The rebuilding mechanism:
Many of us know that placing a structure over a wreck will cost you 50% less time and money. What is annoying however is trying to place the ghost structure over the damn wreck. Can't we have something that "snaps" the structure over the wreck, just like mexes "snap" to empty mex spots?|||Just be glad the build preview turns a different color now. It didn't in supcom 1 (and I think early builds of supcom 2) so it was really anybody's guess as to whether you'd get that 50% back or not.|||But I do remember CT saying in a preview that buildings would snap to wrecked versions of themselves. It apparently didn't make it to the final build. I think that feature would be annoying since it would be hard to place the building anywhere near the wreckage as it would just snap to the wreck's location.|||Well if there was an option like "enable auto-snapping to wrecks" then it would make everyone happy.|||would be nice, but not absolutely nescessary. Another thing that would be nice is a replacement structure actually costing less mass instead of paying full and then getting refunded..|||Agreed. The lack of that features really gimps the rebuild bonus in my view.|||This would be a cool feature, but there should be a button to either enable or 'break' so it defaults to one of the other and you hold say the 'Alt' key to reverse it.
As far as it only costing how much it should instead of refunding, the problem there is that wrecks take damage. Less of a bonus is given if the wreck is partially reclaimed or takes damage, including any overkill. The economic system doesn't account for the wreck's refund amount until you're actually trying to build, instead of checking every frame to see if you have enough resources yet, which sounds way less intensive CPU wise (and the game has to run on XBoX360).|||An earlier great idea by someone, is that you can just go to build mode and click on the wreckage and it will rebuild it. You dont even need to work out what kind of building it was or try to place it properly, you just click on it :)
Hollowpoint, dont try to make up some sort of coding speculation when you obviously arent familiar with it. Numerical checks are negligible (just consider everything else thats going on in a frame). And its not at all certain that it needs to poll per tick (it wouldnt even be per frame) to get it done.|||You know what you're right im not familiar with the engine, but I am familiar with programming logic in general and for the most part it is more optimized to be:
every tick:
1) check for enough resources to build cheapest unit in queue
2) if you have enough: check to see if building on wreck and give discount based on the amount of damage taken by the wreck. if not: wait for next tick
(so far looked up 3 things, amount of resources, the cheapest building in cue, and the amount of wreck damage taken if it exists. 3 lookups, 1 list sort and 2 if statements, *no matter how many buildings are in the queue*)
OR
every tick:
1) check to see if any buildings in the queue (with an engy there waiting to build since wrecks evaporate after a while) are being built on wrecks
2) of these, lookup each wreck's damage in order to decide which would be the cheapest to rebuild on.
3) Compare this to the cost of all other buildings in the queue to see which building is cheapest overall
4)If I have enough resources to build: build. If not: wait until next tick
Now you have 1 lookup for each building in the queue that has an engy there to see if a wreck exisits, if a wreck exists another lookup on the wreck to see how much health it has, make a new build list based on new cost info. Compare costs to choose cheapest building. Lookup resources to see if we have enough. Decide wheter to build
(one data lookup for EACH building in queue to see if posistioned on a wreck, one lookup for each successful hit as part of said if statement (if its on a wreck, see how much damage it had) [already less effifient program by this time], building and sorting of a list (for new costs based on whether it is on a wreck or not), lookup the amount of resources now, and your final if statement)
Choose an algorhythm that executes with the same effeciency every time, or choose one that gets *exponentially* more complicated based on how many pieces are involved? A performance minded person would choose the former.|||hotho11owpoint|||hotho11owpoint|||AdmiralZeech is definitely right there. The overhead of rendering a unit or modelling the physics or using dynamic scripts is miles ahead of this.
A performance minded person wouldn't attempt to optimize a code path without seeing extensive profiling data on the engine and effectively proving that it is a substantial contributor to the overall performance.|||The problem I have with the mechanic, is that it waits until you have 200 mass before it starts to build, and doesnt let me rebuild my mex with 140 mass in the bank/whatever I need to rebuild it. You dont need to no how the programming works exactly to see the problem there.|||I'm all cool with the whole thing that you have to wait till you have the full cash for it (not a HUGE problem) but aligning the damn thing is just too hard, and if you accidentally click just a milimeter away you lose any rebuilding bonus and the wreck dissappears as well.|||Nephylim|||Cygmus|||To the OP: Rotate the camera a little.. vertically; as much as you can.
That helps, believe me. ( for now ).
No comments:
Post a Comment