Wednesday, February 22, 2017

Looking at the message our card sends

In the software world there is a world, a myriad of ways for one device to message another. The Fintech folks likely have a solid bit of work done, in use or ready to use.  But, skipping the formal part, what are the likely fields we will see in a message from a smart card to a pit?

And, as I say before, we are going to get a triplet <token,coin,bot>

Bot is an ID, or method, or port, but it activates an actual snippet of possibly long code, actual software on the server box that runs the trading pit.  Bots sell (as in purchased) , traders love them, and traders have a personal relationship  to the various bots they like around the pits, they name them. Some quants are famous and many more rich because traders pay lots of dollars for great trading bots, and pay for their upkeep.  It is a lucrative business, and at Redneck Systems you are all welcome aboard.  Your bots can be as big as Watson, or simple be the null bot, holding your money at neutral.

Let me repeat quant folks, big big money here in writing these bot snippetss.

Coin? Right now I have that field identifying a value and a set of opcodes. I can think of some ideas for those opcodes, but they are all special instructions to the bot.

Token.  This is the validater, if the triplet has a token that passes the central token tester service, then the endpoint is hardware guaranteed to obey honest accounting.  The token may also give a bunch of other parameters of which I know little.

And, I guess we assume each of these fields may be a set of null to more than one. 

The token field should be a valid address for a ledger service, b ut the ledger service may already have some part of your token registered at the central service.  The coin opcode may be sophisticated to bet without the bot.  If no ledger service is not called out, then after the finite number of bot actions, the coins are placed at neutral.  The central service will be notified of ignored coins after a timeout.

All this kind of stuff, most of it over lapping and redundant, and all of it needing experimentation and thought.  But the experiments are profitable, they gains large enough that the market opens up connectivity requirements, for a fee.  But the ad hocs get integrated into a clean, efficient multi-currency protocol real quick. Been there, done that.

No comments: