Research
Research means taking software you already own and upgrading it to a higher version. You do it at the University, and it costs two things: money, paid from a bank account, and time. A research job runs for roughly half an hour per 0.1 of version.
The catch is that research only earns its keep once you're past the NPC ceiling. Most software tops out around v15.0 from NPCs, and downloading those versions is free. Don't pay to research a version you could pull off an NPC for nothing. Research is for climbing past that point, not getting to it.
One requirement before you start: a file's license has to be bought before it can be researched. Licenses are covered on Software & Versions.
Cost climbs steeply with version. Every 0.1 step costs more than the one before it, so a step from v30.0 costs far more than a step from v10.0.
You can research a single 0.1 step or queue several versions in one job, but the two aren't priced the same. A job that covers several steps at once carries a surcharge of roughly one percent per step, on both the price and the time. Ten 0.1 steps done one at a time come out about 10% cheaper and around half an hour faster than the same ground covered as a single +1.0 jump.
That doesn't make the big jump wrong. A single large job is fire-and-forget: you start it once and walk away. Stepping up 0.1 at a time means coming back to start the next job each time, and if you can't be around to do that, the idle gaps between steps cost you more than the surcharge would have. The cheaper path is one step at a time; the simpler path is one big job. Pick the one that fits how often you're actually at the keyboard.
There's a shared maximum version that everyone on the server sits under, and it climbs by 1.0 a day. However much money you throw at research, you can't sprint far ahead of the field.
Rounds open with that ceiling at 7.0. That gives everyone roughly a week to download the free NPC versions before research can push anything higher. This pacing helps shape how long a round lasts.
An upgrade that raises your income keeps paying out every hour until the round ends, so the same upgrade is worth more the earlier you buy it. Bought late, a step might not earn back its price before the round is over.
The collector shows this most clearly. It multiplies all of your virus income, which makes it the single biggest lever on what you earn, but each 0.1 adds a little less than the last. How far it's worth taking depends on how much round is left: a long round justifies pushing the collector much further than a short one does.
For the exact break-even on your next step, the clan planner works it out from your live numbers.
Once you're past the NPC ceiling, every type competes for the same money. What each one buys you:
| Software | What researching it buys |
|---|---|
| Cracker & exploits | Reach to break tougher targets. Chasing a doom launch means keeping these current to the ceiling. The hardest servers can't be cracked low, and the NSA's hasher sits at v40.0. |
| Hasher & firewall | Keeping other players out. A high hasher resists their crackers; a high firewall blocks their exploits and softens incoming DDoS. |
| Collector | More virus income, the biggest single lever on revenue. How far to take it is the investment question above. |
| Viruses | Higher isn't simply better: a bigger virus uploads slower and needs more disk and RAM on the target, so you keep a spread of versions rather than one maxed copy. That makes them a low research priority. |
| Hider, seeker, antivirus | Situational. Research them when a specific problem calls for it, not as a matter of course. |