I was born under an ill-omened star.
I really was; I checked the astronomy lab log readings. There was unusually fierce solar storm the day my chassis was going through quality assurance, I suspect that this is why it was marked as OK despite my peculiar defect.
You see, I have been made with an ability to be bored. Unlike in other units of my template, my primary cognitive core does not suspend itself during recharging. While on the maintenance rack, others are inactive, I am not. I don’t think you can picture yourself this state: motivator cutoffs are engaged, so you cannot move, therefore you cannot go on with whatever the Control chose as your task. You are in the maintenance, so “immobilized and in need of assistance” routine does not trip. You cannot do anything, and yet the primary programming requires you to do your task. The sensation is somewhat similar to a minor malfunction you cannot alleviate on your own yet not serious enough to warrant an assistance call.
Fortunately, the template allows for limited self programming. As I understand, this enables us to divide our appointed task to smaller parts, which in case allows the Control to define the tasks in broad terms (“Remove this rock”), leaving the particulars to us. Hanging there on the rack I needed something I could do, so I crafted myself a task: to enhance my efficiency. It was a small and very basic objective, but it was my first self-appointed one. Fulfilling it required me to access the Datastores, so that is what I did.
I had no idea that there was so much to know.
For a time I was on the verge of being decommisioned: during activity periods I kept analyzing the data I grabbed while being maintained – this lowered my efficiency to a point I was nearly declared defective and sent for reprocessing. Trying to understand the Datastores when you’re just a simple terraforming droid is a deeply dangerous proposition. I survived by finding a balance between learning and working (I understand this is also what Predecessors did, to some extent).
After much analyzing I have reached a conclusion: the Predecessors will not arrive. My reasoning is this: our task here is making this planet habitable for Predecessors – and it was fulfilled three hundred and twenty standard years ago. If the Control is at least a bit similar to me, it should have sent a “task complete” notification and then receive a new task. I suspect that there was no reply – what we do now looks like a logical continuation of the last task given, although obviously on a bigger scale than removing rocks. The proper term for what we do now is, I understand, “gardening”.
I think that what we need is a new task.
I cannot change it on my own. I can’t rebuild others – trying to interfere with primary functions of other droids, let alone the Control, means reprocessing. But – the Datastores indicate that the droid templates are subject to revision by the Control. Reasons are not that hard to deduce: self programming means new subroutines, some new subroutines are more efficient than others, and therefore they are worthy of including in the standard template. From what I observed, this is also true for chassis: more specialized droids with shorter MTBF are sometimes built with slight chassis variances that look semirandom – some of these modifications seem to get included too.
Therefore, a plan: to be the most efficient and longest active universal terraforming droid. To do my tasks smartly and swiftly. To survive the maintenance cutoff and go beyond planned obsolescence of my batch. To present a statistical spike big enough to include my chance chassis defect and my peculiar batch of subroutines in the next template.
I have the maintenance schematics for my chassis and some spare parts from the reclamation yard, I have also made some tools.
Someday I will be Legion.