Clouds online - no sync (?)

[sniperton]

Originally Posted by dimlee

Yet until people who can see in the game code will explain, we probably can just theorise.

Sure, my explanation is pure maths. I don't even know whether clouds are static or move with the wind, and whether they are generated at once on mission start, or dynamically in run-time.

[Tolwyn]

Um.
They are static objects tied to the map. (Did no one read my post above?)

Originally Posted by sniperton

Sure, my explanation is pure maths. I don't even know whether clouds are static or move with the wind, and whether they are generated at once on mission start, or dynamically in run-time.

[sniperton]

I did. It seems clear that clouds are generated on a "one cloud / one tile" basis according to the CloudType (density) and CloudHeight properties specified by the mission designer. What is not clear is how and when they are created and why they are not synchronized online. TypeClouds only affects their visual appearance, not their density or exact position.

[idefix44]

Hi,

Here are some screenshots (10).
Me (idefix44) and my mate (crapouillot) played online several missions via HyperLobby.
The host is the stock Dedicated Server 4.13.3m.
The clients are the stock IL2 1946 4.13.3m.
We used TeamSpeak to synchronise the screenshots and the result is that we have the same clouds...

[sniperton]

1. Someone who've seen the code said there's no synchronization between clients.
2. Your evidence seems to support the opposite.
3. I don't fly online, just tried to explain what's on.
4. I'll watch with attention whatever this discussion will evolve to.

[sniperton]

Hi Alpha, just to clarify, 'cause there's a slight misunderstanding here I guess.
Dimlee's question related to the exact position of the individual clouds randomly generated/created by the client application according to the mission settings transferred from the server. Only the cloud density/type ('clear', 'hazy', 'poor', etc.) and the cloud height are specified there in the mis file, not the exact spots where clouds should appear on the map (at the specified altitude). There's no such thing as a 'cloud map' synchronized between server and client, correct? Player A and Player B should see the same amount of clouds nearby, but dispersed in a different way. Online players should be clear about that hiding in clouds may not work as it does offline. I simply tried to speculate how much chance they have when they still attempt to do so

[dimlee]

Originally Posted by sniperton

Dimlee's question related to the exact position of the individual clouds randomly generated/created by the client application according to the mission settings transferred from the server.

Right. This is exactly what I try to find out.

[dimlee]

Originally Posted by idefix44

Hi,

Here are some screenshots (10).
Me (idefix44) and my mate (crapouillot) played online several missions via HyperLobby.
The host is the stock Dedicated Server 4.13.3m.
The clients are the stock IL2 1946 4.13.3m.
We used TeamSpeak to synchronise the screenshots and the result is that we have the same clouds...

Thanks. It's hard to judge based on screenshots only but I assume that you made some observations in flight as well.

[idefix44]

Look the 2 grab0001s.
It can't be more clear, I think...

[dimlee]

I have found at least 3 tracks of online encounters where I was chased by enemy fighter(s) in the clouds. Human pilots, not AI.

https://www.mediafire.com/?nhz92xu9svj3edy
(please note readme file)

When I watch these track records I see that in majority of cases the enemy fighter does not follow my aircraft in the cloud. He either climbs or go around, sometimes below. And when he does follow me, he does not shoot. Yet Bf-110 gunner can spot the enemy in any visibility - guter soldat, ja!
Sometimes, as it's seen in SBD vs Bf109 encounter the enemy fighter does not open fire when both of us are in clear blue skies.

My initial conclusions:
1. Clouds sync online DOES exist in some form.
2. Sync is not perfect, probably. "My" clouds and "his" clouds do not match meter to meter. I assume there is some margin error.

Three tracks do not make the representative sample, of course. But at least I know that my "hide and seek" tactic was not pointless

In my opinion, Sniperton's Theory is credible.