Optifleet - automatic fleet contents optimizer

    • Optifleet - automatic fleet contents optimizer

      Optifleet



      P.S. I am new here, thanks to Francolino for suggesting me this board

    • Hello anilatx,

      Your idea is fine, that's not the problem.
      But it's tooo much only useable for other tool developers because it's not comfortable. Most important is that they have to copy / type each number.

      When you want to offer your project for most of all player, i suggest once again that you create a webpage. Then you can extend it with comfortable features like getting all research levels, ships etc with an userscript.

      Regards, Francolino

    • Thanks for the comment. I am willing to add features making the tool more comfortable.
      Hosting public webpage is impossible becouse of amount of CPU-time consumed, but I can try to run some local miniwebserver to enable browser integration?
      Any hints which webserver to choose to get it the light and easiest to prepare/use?
      Where should I look for player info extraction? Contact developer of some other tool like galaxytool/antigame/... ?
      I have experience in direct use of galaxytool (SQL-querying it's database) and it'll be the easiest, but becouse of need for galaxytool it will cut down the number of users?

      The post was edited 1 time, last by anilatx ().

    • Hello,

      There are several projects which needs at first CPU load to calculate their job - WebSim, DragoSim, Osimulate, Ocalc, ..... or new, upcoming projects like antigame.de/calculator/

      It should not a problem to limit the among of ships and loops. Adding a local webserver is also problematic cause users have to install it. You are thinking much to much as developer, best is when you use the normal player "point of view" - Playing ogame is all what they want, they are not interested to spend time to setup tools etc.

      There are about 17k Galaxytool users, i'm sure this won't count down the number of possible users for your project. You can contact me (Antigame Origin) or Omar Hawk (Galaxytool) in IRC channel #ogame.origin and ofc you can join the Origin team to get teaminternal support. (contact benneb or even me)

      Regards,

    • I've released version 1.1, with performance improved and posibility to optimize defense.

      I experience strange problem: as an simulation engine I use GNU Ogame simulator(homepage no longer exists, but source is available for example in Debian Linux repositories), called also ogamesim or csim. In almost all cases it gives exactly same results as SpeedSim, but I've found a case where they differ(after 1000 simulations, so it's not an coincidence):

      Source Code

      1. [Setting]
      2. attacker
      3. Deathstar = 1
      4. defender
      5. Destroyer = 24
      6. Rocket Launcher = 700
      7. Light Laser = 250
      8. Heavy Laser = 256
      9. ==============================
      10. [Average result in Ogamesim]
      11. Attacker
      12. Deathstar: 0.10
      13. Defender
      14. Destroyer: 18.68
      15. Rocket Launcher: 546.69
      16. Light Laser: 195.27
      17. Heavy Laser: 199.88
      18. ================================
      19. [Average result in SpeedSim/QtSim]
      20. Attacker
      21. Deathstar: 1.0
      22. Defender
      23. Destroyer: 12.94
      24. Rocket Launcher: 379.96
      25. Light Laser: 135.96
      26. Heavy Laser: 138.92
      Display All


      Do You have an idea what could be the reason?
      I could port my application to use SpeedKernel of SpeedSim, but it would be lots of work, and it's about 10x slower...

    • Seems as if OGameSim does not consider the 1% rule, which means, that if the damage of the shield is less than 1% with one shot, the shot is not counted at all. Therefore the shields of a death star can only be destroyed by Gauss Cannons, Plasma Turrets, Battleships, Bombers, Destroyers and Battle Cruisers. But if the shields of the death stars are gone, the hull can be damaged by these units as well.

      So, OGameSim will give you incorrect results in this case.

    • Thank's for the hint. Knowing where to look for I've debugged a bit and found that "reflection rule" is honored, but the order of shooting is fixed:
      Deathstar
      Destroyers
      Rocket Launchers
      Light Lasers
      Heavy Lasers

      So Destroyers always reduce shields before defense hits. It is incorrect I assume? Order should be random? In battle_algorithm.doc of speedsim it's not clearly stated

      Source Code

      1. 'Deathstar' attack (power 200000) to
      2. 'Rocket Launcher' (shield 20/20, armory 200/200), killed!
      3. 'Deathstar' attack (power 200000) to
      4. 'Heavy Laser' (shield 100/100, armory 800/800), killed!
      5. [...]
      6. 'Deathstar' attack (power 200000) to
      7. 'Destroyer' (shield 500/500, armory 11000/11000), killed!
      8. 'Destroyer' attack (power 2000) to
      9. 'Deathstar' (shield 50000/50000, armory 900000/900000)
      10. 'Destroyer' attack (power 2000) to
      11. 'Deathstar' (shield 48000/50000, armory 900000/900000)
      12. [...]
      13. Destroyer' attack (power 2000) to
      14. 'Deathstar' (shield 4000/50000, armory 900000/900000)
      15. 'Rocket Launcher' attack (power 80) to
      16. 'Deathstar' (shield 2000/50000, armory 900000/900000)
      17. 'Rocket Launcher' attack (power 80) to
      18. 'Deathstar' (shield 1920/50000, armory 900000/900000)
      19. [...]
      Display All

    • No, the order is correct. But the 24 destroyers can't destroy the shields of a death star with the same techs: the complete firepower of the destroyers will be absorbed by the shields of the death star. And the shots of the defense units are not counted due to the 1%-rule...

    • No, each shot of each unit is counted on its own. So first, the 24 destroyers shoot with their firepower. The whole firepower of the destroyers are removed from the shields of the death star. But the firepower is not enough to destroy the shields.

      Then, each rocket launcher shoots with 80, and each shot is below 1%, so the shots are not counted, same with the light laser and heavy laser... Therefore the shields stay complete in each round and the death star will always stay alive.

    • Could someone explain in detail what's wrong with numbers I've quoted above?

      Source Code

      1. Destroyer' attack (power 2000) to
      2. 'Deathstar' (shield 4000/50000, armory 900000/900000)
      3. 'Rocket Launcher' attack (power 80) to
      4. 'Deathstar' (shield 2000/50000, armory 900000/900000)

      Error is here? Shield is restored between unit type if it's not 0? Am I right that RL(80 power) could damage once it's under 8000 shield?

      I've tried to read it in sources of SpeedSim, but they're not really nice to read, and probably it could be explained in few sentences as You understand it...

    • anilatx wrote:

      Error is here? Shield is restored between unit type if it's not 0? Am I right that RL(80 power) could damage once it's under 8000 shield?


      No, as long as the shield is still there (so > 0) the shots of the RL will have no effect. The shields are restored each round, not after each defense unit.

      anilatx wrote:

      Source Code

      1. 'Rocket Launcher' attack (power 80) to
      2. 'Deathstar' (shield 2000/50000, armory 900000/900000)
      3. 'Rocket Launcher' attack (power 80) to
      4. 'Deathstar' (shield 1920/50000, armory 900000/900000)


      This is the incorrect calculation. The rocket launcher reduces the shield power though it cannot.

    • Bug in engine fixed, patch commited to debian and official ogamesim now has results equal as Speedsim.

      I have released version 1.2 with names translations (using Speedsim .ini files, but they must be in utf8 format, I provide such versions for ones I've found).
      Display Spoiler


      Apart from reports parsing(in development) functionality is rather complete, now I'll concentrate on GUI improvements, becouse I understand that now it's not very ergonomic...

    • Xunfop wrote:

      The application window height is too high, you should make it smaller, my 14" laptop even can't push calc button.

      When it was smaller some users complained about scrolling two much while they have lots of unused display space...
      That's why I created _smallscreen version (same directory).
      This question repeats on many forums, I'll add it as a header to README