbound items (say someone gives out free SD cards in his vicinty, which he then pops into the Cardano) and also not bother to properly dispose of sensitive material (SD and flash cards are supposed to be hammered into a fine powder). Building a magic card that does some magic and saves the results isn't really all that hard. [↩]
- Holy shit. So how much is this corp worth, a company which owns a) a complete reimplementation of a generic micro and b) the skunk worker that did this ? [↩]
- Holy... hell ?
Seriously, this guy is sitting over there building a handcranked factory. How much is a functioning handcranked factory - of anything - worth iyo ? S.NSA is going to end up with its own museum, I tell you. [↩]