Player riposte nerfs and why?

Discussion in 'Time Locked Progression Servers' started by Hamshire, Oct 1, 2019.

  1. Lisfin Journeyman

    Its true we don't know exactly what changes they have made, but being a programmer with 20+ years of experience, I know for a fact comparison checks can be extremely fast, so removing 4 of them would not make a huge difference.

    The true reason this would reduce lag is because its reducing a entirely new attack round when you would of riposted. That alone, not having to do a entirely new offense/defense check would easily surpass the previous 4 checks by a factor of 10-100x.


    I never demanded anything, you are implying something I never said. I only asked, if they have benchmarks, which I honestly believe they do not, to release them, I never once demanded it.

    It has nothing to do with trust, I honestly don't think they have benchmarks to show that the actions they are taking will have the effects they are hoping for, not that its doctored or they are lying to us. I already said I don't think they are lying.
  2. Lisfin Journeyman

    Yes we don't have all the information, but to say we don't know anything is very uninformative and misguided. Plenty of code has been released over the years, some even showed exactly how the attack formula worked.Lots has been backward engineered and is freely available online if you look into code.

    No we don't know what the servers are running on or how many are on one machine, your right. But to think they are worse than 2008 servers would probably be unlikely.

    82 BILLION instructions PER second on 2008 hardware.Yet removing 4 checks per attack is going to make a huge difference. Its not about the dodge/block/parry/ripose check, its about the next attack check that they are saving the processor time on, AKA making you attack less for server performance.

    I have been programming for over 20+ years, have developed multiple games, have worked on many pieces of software that required lots of bench marking. It's not hard to extrapolate how much process time is needed for certain tasks like math problems or getting variables.

    It would be like you saying that because I only worked on Ford cars for 20+ years, I have no idea how any other car works. That is a very illogical statement.
  3. Accipiter Old Timer


    Yeah, well, I'm a VP of R&D and I know I don't know sh*t.
  4. Lisfin Journeyman

    So you come on here and tell other people that they don't know $hit because you don't know it?

    Some people have been more involved with the EQ code than others. I have ran emulations servers, and have worked on the game code that has been reversed engineered over the years. The attack formula was posted on the forums from a dev a few years ago. There is plenty of information out there if you care to look for it.

    Do I know everything there is about the software they have and run, absolutely not. But to say "we can't make informed posts" because we don't know everything is absurd.
  5. Accipiter Old Timer


    Whoosh!
  6. Laronk Augur


    Since you know all the $hits of all the things with the 100 billion gigawatt transflux capacitors. Explain to me this, why does windows sometimes have trouble opening a simple app like a calculator.

    Sometimes things are coded so poorly that what should be simple actions are not simple actions, it is in fact possible that calculating a miss independently reduces resources on a server. Also we don't know how many zones are run by each vm (assuming they use vms now) and how many vm's there are per server or how many combat actions a single server is tracking at a time it might not be a small amount of calculations.

    I really feel like you're just thinking about one group fighting one mob on one physical server.
  7. Lisfin Journeyman

    Simple and honest answer, because you have been downloading naughty videos without protection. Might want to run a virus scan. /s

    I understand, some programs can be designed poorly and have terrible performance. But to say this is a good way to save server performance, is IMO a terrible choice. There are many other things that can be done to increase the server performance compared to changing the way 4 cached comparison checks would work.

    The attack check for offense/defense following the riposte on the miss would take 10-100 times more processing time compared to 4 simple math checks TotalDodge >= MobTotal etc.

    Most of the "saved" performance would come from less hits, not from the comparison checks.
    All this change is doing, is making you take longer to do the same amount of hits to kill the mob. The server will end up doing more calculations due to the fights taking longer.

    I have already shown, a server in 2008 could do up to 82 BILLION operations a second. even if you had a million people swinging a 100 times a second and you removed the 4 comparison checks, you would have LESS than 400 million saved operations, still a long way from 82 billion.

    We all know, 1 million people are not playing on a server, and they don't attack at 100 attacks per second. This is my point, doing math checks will not save much time, computers are very fast at doing a comparison, especially if the variables are cached. Not calculating another entire attack from the riposte is the actual time saver, AKA just making combat take longer.
  8. Machen New Member


    No. That's not how this works, at all.

    If fights take longer but have the same overall number of calculations, that's a reduction in calculations per second. That's a reduction in server load.
  9. Lisfin Journeyman

    So if fights take longer, you don't agree that more hits/misses will need to be calculated?

