hpongledd

  • Content count

    137
  • Joined

  • Last visited

Community Reputation

56 Excellent

About hpongledd

  • Rank
    Member

Recent Profile Visitors

292 profile views
  1. Game Lag Tests

    I think you are simplifying too much here. From a job perspective its a guaranteed 2 trips, from a task perspective its 1 trip per task. Pathfinding doesnt know about the whole job stuff, so it will treat everything as 1 trip. The job perspective can be brought in by Klei and be considered for calculation, but we are talking about top level archtiecture here. Something that had to be decided at the beginning of the project. If they choose another approach, changing it will be close to impossible. I dont know which parameter Klei choose to be dependent on for path calculations. I hope the people arguing know what Klei choose. Again, simplified to a point its not practiable anymore. For example: a dupe is standing on a tile where a low priority task is, but two tiles away there is a high priority task. Which one should now be choosed? Ok, you could say "pick the highest priority task in a range of 8 tiles". Problem solved. Now, 9 tiles away there is a high priority task, 3 tiles away there is a low pririty task. This low priority task happens to be storing, which include walking to a storage container, maybe 30tiles away. do you recalculate the possible options when you are at the low pirioty task tile or do you execute it nonetheless? Questions, questions, questions, trade offs to take a lot of not foreseen situations, which raise another cople of questions and another buch of trade offs. Maybe a travelling salesman approach wouldnt be that bad. SInce you can calculate what it would cost me solving a low priority task before a high level task and if its more worth doing the high priority task first.
  2. Game Lag Tests

    Lovely. Its called trolling and playing with words. Beside that it should go without saying that the part i find hard its not implementing the algorithm but rather using in a "real environment". Theory is a nice thing but unfortuntely i never studied in my life and im more of a idealistic/pragmatic person. I never talked about game environments, did i? But yeah, i'll give you that one, since you are right, i guess I seem to be saying, that i never carred enough to deep dive into the topic and to get a feeling how programming somehting like that in a real case scenario is. So if someone says "its not that difficult", he should maybe consider the implementation in a real case scenario not some "proof of algorithm" solution. Im smart enough to realize that i also need two hours to create a console application, copying an algorithm into it and maybe an input for start/destination, but where is the benefit of this for a real case scenario? I tend to be arrogant while trolling. But a quick search on google brought me to a bachelor degree thesis called "UTILIZING SWARM INTELLIGENCE ALGORITHMS FOR PATHFINDING IN GAMES" Without reading the whole paper, i'll just quote this part from the Summary: To remain fair, as far as i understood, his conclusion is ACO is faster at calculating but less optimal paths were taking. Also to under some circumstances calculations were faster with A*. So....my trolling holds up, to some degree, to a 5min scrutiny. Regarding using ACO for ONI: His tests were made with an 512x512 grid sized map. If i recall correctly ONI maps are smaller? SI path calculation tends to be becoming slower with the ratio of obstacles increasing. As far as im concerned i would say the trade off less optimal paths for faster performance is acceptable and could be a better solution for ONI, if someone would run tests to check at which rate of obstacles A* and ACO will have equal speed, based on this number ACO could provide more value. Only because A* is the "standard" it doesnt mean its the best option and also only because ACO isnt mentioned as a possible solution on your wikipedia list it doesnt mean it cant be one, ever heard about "thinking outside of the box"? The statement of my intial post was "Klei wont do anything about it, because they dont have a ******* clue either, elsewise someone would have done it, a long time ago." Another reason could be (probably more realistic) that they can but wont because its not worth it. Why? It costs them a lot of resources and the benefit is, lets generously say, 10% of the player base will have better frame rates on their big bases. Therefore, i think its only fair, if i troll back isnt it? Its the same kind of discussion we had a couple of months ago where R9MX4 (or whatever his name was ragequitted the forum because Klei did not thread him right and ignored all of this wonderful exploits of bugs he made, displaying serious issues with the game. (for the ones that think this is sarcasm, it isnt. I truely admire what he did for the community and Klei, he has my respect). But damn son get a grip and keep your **** for yourself. Klei is serving 1.5 Millions Players, not only those 30 narcisstics persons that love ONI and get offended if Klei doesnt say "Thank you!" everytime they post a sophisticated bug which only appears with deep game knowledge.....And in the end, maybe for game developers this one is less true, its all about money, they wanna make money, period. They dont care about you, me or the whole community. The handle our **** because they know, we give them money, therefore they need us to pay the bills and we need them for our entertainment. Or would you say those conclusions/assumptions are wrong? But again, please remember: i dont have a ******* clue about what im talking ;D edit: god....im enjoyiing this way too much....
  3. Game Lag Tests

