oh yea btw i tried sv_doublejump 1 in your latest SVN build it still sticks me to the ground, unless i use sv_pogostick 1 then it works do you know why that is? you can claim that as often as you want I can't fix it because it never happens to me to me it is working perfectly whatever I try hmmm really weird if you manage to fix it in the code, fine, submit a patch then im sure morfar has same problems but before I see it when playing on my own system, I can't do anything about it no i was just wondering if you knew why but sv_doublejump IS a bug I am not surprised if it may fail that way sv_pogostick replaces perfectly working code with buggy code as used e.g. in Quake 2 *sv_doublejump, sorry ok it is intentionally buggy, actually its just it works with pogostick which i find strange :) to allow jumps it should not allow yea with pogostick i can still manually jump if i want to doublejump + pogostick is dangerously broken for me that actually helps me more because i gain a little bit more speed when i do it manually really? I often get double height jumps when jumping from normal floors ah both together just don't make any sense yes i had the same problem, perhaps i should explain I have model/hitbox scaled down I know why it happens but I don't know why it does not happen in Quake 2 to try to work on Q3 defrag maps WRONG expect major breakage then scale the maps up, not the players down but go ahead, has nothing to do with that ? the client does not know the hitbox size if you change it in QC code the client's prediction thinks player has another size and does weird stuff well i have physics working on defrag maps NOT q3 ones. for the most part I just mean what you do is useless if it totally breaks with cl_movement so do NOT do that hitbox change thing instead, change the maps (q3map2 -scale ...) but that may then also need reparametrizing the physics which is possible but some work ok the reason the model (and hitbox) is scaled down, is because i don't want to scale Q3 maps up, due to distribution. e.g. if you scale up your maps by 1.5, sv_maxspeed would have to be multiplied by 1.5 too to match it but hitbox downscaling does not work so forget it how do you mean it doesnt work? :o it totally fails with cl_movement and without cl_movement on a server is mad in what way does it fail with cl_movement can you give me like an in-game example it thinks player has another size you stand on a floor, and prediction thinks you're in air it would "shiver" all the time while you move worst if you are near a wall I set cl_movement to 0 and the shiver is gone that's no solution it's fine LOCALLY but bad on a server you aren't really going to release such a mess but if you think you must do it change "gameversion" so normal Nexuiz clients don't see that thing on the server list and patch the engine to use your hitbox it's somewhere in cl_input.c then distribute that patched engine together with your mod note that it will be incompatible with normal nexuiz servers because of the hitbox size tZork made me a progs.dat to scale the model and hitbox size sure but the engine does not know about it that is the problem that totally breaks cl_movement and turning it off is NOT a solution what about if i want it for offline play Not online play. then it's fine but then better make it not even SUPPORT online play e.g. by letting only one player join, or by changing gameversion anyway, you find the engine's hitbox sizes in cl_main.c VectorSet(cl.playerstandmins, -16, -16, -24); starting from that line before i go on about online play, I tried sv_pogostick 1 with sv_doublejump 1 with cl_movement 0 on one defrag map i got some very funny results, i went up these stairs and it sort of launched me in the air as you said before this combination is known broken and no way to fix it it is that way by design and no offence, but i know nothing about code, so i wouldn't begin to look through that but i know people who would so thanks for writing it anyway. ah, now I see why Q2 didn't have to "fix" it Q2 has no pogostick basically, these two options are incompatible one question, if sv_doublejump 1 worked (here anyway) without pogostick, what would that fix, if anything? ? sv_doublejump does work without pogostick, so? or rather, it does for me no idea why you are having a problem with it does for you, but that us very strange because im not the only one having problems with it it feels just as strange to me I tried to reproduce your problems for two minutes as one of you said it even happens when standing on the same spot and repeatedly jumping it never happened for reference, I tried it on basement basically the rule is if I can't reproduce it, I can't fix it sorry, but impossible brb ill give it a quick try on basement.. get someone who has the problem AND can fix the code ok basically how am I even SUPPOSED to fix a bug i can't see? yes i understand :) if I make a "fix", how am I supposed to check if it did anything? :P yes dont worry :) you knw you were saying before about the hitbox thing if you change it in QC code the client's prediction thinks player has another size and does weird stuff is there anyway to fix that? yes, by patching the engine to use your new size no other way exists the engine can't support both because the protocol does not network that information so would that fix this thing when i sometimes jump twice as high as you said? :P that, what? erm the twice-as-high-jump bug is fixed by not enabling both pogostick and doublejump AH these two are incompatible the only time i get that problem (with my hitbox reduced) is going up stairs on this defrag map it like flies me up way too high but with both off, it works as normal. the problem is i cant try with pogostick off...because as you know for me doublejump isnt working correctly :( however get someone who has the problem AND can fix the code i will perhaps look into this at some point and see if i can find someone, but im guessing thats not going to be easy.