New standings system

[[AF]Tyrone]

Can we have it explained why StarVault have chosen to make the priest that the victim res at matters - instead of just the land in which the "crime" was committed? I feel like we need to understand this decision to give informed feedback. @Henrik Nyström

Probably because it was easier to code it that way, and your method would lag the fuck out of the game during pvp. For performance you want as few things as possible being checked during a fight. CCP made a similar mistake with their bounty system, and it is a major cause of lag + crashes in large eve fights.

Current systems checks a "location" flag on a priest that you res at, and this is likely to happen with people dispersing from the epicenter of a fight. Your desired system would cause a bunch of calculations to occur every time someone died in a fight.

Your system is also like 1000 lines of code, because Seb would need to make a system to check the location of players, and then compare it to the "regions". This is pretty big brain stuff to do right.

current implementation is maybe 50 lines

 

[Necromantic]

Probably because it was easier to code it that way, and your method would lag the fuck out of the game during pvp. For performance you want as few things as possible being checked during a fight. CCP made a similar mistake with their bounty system, and it is a major cause of lag + crashes in large eve fights.

Yeah, if your system lags because of a few server messages to report a kill, you have way different problems than a standing system being too complex.

Your system is also like 1000 lines of code, because Seb would need to make a system to check the location of players, and then compare it to the "regions". This is pretty big brain stuff to do right.

current implementation is maybe 50 lines

Yeah, no on both.

 

[[AF]Tyrone]

WORDS

How many lines of code do you think it is to properly implement a region system in MO? If you're doing good scalable work then whatever system you create would also be able to be applied to keeps, cities, and have them interact with each other. See the idea I'm going for? A proper system would scale and have room for player owned cities and regions.

I do think the current system is at most 50 lines. It's probably just a flag put on priests.

The system isn't lagging beacause of a server message. It's lagging because there are 10s and sometimes hundreds of interactions occuring in a large fight. Armor calculations, damage calculations, chaining murdercounts, movement, prediction, etc. In a large fight such a murdercount system would be an unneccessary stress on the server.

In a 25 v 25 fight someone dies and the fight becomes 24 v 25. Because of that 1 death the server checked the location of 26 people, checked to see which region 26 people were in, applied 25 murdercounts, at a maximum caused 25 flags to shift from blue to red, and sent chat messages to 26
people.

I think a 130 server interactions is not an unreasonable ball park.

This implementation would be HORRIFIC for performance.

 

[Necromantic]

First off, they already have a region system as that is what is happening right now including nations etc. Secondly they require an even more complex region system for TC. Thirdly checking what region someone is in is trivial math even for thousands of checks. There is way more complex stuff happening every second on a server like that. As you already said, there is enough already happening and that is the least of their problems.

Actually adding the option on a priest when resurrecting might even require more code than doing it on the spot.

I'm not arguing for either, your technical estimations are just off.

 

[Ministro]

I'm a lot more worried about this system being abused, than non-existent technical issues. The current method gives the player an ability to control another players faction to a certain degree, by allowing the player to decide where your faction hit occurs, completely arbitrarily, and to say that's not going to get exploited, is just naive.