Looking at trader bots

Now writing any nest block code, sketching bots, mostly exploring how a typical algorithm would do things I call normal statistics.


for bid in the_Pit(by_price_range,using_dict) :
 mylist.append(bid)

The bot has a collecion of bids, go through the list, prepare a bid and:

for bids in the_pit(good_to_trade,using_dict) :
 if trader(bid,self) : break

Selector functions, by_price_range. You wrote it in standard python, looks at a number, selects it based on some bounding function. Good_to_trade? What's that? Dunno, you wrote it send me a copy. I presume it looks at each of the open bids, compares them to some statistical threshold you computed, then attempts the trade.

The only thing new? Pay attention to your descents (for loops and while loops), they cost pennies.  And remember, the bids are not static, between, and during for loops, the tree changes, other bots run the graph.  You do not get 'lock', unless its in the contract, and you gonna pay big for it..

On and off pricing

On the graph, you pay contract price, off the graph, free. If you are a tegular tradfer, the site owner is giving you compute power one fiber hop wsy from yhe pit.  So, run a large code, in your own protected space, all day long, and whenever you need some input, run he graph with for loops and bid selectors, and collect sub graphs,in their entirety, for the cycle pridce.  The pit boss can only charge for cycles on the tree..  If we think about we can see why. Any pricing outside the pits is not well rounded, red/greer won't work,a lot of the timer. We get entropy gaps and queue cycles..

Notice the terminology, almost always, a lot, tiny.  The contract terms should be in base quants, dollars is a dynamic value, pricing on the net is in quanta relative to the nature of the pit boss.  If it just promises the disordered bid list, then it will charge by the copy.  Depth, it is a generic quant and any bot can take depth and get an approximate ratio of pit cycles available.  The boss should keep some simple running total, like bots on the graph, current depth, typical  bin size, in unts of singletons.  There is no master knowledge, python does not count, and if they count, they break the contract.  The pit boss has no more priority than the trader bots.  So, all bot have he same precision on current pity variables, and are free to make counts of things in their for loops.

No calls back to the interpreter, really

If the bots ate going back to the interpreter, then we put an undue burden on the interpreter code, it has to meet well rounded condition.  If the interpreter gives great info with a short call on some things, believe me half the bots will be here half the time.