Well I notice when spectating the crosshair often doesn't track the model accurately. It does look more noticeable than usual in that video though.

My rail aim went from outstanding to shit after the hitbox increase. Should be the opposite. I think they dont have the hitboxes centered or moved them over or some shit so i have to relearn what will hit now.

nope, you just suck

<3

>My rail aim went from outstanding to shit after the hitbox increase.
same, I dont have perfect aim but my rail is fucked up
i am also noticed that i got the hits if i shoot behind the model

big hitboxes + antilag, nothing weird in that demo.

i've read that the demos are bugged and you need to record live gameplay instead of demos :/

I've read way too many people too lazy to spend 3 minutes trying to understand the meanings of pov, serverside and antilag that blame this kind of vids on bugs.

This, really.
Considering the amount of nerdiness that goes on here regarding cfg's, mice, monitors and other technical minutiae that are just placebo a lot of the time, it's surprising that so many people here on ESR still don't understand what antilag is about (and, despite this, still generate massive amounts of whine on the subject)

Well I considered that, but when I was spec'ing this live it looked exactly the same, which is why I went back to view the demo. I've also had situations like this playing myself so I dunno... Meh

cg_truelightning pls

Imo, there are several reasons for this.

The "Anti-lag" someone was talking about earlier in this thread includes backwards reconciliation by the server; when it receives information from a players pc it checks back by a set amount of time into the current game's history to see if that user has hit anything since the last time it checked.

This results in the rail hitbeep being non-instantaneous (as your data has to be sent to the server, server checks it, sends back results to you). This is why when you, or the player your shooting is moving fast, things can seem very out of sync. A good example is when you moving around 600 to 700 ups, your opponent rails you (from his screen, just before you go behind a wall), and your computer registers the hit when you are already behind the wall. I screamed BULLSHIT many a time until I was informed of what it was.

The second reason is that it looks even worse in a demo, as it is not correctly synced to start with. There is an ancient thread about this exact matter, where someone posted an excellent video highlighting the differences between a live recording and a demo playback.

I know you havent done this here, but playing against bots is also worthless to analyse hitbox registration, as their position only get updated every 10th of a second or so (originally to save cpu resources when Q3 was released).

CPMA also had backwards reconciliation, but to a smaller degree than QL. This is why it was more noticeable in cpm, and looked awful in a cpm demo etc :)

So in short; this is a well known "problem", and any "evidence" will be disregarded unless it's a live recording, as it's known that demos are a useless medium for use in analysis of hitbox registration; even when it looks absurd from demo playback.

Hope this helps :)

dunno why people now speak of antilag for cpma, ( had never heard that until 2 weeks ago ), all I could find in the docs is massive interpolation/extrapolation.

I can very possibly be wrong though, the documents haven't been updated in ages and most of my experience is limited to either lan or 15ms

I remember an arqon post about it in a post near the start of QL's existence; he said that cpma does have it, but to a much smaller extent than QL.

cg_lagHax <-1|0> (default = -1)
A combination of adaptive prediction and an updated version of the
famous "50ms hack" we introduced way back in 99v6 that also does small
amounts of lag compensation. Capped at 100ms no matter what: this is
intended solely to make European / EastUS v WestUS / etc games a bit
less of a hassle, not to hack dialup players into aimgods at the expense
of everyone else. 0 disables it, -1 means "as much as I'm allowed": it's
naturally adaptive. You'll lose some of your "feel" for lag, which
messes up your RL aim, etc. This doesn't suffer from the CS/etc problems
of "total BS" shots that piss everyone off; it's not trying to be some
panacea for modemers; and I'm honest enough to call it the hack that it
is instead of pretending that it magically makes lag suddenly not exist,
but all in all it's a pretty nice end result. If you use this, any form
of nudging will generally make you LESS accurate if your ping's under
100ms, because it'll screw up the adaptive calculations.

this is clientside, nobody would be retarded enough to implement clientside antilag that could be hacked very easily to make every single hitscan register as a hit. Especially not arqon ( inb4 fanboyism accusation )

client decides it's a hit -> server is retarded
client sends hitscan -> server gets the order -> rolls back in time according to the ping -> WAS it a hit yes/no -> sends confirmation or denial.

laghax always sounded to me like extrapolation to kill the 30-50ms lag built-in the engine.

Stuff gets predicted by the client. If the alg is good enough and your ping isn't utter shit or unstable, models have a very decent to very good probability of actually being where your client displayed them/ to be there when the server gets any orderyou might send, hence you don't have to lead with hitscans anymore but it still fucks your rl aim cause you still have to compensate for the projectile speed/lag.

+added jerkyness. iirc using smoothclients on top of that makes the ennemy position slightly less reliable overall ( even though more prone to obvious errors like people warping into walls ) but those imperfections made it less jerky.

I don't have this problem that much, but if i do it happens only on ping 50'ish opponents with good movement. In my experience these conditions throw the hitbox and model out of of sync.

