they're basically the same but the extra .2 on the ratio keeps glitchy servers from impacting hitreg as much in exchange for like 3ms network latency

you can see the difference by watching the interp on net_graph 1 and seeing whether it's flashing yellow instead of having a fixed orange/white. if it's flashing yellow, that means hitreg will be affected in some situations on that server with the settings that the net graph is displaying

>Didn't people used to say not to use smooth? What are the downsides?

with a low cl_smoothtime the difference between cl_smooth 1 and 0 is nearly nothing

except on glitchy servers having a small amount of PoV smoothing is beneficial on high refresh monitors because of microstutter

cl_smooth doesn't affect latency or timing, it just smooths out forced viewpoint updates from the server. because of input prediction, getting knocked around (i.e. airblasting and rocketjumping) normally causes spontaneous jumps in your final point of view after it finishes running prediction for that frame, but cl_smooth makes that spontaneous jump take a period of time instead of being literally instant. it doesn't come into play anywhere else but bad lag or extremely tiny position adjustments due to precision/framerate differences between server and client
>Also, does 1.2 and 2.2 still not work on certain servers?

As long as servers can restrict cl_updaterate and cl_interp_ratio (which is still the case afaik) you have to specify both

cl_cmdrate 66
cl_updaterate 66
cl_interp 0.018181818 // or 0.033333333
cl_interp_ratio 1.2 // or 2.2

cl_predictweapons 1
cl_lagcompensation 1 // DO NOT EVER, EVER, ****EVER**** DISABLE THIS
cl_predict 1 // DO NOT EVER, EVER, ****EVER**** DISABLE THIS

cl_pred_optimize 1 // Slightly improve hitreg in certain situations at high framerates at expense of an extremely slight performance reduction (default is 2)

cl_smooth 1 // Whether to smooth out the location of the PoV when it's knocked around by the server: rocketjumping, airblasting, etc
cl_smoothtime 0.0075 // ~7ms PoV smoothing window, should be enough to prevent microstutter without affecting reflexes. Default is 0.1 (100ms)
rate 120000 // Prevent choke when it's not necessary. Lower this value below the default if you have extremely low bandwidth internet (e.g. 56k dialup: 56000)

glhf

yeah but the command could theoretically do unrelated things first if it's hooked up to anything at all despite that fact so that's what you should look for if people are claiming results

Kotaku is the least garbage? What?

bad servers restrict interp ratio so you have to set interp as well, but some servers restrict updaterate so you have to set interp ratio as well

updaterate 66:

cl_interp 0.0152
cl_interp_ratio 1

cl_interp 0.033
cl_interp_ratio 2.2

posting in a thread

the tf2 team has literally shared gg2 on the tf2 blog before

too bad the gg2 forum is dead and nobody has any motivation to contribute to the game anymore

tho people still play it for some reason

- a gg2 dev

Compare against QWTF instead of TFC maybe

Not saying something will happen at that time but imo it's a more fair interval considering QWTF and TFC being mostly similar

too sweet, same with AZ sweet tea and gold peak sweet tea

lemonade iced tea is fine though

thats coming from someone who has basically lived on soda for years

I don't get sleep paralysis but when I get colds sleeping feels like I have no lungs and cannot breathe, even though I can actually breathe mostly fine

Pixel skipping doesn't exist in FPSs. Skipping over models matters, but pixels are purely part of putting the gamestate on the screen.

Just don't raise your ingame sensitivity setting above a level that makes it hard to put the crosshair on top of still targets while you're strafing in the game you're playing. If you need to set it higher than that for comfort reasons, increase your DPI instead, or use (gasp and shock) mouse acceleration.

If 4.5 doesn't make it difficult to aim at models while you're strafing, and your mouse runs better at 800 dpi than 2400 dpi, then even that is fine, despite the "pixel skipping".

the part of last with the control point up to the pillar is ok

i get about 75% the frames of windows on arch and a phenom 2 x4 and a radeon hd 5770 with mesa

second CP needs to be not hard to defend against mid if you have uber, but hard to defend against mid if you don't. spires are an easy way to make it hard to defend without an uber without making defending with an uber harder

if you have more than three routes between last and second, it's going to be impossible to make sure nobody's hiding in them when you push out of last, so don't do that

elaborating on the previous statement, if you have a ton of corners in all the rooms between last and second, that will dramatically slow down people who are checking for lurkers and pushing out of last will be slow

there needs to be a demo rollout with at least a lunchbox on the way to mid, add a pill if there's serious risk of taking fall damage

mid needs to be less detailed than other areas of the map because it will have 12 players on it at once at the beginning of every single game, and people will rail all on your map with no compassion if they lose the map after taking framedrops at the midfight

valk's guide is pretty good for more concrete advice.