Weather?

[frederf 0]

Maddmatt, you are basing your opinons on some poorly-made missions. Of course if the server controls AI that are in clear weather they will see clearly but players in localized weather won't. If the weather exists and is syncronized accross all clients and the server then everyone is on an equal playing field.

Of course the weather has to be "global" in a locality sense to avoid problems but it doesn't in the geographic sense. AIs can be selectively handicapped by geographically localized weather effects.

One of the things that I always wanted is for native weather pattern support built into the map files for seasonal and day night weather cycles. If the Island is in the great britian area and it's February mission date it should default to seasonally appropriate (Probabilistically) weather. Fog might happen in the morning and return again at night for example.

As for localized weather mattering at the transition, first much more realistic and gradual transition zones would make a lot of the "abruptness" (again caused by simple-minded mission makers) problems go away. However there might be a need for weather to be a factor even if you're not currently in it. AI on a clear hilltop might have their vision obscured of targets in a fog-shrouded valley floor.

[Maddmatt 1]

...but it doesn't in the geographic sense. ...

Yes it does :confused: How do you achieve different weather in different locations then?

AIs can be selectively handicapped by geographically localized weather effects.

How? Never seen that done.

Or do you mean you are suggesting that for ArmA 2? I don't think that is possible in ArmA 1.

Edited May 19, 2009 by Maddmatt

[whisper 0]

Not possible in ArmA1.

1 weather setting on the server, meaning all AI groups will have the same weather setting wherever they are, if it's clear weather, AI groups inside the foggy area have advantage over players in that area, if it's foggy weather, AI groups ouside foggy area are in disadvantage compared to players

[frederf 0]

Yeah I mean it's theoretically possible in A2. I think you could setSkill dynamically but that wouldn't really be the same thing.

[CarlGustaffa 4]

Ok, I get it now. So localized weather is a rather bad idea due to AI cannot have local weather. I will not touch setSkill array for this either, as I feel it belongs to other stuff and could possibly conflict those.

Having an interrest in general aviation, weather reports and forecasts are very important. I this these syntaxes pretty well. How does a military weather report and forecast look like? Like aviation TAF/METAR. I'm currently using an 'aviation influenced' textual approach so that everyone should be able to understand it. If military style include lots of abbrevations and looks incomprehensible to the common gamer, then a more 'folkish' textual version should also be shown.

ECS had a nice idea for weather and its execution. But unfortunately it doesn't allow any typr of user control and it became undesireable for me to use. If I as a mission designer wants to control weather for a paticular mission, the probability based year cycle weather is not possible.

So, how would a perfect weather system/script look/act like?

[frederf 0]

Military weather is differently stated at different levels. You might want to look at an OPORD template for a mission-focused example of a weather informative. METARs are fun to read to me but it's likely not that realistic nor accessible practically. (It would be cool to be able to tune in a weather station if you were in an aircraft).

Weather handling becomes a lot less important when weather change time, change severity, and localization are all relaxed to both more real and more manageable levels. If it takes several hours for weather to develop from one mode into another and you don't run through 17 different seasons while driving for 10 minutes on a road then weather prediction doesn't have to be that detailed.

ArmA2 should have a more homogeneous terrain so we won't be trying to simulate dust storms on a desert while having heavy rain in a northern forest going on simultaneously.

My ideal weather script would probably handle weather globally that was function-based for established trends that dealt with both the day/night cycle as well as the 12-month cycle. Bell curve probability distributions that traveled on cosine functions could handle the daily "flavors" well.

Say 0.0 is clear weather and 1.0 is the worst weather. Say perhaps the mission is April 12th. The function could determine that today is somwhere in the range (randomly) from 0.30 weather standard deviation +/-0.15*. That means 68% of the time the weather will be between 0.15 and 0.45. It could be outside of this but the chances are much less.

*Note the variability of the weather can also change over the year where some times of year are more unpredictable as others.

The number you determine will be the weather for noon April 12th. You do this again for April 11 and April 13 and then simply transition the weather over the course of the day gradually from the 11th/12th average at midnight to the 12th value at noon to the 12th/13th average at the second midnight. This should work for missions that are up to a day long. You could get more clever to have "week long" or longer randomly generated trends.

The day/night cycle should primarily deal with fog I would think. Given the day of the year, what weather was calculated, you can have a generated maximum fog level and have that come at go around about morning. This should work any day.

If you wanted to be more clever you could generate random temperature values for high/low/average and actually calculate temperature lapse rates to see if there'd be fog or not (and when) or how the weather might look. A hot day that had a lot of weather might allow a thunderstorm to be unlocked.

[S!fkaIaC 10]

Military weather

Military weather forecast:

"In the morning some random bullets that could turn in the late afternoon into a serious metal storm.." :D

Seriously:

In ArmA 1 there are some missions with local fog and rain areas shown at map by markers.

Hence I assume we can have "local weather".

Still my Q: will it rain also inside house as in ArmA 1?

[Landstalker 10]

S!fkaIaC

someones awnsered that question and NO is the awnser it doesnt rain in building anymore or supposidly under big trees which to me would give an awsome feeling when its raining hard and me and a squad or me on my own take to a big tree for time out.

[Raficoo 0]

look at the clouds in this video

http://www.youtube.com/watch?v=TVWe05sPXPM&feature=channel

3D volumetric if you ask me :/

the beginning of the video reminds me of Crysis lol

[Alex72 1]

Yep awesome clouds and confirmed no more raining inside, under shelters like big trees etc. :)

[S!fkaIaC 10]

Thanks Landstalker and Alex. Missed it.

[aus_twisted 0]

Military Weather

[frederf 0]

In ArmA 1 there are some missions with local fog and rain areas shown at map by markers. Hence I assume we can have "local weather".

This isn't exactly true. There was "locally triggered global weather" which is different from "local weather." It was done by changing a client's global weather to a certain value after stepping into a trigger. Any AI under the client's locality would be affected by this weather change even if they were outside the weather. This is especially bad for AI under the server's locality (most AI, esp. enemy) which all have to be in the same weather.

Those missions were really goofy where a rainstorm would only be in a 500m radius and you can't see it until you're in it. Such severe microclimates are more destructive to believability than they are worthwhile.

[CarlGustaffa 4]

I agree, but that's the only thing we've got. With 3D weather in Arma2 I think this opens up a lot of new possibilities.

When I joined Domination earlier and I was running ECS, the ECS weather would overwrite the mission weather. ECS weather was said to be quite realistic with seasonal changes etc. However, this mission continously starts at the same day, and the weather would be pretty much the same every time -- quite bad. So bad it became annoying to play the mission.

So unless there is a quick way to set the 'weather date', controlled or randomly, the seasonal weather really doesn't work well for most missions that take place at a certain date.

Perfect weather for me would be a system where every time you start a mission, a random set of behaviours are generated.

LOL @AUS_Twisted :D

Edited May 22, 2009 by CarlGustaffa

[S!fkaIaC 10]

Any AI under the client's locality would be affected by this weather change even if they were outside the weather.

Arrg! Did not know that. So simply wrong.

[frederf 0]

That's what I was getting at with my seasonal-but-variable weather is that you can have weather that takes a hint from the date but still has wiggle room to be somewhat different every time you play. A proper mission should have a "Pick a Date" option on the mission assignment screen.

Localized weather is not an issue if you make all of the changes gradual in both time and space. It's really the sharp edges and the overall difference between mix and max values over the used battlespace that make the Evolution-style weather so stupid. There are weather fade times used in the scripting language and they really should be 10+ minutes for any change but too many people have 10 second change times.