autopilot.js: Focus on hack augs only once we have enough augs to unlock Daedalus

[alainbryden]

In the interest of exiting the bitnode as soon as possible, once we have enough augmentations (count) to unlock Daedalus, we should really only go for hacking augs to meet the last requirement as fast as possible (money comes pretty easy)

• Note there 53 hack augs, 45 of which are from "easily obtainable" source (doesn't require the -a arg in faction-manager). This may vary with gangs randomization, but won't be far off.
• We need to be able to manually run faction-manager --purchase (rather than relying on ascend.js) so that we can choose the arguments we want (such as stat-desired)
• We have to change faction-manager to only output the summary file when requested, so that other runs don't overwrite the version from the args we chose
• We need to add to the output file some extra information, like how many desired augs total there are, how many within rep, and what the cost to the next aug is. We can use this information to make better decisions
• We might have to sanity-check that if we're buying up the last of the hack augs, or the next aug is significantly more expensive (e.g. Qlink) we don't let the "min augs per install" config hold us back.

[driderr]

Autopilot recently installed augments just a few minutes short of getting the last aug in one of the factions while gaining rep (to 100k), which was a huge waste of rep/time. Maybe bullet 4 will help fix that.

[alainbryden]

Autopilot recently installed augments just a few minutes short of getting the last aug in one of the factions while gaining rep (to 100k), which was a huge waste of rep/time. Maybe bullet 4 will help fix that.

Hmm, currently the only script that knows anything about how close we are to earning enough rep with a given faction to get its next/last desired aug is work-for-factions, and it doesn't have a line of communication with autopilot.

You're right, and there are a bunch of edge cases like this to hammer out over time, but right now I just have them as TODOs in code because I'm still trying to hammer our the "80" of the 80/20 rule.

For now, if you have time to micromanage ascension, I suggest diabling autopilot and just running daemon - call ascend.js manually when you think it's a good time. Autopilot is currently just intended to do an "okay" job in my stead while I'm AFK