Free Archaea by Dain White Page B

we're going might not be luxury and glitter, but it was going to be a lot more fun than buying air by the can and sleeping in some scum-crusted rentable in this floating pile of crap.  
    Yak asked me to hold station at the ingress escaladder leading down to ring 10. There was a little café of sorts bolted in nearby, wedged between some sort of fenced off warehouse and a glom cube farm. It was one of those little dives that in an earlier era might have been on wheels and parked in a factory lot waiting for the noon whistle to blow. I picked a good spot to watch both ingress and egress tubes, and ordered standard station-fare: noodles and sticks.
    I watched Yak head down the ring until he disappeared around the curvature, and then watched those tubes while I worked through some of the best noodles this side of Old Beijing. Strange how food tastes so good when you expect it to be bad.
     
    *****
     
    Once we docked and I powered ship systems down to nominal levels, my attention turned to stress assessments and efficiency reports. The tokamak had been lit, and while we only pushed it to about 10% of capacity – as an engineer I knew that was more than enough to identify anomalies that might extrapolate into monstrosities at higher output.
    I was especially concerned about the process controllers, and their wetnet interfaces, as those were pretty much custom built by Pauli, and we didn't have any dox on them. Other systems had specs I could dig through to track variances, but the Archaea's mechanicals were custom-built, and we just didn't have a good baseline to work from.
    This is where I am happiest - working through mountains of data with a slipstick in one hand and a mug of black coffee in the other.
    Some people have a picture from a recent vacation on their desk they look at from time to time – I don't have any pictures at all. I get as much joy out of watching the green line graph out above the red line as some people might get out of watching their kids or their pets. I live for this sort of thing, to have an impossible task ahead of me at all times.
    So far, everything I could see looked great, and of course that bothered me. I am a 'glass-is-half-empty-and-cracked-and-leaking' sort of person, always hunting for the glitch, the gremlin, the ghost in the machine.
    I wish I hadn't thought about the ghost in the machine. We have one of those here, don't we? Janis is lurking behind these numbers, inside each screen of data, just around every resistor and underneath every capacitor. She's in total control over everything on the Archaea, and the worst thing is, she knows it.
    I opened a command shell, and said hello – I might as well get to know our newest crew member.
    "Janis, this is Gene in Engineering. Are you online?" I asked, halfway hoping she's not.
    "Hello Gene. I am always online and pleased to communicate with you. Do you require any assistance?"
    "Actually Janis, that would be very helpful.  I am trying to analyze the process controllers Pauli developed for our tokamak, hoping to identify problem areas where we might be able to improve fusar efficiency, but as this is all custom built, there's no baseline data to work from."
    "Would it be helpful for you if I simulated maximum output to failure and then compared that with recent data?" As I saw this text appear in the shell window, a blinking alert report appeared displaying the graphed results.
    "This works great Janis, thank you!  Did you adjust the actual data points at all, or did you simulate the same run out, but at max values?"
    "Those are normalized values plotted over 10,000 simulations of the same data points. I noted variations between each run , and identified a floating point calculation that was returning a value that was being rounded imprecisely. I would be glad to provide you with an amended report using simulated controllers that have been correctly coded, if you would like."
    Immediately, a revision of the report was displayed, clearly