Only once it happened that often on a server that it was making me rage and give up. I rechecked the demo, and in the demo i saw exactly what was happening on my screen.

http://www.youtube.com/watch?v=ADKgiwazqOs

It happens in so few occasions.. i get more annoyed by all these warpers in various shapes and sizes, some even able to cause hardware lag on my client, and the fact that i have to play with a simulated +20 ms to be able to have a decent LG.

is that taken from demo?

same happens for me and I highly doubt it a config problem. Some sort of server bug or just game bug.
Annoying bug for sure.

Nice lg and apparently it works in your favor so.. :P

My lg aim is as shit as never before now days. After the hitbox update my lg acc went quite up, I'm not a big lg aimer, so I was pleased with some bigger numbers/better aiming. And now, since some time, my lg is even worse than before the update :s

I've tried changing sens/cross/gun_xyz/lg_style but nothing helps :)

Haven't noticed any big differences rl/rg tho.

Could this be related to cl_timenudge?

just aim to the position where the player was 0.2secs before. problem solved

The main reason is probably antilag, you can see he's aiming basically just as far behind the model the whole time. the spot he's aiming is where the model is stationed on his screen.

This is also enhanced if you have the lean thing on so keel dips a bit out of the hitbox when moving.

That is unless ql has suddenly gone to crap, can't log on atm and haven't played for a few days. This however is still the reason demos and anything not directly portraying the shooters screen doesn't count when discussing registry, you basically need to record it while it's happening from his pov (demos and spectating don't count).

I hit some ridiculous rails though I'm not even on the model. My buddy is experiencing the same and to be honest I don't like it at all. It is so fucking annoying when you want to dodge a rail and someone with slow reactions still hits you.
Bigger hitboxes doesn't improve the flawed netcode and lately I'm really pissed off by ID's "support".
I somehow get it now why so many people complain :x

Ah and I forgot rail hits through ledges. I mean you can even shot greens through them. Try it out on ztn ra and throw some down to mh.

I was hitting around 50-60 rail, and in the last 2-3 days my rail has been like 25-30 IF that. I've noticed a ton of problems with the match browser, and weird bits of lag that weren't there before either. Aren't ID moving some shit around or something? That might explain it, but it's not cool to have this many problems.

id are not moving the gameservers. You've just been missing rails, that should explain it.

yep experienced that too

This is a demo so it's useless for netcode analysis

its not a secret ql netcode sucks... maybe they will increase the hitbox (again) to a size no1 can miss -> less whine... profit!

nigga what?

I'm hitting fine, no problems.

Meanwhile, people have always complained since QL started. There is no difference between complaining now and before the hitbox changes. Or the previous changes.

Just wait until the next update and more people will believe that suddenly there have been some changes and that it's causing them to play bad.

Great that YOU are hitting fine.

Not great that meanwhile, people have always complained since QL started. Did people (as in even 1 person) have such complaints on q3 vq3?

No.

On q3 osp? -> No
On q3 ra3? -> No
On q3 cpm? -> Hell no
On q4 vq4? -> No
On q4 xbattle? -> No
On q4 q4max? -> No
On QL? -> Yes

Working code is out there, free to grab.

Actually they did complain.

q3 1.17 to the new patches, to the early versions of osp, to the later versions, to cpma, and now quakelive. The only time I wasn't really around to see the complaints was during quake 4, but I honestly don't believe that everyone playing it was happy, considering it only became playable after a year or so of work, and whining.

if you are specifically speaking about netcode, original quake netcode and OSP was awful, even on my 40 ping now I would dread playing on it.

CPMA netcode was fine, but fuckit, the xerp complaints was endless, thread after thread about the netcode.

Yes, the working code is there, go play if you want. I'll stick with quakelive.

Jeez i must have not been reading the community forums in those days then. Or wait.. i was.

A quick google brings up this regarding CPMA:

http://www.esreality.com/index.php?a=post&id=1518108
http://www.esreality.com/index.php?a=post&id=1280666
http://www.esreality.com/index.php?a=post&id=1887513
http://www.esreality.com/index.php?a=post&id=1521589
http://www.esreality.com/index.php?a=post&id=1529836
http://www.esreality.com/?a=post&id=1520770
http://www.esreality.com/index.php?a=post&id=1889449
http://www.esreality.com/index.php?a=post&id=1525458
http://www.esreality.com/index.php?a=post&id=1144625
http://www.esreality.com/index.php?a=post&id=1584312

It's a bit harder to find older stuff before the pre-CPMA period, but I very clearly remember people aways complaining about anything and everything.

I actually opened 2 of your links, because i have not played CPMA myself. The first post is about someone not liking CPMA, the second about someone not liking OSP.

I don't like duel. And i don't like chess either.

And i also don't like the ---> N E T C O D E <--- of QL.

