How long we want the command delays to be?

Discussion in 'Planetary Annihilation General Discussion' started by qwerty3w, February 20, 2013.

?

How long you want the command delays to be?

  1. none

    77 vote(s)
    82.8%
  2. 0.3s

    10 vote(s)
    10.8%
  3. 0.5s

    2 vote(s)
    2.2%
  4. 0.8s

    0 vote(s)
    0.0%
  5. 1s

    1 vote(s)
    1.1%
  6. 1.3s

    0 vote(s)
    0.0%
  7. 1.5s

    0 vote(s)
    0.0%
  8. 2s

    0 vote(s)
    0.0%
  9. 5s

    0 vote(s)
    0.0%
  10. more

    3 vote(s)
    3.2%
  1. Gruenerapfel

    Gruenerapfel Member

    Messages:
    161
    Likes Received:
    0
    i think to make the game more fair between players with differnet connection a "forced" latency would be nice. everybody has an minimum ping of like 100 and only players with realy crappy internet will suffer.
  2. BulletMagnet

    BulletMagnet Post Master General

    Messages:
    3,263
    Likes Received:
    591
  3. drsinistar

    drsinistar Member

    Messages:
    218
    Likes Received:
    0
    No. Lag compensation is a bad idea.
  4. godde

    godde Well-Known Member

    Messages:
    1,425
    Likes Received:
    499
    Why? 100 milliseconds isn't much. SupCom had 500.
    It is easier to adapt to a set delay rather than one that is changing all the game.
    How does Starcraft do it? Micro is highly proficient in Starcraft and if 1 player have much higher latency to the server than it can really be considered a disadvantage.
    That said, PA isn't going to focus as much on tight micromanagement so I doubt that high latency is going to be a big disadvantage.
  5. igncom1

    igncom1 Post Master General

    Messages:
    7,961
    Likes Received:
    3,132
    Just have acceleration rates for units.

    Problem solved.
  6. KNight

    KNight Post Master General

    Messages:
    7,681
    Likes Received:
    3,268
    That should already be part of the simulation aspect, but I'm don't see how it solves anything?

    Mike
  7. igncom1

    igncom1 Post Master General

    Messages:
    7,961
    Likes Received:
    3,132
    :lol: Exactly!
  8. sylvesterink

    sylvesterink Active Member

    Messages:
    907
    Likes Received:
    41
    There are better ways to do away with micro than delaying responsiveness to a command. ZK implements a few.
  9. godde

    godde Well-Known Member

    Messages:
    1,425
    Likes Received:
    499
    igncom got a point in that low acceleration makes high intensity micro like stutterstep and projectile dodging harder to perform.
    Even though SupCom have 500 ms delay you still see intense bombermicro and projectile dodging at times. You just have to give the commands earlier.
    If you want to decrease the viability of projectile dodging it would be good if you added turn rate acceleration as well. You can't jinx as fast back and forth then.

    Although I'd still say that parts of Zero-K are micro intensive. Like raiders vs raiders.
  10. BulletMagnet

    BulletMagnet Post Master General

    Messages:
    3,263
    Likes Received:
    591
    Low acceleration can be a brilliant idea, but the Supremilated mod (gosh, that's getting old now) tried it and it broke the path finder when you had blobs of tanks. It broke doubly hard when you wanted to formation move them too.
  11. igncom1

    igncom1 Post Master General

    Messages:
    7,961
    Likes Received:
    3,132
    How so (I never had a chance to play it) as I am curious?
  12. godde

    godde Well-Known Member

    Messages:
    1,425
    Likes Received:
    499
    Haha! Don't get me started on how much more complex pathfinding have to be with low turn rate acceleration. But... Impossible is nothing!
  13. BulletMagnet

    BulletMagnet Post Master General

    Messages:
    3,263
    Likes Received:
    591
    The path finder wants to keep units close to each other (kind of important when you want a formation). Units in the front edge of the blob could freely drive off, while those behind are blocked momentarily by the front. Since units have a slow acceleration, those blocked take too long to move, then the front units stop and come back because they're straying too far from the formation.

    Now the front units are back in the blob, and blocking the units again. Derp.


    I suppose this could be fixed by increasing the distance that units were allowed to stray before returning. But at the time, I don't think anyone knew if that was possible.


    Also, this was without flow-field pathing. Whether or not the same thing would happen with FF is purely speculation.
  14. godde

    godde Well-Known Member

    Messages:
    1,425
    Likes Received:
    499
    The problem in SupCom was that a unit driving into another unit from behind would lose all its' speed. This doesn't happen with flow-field pathing so I doubt it would be an issue.
  15. drsinistar

    drsinistar Member

    Messages:
    218
    Likes Received:
    0
    He was suggesting forced latency. Would you like the game to raise your ping just because? I certainly would not. I'm not arguing over whether I like micro or not, I don't really care. I do care about a forced latency though.
  16. godde

    godde Well-Known Member

    Messages:
    1,425
    Likes Received:
    499
    I would prefer no latency at all but that is most likely not technically possible.
    SupCom had half a second forced latency. If the host could set the latency to a tenth of a second to keep everyone on equal terms while the orders are executed on my computer one tenth of a second after I give the commands, I would prefer it that way.

    The argument that PA isn't going to be microintensive could go both ways:
    "It doesn't matter that your orders are executed with latency delay because micro isn't that important"
    "It doesn't matter that your orders are executed with 100 ms delay because micro isn't that important"
  17. drsinistar

    drsinistar Member

    Messages:
    218
    Likes Received:
    0
    I understand that SupCom had a forced latency, I'm just saying that I don't want it to return in PA. There is no reason for there to be a forced minimum latency. However quickly you can send information to the server should be your ping, not some predetermined amount. Whether it makes micro hard or not, really does not bother me. I honestly do not care about micro's presence. Besides, I imagine that it's one less line of code to not include forced latency. Uber could better focus their attention elsewhere.
  18. Raevn

    Raevn Moderator Alumni

    Messages:
    4,226
    Likes Received:
    4,324
    Forced latency is only needed for peer-to-peer models, since in that system all the clients must be sync'd to each other (so you'd get crazy lag spurts without some limit). However, not including it in PA in some way would give people with lower pings an unfair advantage. I'd imagine some middle ground is needed (say an enforced latency, but reducing it from the 500ms it was in SupCom).
  19. asgo

    asgo Member

    Messages:
    457
    Likes Received:
    21
    an option for a short forced latency for sake of fairness might not be a bad idea.

    Besides, the client server model might have a say in it, independent from your ping the server's effort to keep things synchronized might add some delay anyhow.

    We should consider ourselves lucky, that they probably don't implement the bureaucracy communication model: write, sign, fax, end of work day, copy, stack,weekend, send response ;)
  20. garatgh

    garatgh Active Member

    Messages:
    805
    Likes Received:
    34
    Im realy starting to hate how everyone just ignores my posts now days. Maybe i should start borderline trolling just to get some attention. :roll:

    Anyway, time to repeat myself: Its been confirmed (well somewhat confirmed atleast) that a delay will be present, it will be shorter then supcom's but still longer then TA's.

Share This Page