Welcome to Keen Software House Forums! Log in or Sign up to interact with the KSH community.
  1. You are currently browsing our forum as a guest. Create your own forum account to access all forum functionality.

Long, slow death of the worlds.

Discussion in 'General' started by Ghostickles, Mar 1, 2017.

Thread Status:
This last post in this thread was made more than 31 days old.
  1. Ghostickles Senior Engineer

    Messages:
    2,077
    After a few months of playing a few servers I have noticed a sneaky little problem with the world files, they keep growing larger and larger and larger.
    On a server the blocks a players places will de-spawn shortly after their claim expires. This is great. The voxel deformation stays though. I have noted a few worlds with a very large file size but very little existing player builds. I have to assume this is all voxel deformation.
    It would seem to make sense to repair the voxel as well as remove the blocks after a claim expires, although the voxel repair could possibly take a bit longer, to conserve file size. Unlike in SE where a complete voxel refresh takes care of this problem; Medieval players earthworks are often an integral part of their fortifications. Simply refreshing the voxels would destroy hours to weeks of game play in some cases; filling moats and caverns otherwise used as a valid game play elements. Simply not an adequate solution for this particular game.
    What I would like to ask; is there a solution to the ever increasing file size due to abandoned voxel objects without disturbing active players?
     
  2. Forcedminer Senior Engineer

    Messages:
    2,227
    Hmm.... :/

    wonder if ME and SE could get a feature like what rust has.

    the feature known only has....."WIPE DAY."
    the final day on the server before everything is reset.

    but for the sake of ME and SE it could just reset the voxels in planets.... :/
    hmm....underground bases....

    wonder how hard it could be to add a planet reset feature with a feature that if a base has that....king statue and in SE's case a beacon attached it won't refill voxel around that base?

    hopefully this isn't another case of...decent idea...loved by fans...ignored by devs.




    below is a link to what the term in rust "wipe" means....in a nutshell....server reset...all bases...inventorys...reset...enjoy the new smoother server experience!
    http://steamcommunity.com/app/252490/discussions/0/365163686059456656/
     
  3. Monkworks Trainee Engineer

    Messages:
    28
    Interesting findings Ghostickles.

    I agree! That the voxel should restore itself after a timer completes. Could be set by the admin for maximum fine tuning. A few days to a few weeks, whatever.

    A side issue related to this; I have 6 active Claimstones. But a 7th is reported by the game. Using the option to logon near your Claimstone , takes me to nothing. I spawn into the voxel and then relocated outside the voxel. But there is no 7th Claimstone. You are witness to me moving from one site to another recently Ghostickles and this alternat Claimestone takes me back to my old base. I believe there is a faulty clean-up string here in the code. I propose that maybe this is also contributing to the slow death....having lots of Shadow/Ghost Claimstones, would inevitably enlarge the put through.
     
  4. Taemien Apprentice Engineer

    Messages:
    468
    World files getting larger seems like an issue. But is it really? I've ran a server with a few friends and its only ever been a few MB. How large are your servers' world files getting? I mean if it takes a few months to get a file to 100MB, I just don't see an issue there.

    But with that said, they could optimize it like they did in SE. My friend's DS didn't even get to 1MB after 2 months. And we did some pretty decent voxel alterations.
     
  5. Scorpion00021 Senior Engineer

    Messages:
    1,411
    Perhaps storage algorithms could be optimized to reduce world size. Does anyone know if ME sends the entire planet's voxel deformation data to players when they join? If so, perhaps this could be modified to only updateplayers with deformation data that would be reasonably visible.

    The point here being: The disk size of the world should not matter, within reason, if it isnt being transmitted at all times.
     
  6. Ghostickles Senior Engineer

    Messages:
    2,077
    Already seeing 11 and 13 MB world files. These can continue grow at a rate of about a MB a week, although it goes faster or slower with player load. Connection and loading time increases with file size.
    In comparison Fountain Core, one of the longest running, most populated SE servers has a world file download of just over 500 KB.
    Its an issue. :)
    --- Automerge ---
    This ghost-claim thing is weird, maybe file a bug report?
     
  7. Monkworks Trainee Engineer

    Messages:
    28
    maybe a network hiccup missed the clean up on the waypoint when I deconstructed my claimestone... Actually, i let it timeout. Therefore, to be more accurate and because this is a known issue, with the 50 days thingy bug, claimstone..we should wait.
     
  8. Taemien Apprentice Engineer

    Messages:
    468
    SE has a new set of netcode and other optimizations I don't believe ME has yet. Once those go in, we'll likely see an improvement.

    There's alot of things ME has that SE needs and vice versa. I'll be happy once both games have both.
     
    • Agree Agree x 1
Thread Status:
This last post in this thread was made more than 31 days old.