I think maybe you misunderstood my first post. You are correct in saying that the netcode for Quakelive has always been like this. I.e. the same since it has been released. That was not my argument.

I am saying that there is no difference between how it is now, and how it was before the (latest) hitbox changes.

The rest of my argument that people imagine *new* problems where there are none. I believe that sometimes the netcode can force you to have to aim behind the player, as I have experienced this myself, but that was not what I was talking about.

At the same time, not relating to my previous argument, the current netcode is far far better than it was in OSP, unless maybe if you are playing with a sub 20 ping. (Interestingly though, I remember hitting "shadow rails" all the time in OSP, even on lan.) CPMA netcode was better, but it still sometimes felt weird and had anomalies, I think that the difference is negligible between CPMA and quakelive. Of course this is just my opinion and I can respect that you don't like the current netcode.

good ol' xerp thread

Would it be possible to have and FPS game where your client just draws the hitbox instead of an enemy model? Sure it would look ugly but it would be efficient, if it's possible.

Ultrono?

Wow hadn't heard of that. That game looks like pure awesome!

Well it was made by czm...

Doom 3 did draw the hitboxes.

All we need is uncapped cl_timeNudge and it'll be all good.

Seriously, fuck lagHax/xerp and all other bullshit that's supposed to fix things. All needed to be done to QL is remove all the crappy "laghaxes" and do some optimization improvements.

moral of the story: dont play ql

:D

its not about netcode. see how keel fit hitbox when is standing. always hitbox protrudes from the right side. id should change models or refit hotboxes

Be a man and stop blaming netcode for your suckage. There are certain cases when it doesn't work correctly, e.g. enemy is warping or MAYBE at pings < 10 (still no 100% proof for that though) but 99% of the netcode whine is just stupid. As someone already said, demos can't be used to analyze netcode. "Last few days" is stupid too, as there were no updates at all which could have influenced netcode. Just doesn't make sense.

Some settings to try:
cg_predictlocalrailshots 0 //1 is not as accurate
cg_railtrailtime 0 //don't get fooled by railbeam
cl_timenudge -20 //worth a try
cg_forceenemymodel keel/bright //tankjr isn't good for rail
cg_playerlean 0 //better representation of the hitbox

*facepalm*

As I said... The reason I went back to view the demo was because that's how it looked ingame. Sure the demo may 'enhance' it, but it still looked completely off 'live'

Just because there were no updates recently doesn't mean it wasn't slow and fucked

The connection between your pc and the gameserver is not influenced by the server which hosts the website and database.
facepalm indeed..

I wasn't talking about the database website, but okay

bump. awful game

cry more motherfucker

It's so bad that you tracked down this 2 year old thread

Come on noctis...

back in the day, there were netcode threads without Weird

Last few days of playing were really weird for me too, I've hit some rails that are just impossible to hit, even when they were drawn on the screen they couldn't have hit yet the sound registered a hit. The again rails that were right on the spot, that I've hit on 3/4 players on the opposing team somehow miss the forth player.

For instance I've hit a shot on hidden during warmup dropping from the bridge to red armor stairs flicking to the bridge on the enemy. When the rail was drawn it was visible that it hit the outer rim of the bridge, just bellow the enemy, but still, way bellow. It registered as a hit.

To be honest there is obviously a way to abuse how your connection interacts with the netcode, and how that combination interacts with your opponent. On some people I can't seem to miss, and some can't miss me, and some whatever I do it's almost like they are not where they are drawn on my screen. It comes down to playing correctly and good, rest is really just random stuff going in or out of your favor.

Say what you want about the netcode being fine but if someone has a ping around 50 (my connection being <35) my lg goes right through them. Its as if their model moves 150% instead of 100%. Their dodging speed seems way faster and my lg just misses their model even when it is sat upon the object.
Even when they are falling or going up a jump-pad my lg seems to drop as their model moves.
At first I assumed that it was just people dodging until terrible players and non-dodging (JP / falling) made me think otherwise. Now every game I play against someone with 50 ping they ghost my lg and I can't deal with it.

**Should add that if I have 45-55 ping it moves to around ~63-73 where this happens

FUCK

I have two demos where the rail is dead center of a model and does not register as a hit. The first demo is me playing on practise mode against a bot - http://4seasonsgaming.com/forums/download/file.php?id=1915 . Now i know the issue about demos and how they are not the best method to analyse these sorts of issues and also the fact that prac mode isn't the same as lan but it's literally dead center on the model. And the other demo is rapha vs cypher on lan - http://demos.quakecon.org/qc2009/skrapha(POV)-vs-SRSrazerCypher-ztntourney1-2009_08_13-18_54_56.dm_73 . It happenes at (0:10). The issue with this demo is i think they had some interpolation issue for hitbox models which weren't fixed back then? But even then they have recently fixed it so shouldn't it also apply to demos? I have viewed the rail shots in timescale .0125 which is the minimum limit and they are both dead center on the model.