    As i said i dont understand a single word, but i can imagine that its easy stuff for a mathemitician and maybe implementing the function is trivial, but i can also copy paste a function into my code, useless **** but hey "i implemented the function" ;). Also i dont think you need two hours most. I dont understand a thing about it but the sole fact it was only google and a professor who wrote this algorithm i think it needs a little bit more then two hours but in case you really are that good, here waits a big business opurtinity for you (my former company would be more than happy to pay 1000 dollar/month if you can provide a better and more performant solution than they currently have, wanna have the contact? Not to talk about all the other taxi, deliver companies who are struggling with the exact same stuff) ähhm ok, as i said i dont know anything about it, but im a software developer and i guess calculating the best routes over a whole country, everytime an order comes in, with an average output of around ~10 order per minute is more cpu intense than the.... lets say retrigger a calculation over 20 possible routes every tick. Therefore i concluded that you have exactly the same knowledge about the topic as i do. Ergo none. edit: and not to forget you solve a an issue that a lot of mathematician are struggling with in "2hours most".
  4. Game Lag Tests

    Wihtout reading the whole topic...i'll just post my two cents. Klei will most probably do nothing about pathing optimization. Reason? Although its just an assumption, im pretty sure it is like that, they are using a framework. Blackbox. Period. Unless Klei has a (really) big brain in their company nobody will understand whats going on in this framework. Google "ant colony optimization algorithm" to see a glimpse on how complicated this **** can be. I assume they already have contact with the developers and most probably they dont give 2cents about klei. Conclusion? Nothing will happen. Funny side story: We needed a framework with the above mentioned algorithm. We could only find two ppl/company who did that. One was google and wanted 10k/year for a licence. The other one was a random professor at MIT who gave it to us for free.
  5. Maybe im too dumb to understand, but you say from top to down every 6th tile but the last pictures shows a bigger spacing. I thought the count starts at the tile and not where the plant ends? Or what am i missing here?
  6. PIPs not planting?

    how did you create the tiles?
  7. stored is ok. loose is bad. If you can fill your storage compactor with only one kind of debri even better (dont ask me why)
  8. ??????? Debris and Path are still the biggest issues and by fixing those i easily get my colony from 15-20fps to 30-40.
  9. as i already mentioned yesterday. I understand they need a gantry if entering with a atmo suit but why cant they leave without a gantry if they have a jet suit? Makes no sense imo.
  10. [Game Update] - 299550

    Yeah they enter without any gantry, but i could observe how sometimes they refuse. Reassigning helped. Currently im deconstructing and reconstructing the command capsule for exiting. And most probably the reason Klei doesnt want to deal with this bug (which is around since loooooong) is because its annoying and they dont have any idea how to solve it OR they just dont like me. Eitherway, i'll post this to any game update till they ban me from their forum or they fix it.
  11. [Game Update] - 299550

    It happens also on the edge of the map and my robo miners (2-3 at a time) just start lasering it and it keeps going and going, till a dupe gets there and digs it out. I know that. My problem is: 1. i dont have any gantry and i dont want to build them 2. i designed my rocket silo without gantries 3. i restricted the pathing for jet packs to solely the rocket silo for a reason
  12. [Game Update] - 299550

    I'll just quote myself. Bugs encountering on this preview: - Oxylite doesnt get grouped and lays on the ground with 10.2kg packages. Sweep command doesnt work for some "packages", i need to manually select it and click the sweep button - Dupes with Jetpacks wont exit a rocket, even with a gantry they just reenter the rocket instantly (please please fix this! i hate gantries) - Shove vole doesnt move an inch (navigation looks correct) and lags at "expelling waste" - not sure if intended but the message "Unrefrigated food" opens the food tab - it appears that sometimes regolith tiles, especially at the edge of the map and on a rocket engine get digged out forever without any success (like snow tiles some time ago) - Door settings for dupes returning from space are resetted and need to be reconfigured (very very annoying and sometimes deadly) - Sometimes debris falls through tiles or gets stuck in tiles - if liquid tanks are set on an amount below 900kg, clicking on it periodically ejects it - rocket oxydizer tanks are filled but rocket is in space - copying door settings causes the game to crash Nice to have: - Rotation of Reservoirs - Rotation of metal refinery