    Yes I can agree with you, that amount PER second will be reduced, however if more people are in combat longer, it can increase the server load because they no longer kill the mob as fast, causing them to swing more times and stay in combat longer, creating even more processor work than before.
  10. Laronk Augur

    But what you're not understanding is that even a simple divide isn't one operation it's many operations even at the assembly level. The Everquest server is abstracted pretty far from assembly level. There's an operating system in between and then there's some kind of engine that it's running on.

    Now there could be even more things there could be an interface to that engine.

    You have to actually call the goblins attack from somewhere and then call the tanks's evasion AC from somewhere those calls are not just single operations. Then there's some kinda random dice roll function, also not a single operation.

    I guess we're talking about riposte which is a swing that has all those operations too,

    Even if we're just talking about the single "did I" riposte we're grabbing a stat from somewhere which is a call to somewhere since there's a skill involved, and there's still at least 1 multiply or divide function and its likely not happening at the assembly level so there's more operations in between.

    My example earlier with calculator being slow was more that the calculator isnt running at the processor level and neither is everquest.
  11. Lisfin Journeyman

    True, they might not be a single call at the processor level, however if these variables are cached, and I am betting most of them still are at the start of the combat shuffle bag creation, most of the variables are simple comparison checks, which are still extremely fast, and can be even faster if doing a byte comparison.

    Excluding 4 comparison checks will not effect performance nearly as much as the resulting attack you are removing.

    This is why I was asking if they have any benchmarks of this because I don't think it will have the effect they are looking for. Also considering, combat will take longer, allowing more swings to be calculated total.
  12. Machen New Member


    No, no, no. This is not how any of this works.

    Outside of raid content, which is a relatively small minority of what happens on a server overall, players don't keep playing until they have completed a certain number of fights, then log off. If fights for some reason are taking longer, then players will complete less fights in the time they have to play. If tomorrow, devs reduced the number of hits in all combat by 50%, thus doubling the time it took to complete each fight, players are not going to suddenly double their play time to result in the same total number of hit calculations.

    But I'll edit to add, that EVEN IF your premise were true and we'd end up with the same total number of calculations, spread over a longer period of time, that would STILL result in an overall reduction in server load, because many of those fights would now be taking place at less busy times of the server. Toward the end of prime time, when people are logging off, or the wee hours of the morning. Server load is obviously most critical when the server population is highest.
  13. Lisfin Journeyman

    Your assumptions are incorrect because you are forgetting about the mob attacks.

    If you reduce your DPS by removing riposte on miss, the time to kill a mob is going to increase, I hope we can agree on this.

    If each mob takes 5-10 seconds more time, how many more attacks are there in them 5-10 seconds?
    Yes it takes the SAME amount of attacks for the player to kill the mob, however the mob is attacking you for the 5-10 more seconds each time also.

    If it would of died 5-10 seconds sooner from more dps, all the attacks/misses it would of done to you would of never happened. If you have many groups taking 5-10 more seconds to kill mobs, and the fights last longer, there will be more calculations from all the extra mob attacks.

    Yes if it was only you attacking, it would not matter, but the longer the fight goes on, the more attacks the mob will do against you, causing even more processor time compared to killing it faster.
  14. Corydon Augur

    Here are a few full stops free for you to use in any of your next posts. You are welcome and thank you very much for using them.
    -----> . . . . . . . . . . . . . . . . . . . . . . . . . . . . <-----
    Nolrog likes this.
  15. Hamshire Augur

    Doesn't look like this mistake of a change is going to be reverted and I don't see a donation option to help improve servers so the two accounts ive had active since RAF came online what feels like 10 years ago? Are now canceled and this one im sure will join it soon.
  16. Bewts Augur

    So they just reordered the calculus such that a miss stops all subsequent avoidance calculations? This would be the equivalent of reordering a DA check to be the first check instead of one of the last...

    Outside of marginal losses of DPS from ripostes for player characters; I fail to understand why this is such an concern for gameplay mechanics?

    As a tank in say week one of Velious I’d be thrilled that AoW won’t be sneaking in more max damage rounds via ripostes off my misses between its quad rounds.
  17. Zanarnar Augur

    Cool, can I have your stuff?
  18. That0neguy Augur

    They aren't upset over that. They are upset that they cant pull massive pulls of grey mobs and ripo them down.
  19. Hamshire Augur

    Begging for anyone's stuff on the EQ forums in 2019 :D
  20. Zanarnar Augur

    I could say the same about rage-quitting because you can't riposte grey cons to death as easy :rolleyes:

    Also, asking <> begging