I know this saying 'no pain no gain', but the way how currently colonisation works is not sustainable. Neither good for CMDRs nor for the game itself.
This reddit post inspired me to write here. A suggestion to tweak colonisation mechanism further.
A bit of background
So far we see many 'zombie' systems that are not and will not be developed by an architect any further.
Reasons are plenty: simple 'too much effort with hauling', 'this outpost was built just to move away from the bubble anyway, i actually want to build my systems in XYZ sector', 'i don't want it / need it anymore'. Plenty of beautiful systems are at threat of being forgotten zombie systems with one station only.
Suggestions
1) Instead of relying solely on architects work (hauling...) - automate hauling and delegate it to NPCs that will (using existing trading network mechanics) deliver cargo to new stations.
Initial station still will have be built fully by a CMDR. But subsequent developments - goals set by an architect - will be automated by NPCs.
It will require more time to develop system by NPC efforts only so a CMDR would still have motivation to speeding things up by doing the hauling himself.
2) create a way to modify system infrastructure (upgrade existing stations t1-> t2, change station types agriculture-> industry, extraction -> refinery etc) so the architect is able fix some mistakes or to adapt system to changing environment (market, neighbourhood systems, own capacities and resources).
3) increase weekly payout for developed systems to give incentive to architects to develop it further.
4) create some metric on 'colonisation attractiveness' of a system (interesting POIs? amount of slots? variety of slot types?) - anyhow - with clear criteria on why a system is an 'attractive colonisation target'. Then have more demanding rules on developing such systems: system will be put to decay if is abandoned by an architect, or architect will be forced to sell the system at some (well known in advance) price, after a period of inactivity.
5) Another metric could be a 'developed system threshold' above which any penalties will not be applicable anymore.
6) create a ranking system for architects? if it goes below certain point - an architect can't claim new systems. And the opposite rule - if the ranking is good there could be more claims made - Reward architects for more effort.
We have 400 bln stars in galaxy, putting a leash on architects or penalising them for 'not developing system enough' seems ridiculous, yet having hundreds of systems as zombies, or blocked for further development is a waste too. A balanced approach is needed and any effort that has been made by CMDRs should be appreciated.
Thank you!
This reddit post inspired me to write here. A suggestion to tweak colonisation mechanism further.
A bit of background
So far we see many 'zombie' systems that are not and will not be developed by an architect any further.
Reasons are plenty: simple 'too much effort with hauling', 'this outpost was built just to move away from the bubble anyway, i actually want to build my systems in XYZ sector', 'i don't want it / need it anymore'. Plenty of beautiful systems are at threat of being forgotten zombie systems with one station only.
Suggestions
1) Instead of relying solely on architects work (hauling...) - automate hauling and delegate it to NPCs that will (using existing trading network mechanics) deliver cargo to new stations.
Initial station still will have be built fully by a CMDR. But subsequent developments - goals set by an architect - will be automated by NPCs.
It will require more time to develop system by NPC efforts only so a CMDR would still have motivation to speeding things up by doing the hauling himself.
2) create a way to modify system infrastructure (upgrade existing stations t1-> t2, change station types agriculture-> industry, extraction -> refinery etc) so the architect is able fix some mistakes or to adapt system to changing environment (market, neighbourhood systems, own capacities and resources).
3) increase weekly payout for developed systems to give incentive to architects to develop it further.
4) create some metric on 'colonisation attractiveness' of a system (interesting POIs? amount of slots? variety of slot types?) - anyhow - with clear criteria on why a system is an 'attractive colonisation target'. Then have more demanding rules on developing such systems: system will be put to decay if is abandoned by an architect, or architect will be forced to sell the system at some (well known in advance) price, after a period of inactivity.
5) Another metric could be a 'developed system threshold' above which any penalties will not be applicable anymore.
6) create a ranking system for architects? if it goes below certain point - an architect can't claim new systems. And the opposite rule - if the ranking is good there could be more claims made - Reward architects for more effort.
We have 400 bln stars in galaxy, putting a leash on architects or penalising them for 'not developing system enough' seems ridiculous, yet having hundreds of systems as zombies, or blocked for further development is a waste too. A balanced approach is needed and any effort that has been made by CMDRs should be appreciated.
Thank you!