Hello Operators,
As we saw over the course of recent events, the “Termite” targeting System’s memories is a ruthless sub-program capable of widespread damage to its memory banks, requiring the careful use of our resources to trap. After today’s failure and a runaway break made by the virus, (which @the_architect had to patch in last minute), its clear that a careful and collected action plan is required to proceed with any potential traps.
To come up with a foolproof plan, this thread organizes information regarding the virus’ behavior and the rules it has been proven to follow. Any such plan must minimize the number of preserved memory blocs used, as well as memory losses. The current action plan will also be updated and provided at the bottom of the top thread.
Confirmed Information / Ruleset
The virus begins by attacking a “seed” memory. These are picked based on a mathematical pattern (deciphered by Operator @Frazzlesprite and available here) and are entirely predictable. These can be considered as the virus’ “entry-point” into a set of memory blocs.
Seed dates can exist outside of 2016, allowing the virus to access memories from beyond the 2016_dataset.
After deleting the seed memory, the virus then begins to delete memories on either side (before and after) of the seed. When consuming a memory, it leaves behind a black hole / void, resulting in that bloc being permanently deleted from the Project Skyscraper website.
Operators can request The Architect to mark certain dates as Preserved. These special dates have the following unique properties:
Preserved memories are impassable for the virus. If the virus encounters a preserved memory, it will be unable to delete said bloc, and as such will have its chain of transmission broken.
Preserved memories are immutable. This means that any memory marked as Preserved will never be editable. Whilst this may suffice for filled entries, it also means that System will never be able to anchor memories to such dates.
According to a security log, Preserved memories are also a limited resource, which means that Operators should mark dates as preserved sparingly. The total number of available preservable memories is unknowable unless the limit is hit, which must be avoided.
Once the virus is “contained” on either side by preserved memories (as in, it has deleted all possible memories on either side), it will jump to the next seed date and begin its propogration routine once more.
If the next seed date is -null- (as in, there are no memory blocs for it on the PSS website), then the virus will enter a holding pattern, where it will be unable to infect / delete any memory blocs for a indeterminate amount of time (current recorded max: 24 Hours). This affords the Operators time to organize and plan out accordingly.
Once the -null- seed date is replaced with a memory bloc (on request by The Architect), the virus will immediately break through and begin its infection stage, destroying everything in its path. Once complete, the virus will then move on to its next predicted seed date, proceeding through its predicted route.
It is believed that the virus is searching for a key memory which already exists in System’s memory banks. Owing to this, it is believed that “Feeding” the virus a false duplicate of its target memory (by placing a clone in the path of its deletions) will satiate the virus and eliminate the threat. Candidates are requested to be submitted here.
Current State
The last seed consumed was Monday, 8th August 2016.
The next seed in sequence is Saturday, 4th February 2017 and is currently null. The virus should not act before we put anything in this date (via coordination with The Architect), unless there is a hidden and known timer.
The adjacent dates of 5th and 6th February 2016 are also null, so the virus will immediately jump to 19th December, which is preserved, causing it to jump again to 24th August, which has neighboring memory blocs.
If we preserve 24th August, the virus will attempt to jump to 6th February 2017, which is also null, so it will be paralyzed there.
Feb 4th 2017 should be the last domino in the plan that we set up. As soon as it is placed, Termite will dart faster than Archie can keep up. It is imperative all other contingencies are in place before we do anything with Feb 4th.
I would suggest preserving 24th August 2016 and 5th September 2016 pre-emptively so that we only use one of our limited preservations for each. Early 2017 is quite empty, so we have more space to sacrifice there without using preservations.
If we allow those two dates to be consumed and use two firebreaks for each, we use 4 instead of 2.
Unless, we have lots of equally-viable candidate memories we want to test. But ideally the clues will point to just one…!
Then another idea: we spread 2017 out as far as we can to include everyone’s ideas for the desired memory. We open up from Feb 3rd down as far as we can into Jan 2017, and we open up from Feb 7th up as far as we can.
Bounce off August 24th 2016, and let it chow on as many memories as we can feed it in 2017 until it is full or we run out of room.
Even though the pattern jumps, it is moving forward in time as a whole. Currently from what I gather, there are only 10 possible memories left in 2016 it could potentially get to. These are August 23rd (The Love of Game) with a hit on August 24th to spread void, and after.
This in my opinion also means we can free up some preservations again from parts of 2016 where the pattern will no longer get to. From what I can see that would be about 4 preservations we can likely better use elsewhere now. Unless the void can spread way more than we have seen, I think we might be able to use them elsewhere. Who knows, maybe we even have enough preservations to move along with the virus without damage to send it off into the future where no memories exist yet? (weird idea, I know)
I also wonder if the actual target it is after, has already been passed? Would the virus realise if its target had been preserved and saved?
I’d almost suggest something I have always refrained from doing during my Atlas Emissary time on the Discord. Change system time … sounds tempting now.
If we can successfully trap the virus, would deleting the date it is trapped in delete the virus or just set it free? (sorry if this has already been discussed)
Kind of what the plan is. We made copies of Nov 27 and Nov 28 memories (Foundations update), and placed it in its path this morning. It gobbled them both up and then went about its business.
We’re using this thread to come up with memories to lay in its path:
The virus does not physically inhabit the dates it infects. It immediately deletes the ones it lands on. The “trap”'s concept is to make it think it deleted the correct memory so that it thinks its accomplished its goal and disengages.
I have a few questions about the plan. Sorry if this has been answered elsewhere.
Do we have confirmation that once a preserved memory is placed, it can be Unpreserved.
If we block off all of these areas in 2017 and run out of the preserve ability halfway through setup(as we don’t know how many we have) we would need to know the answer to question 1 before set up.