Home Trending Latest threads New posts Developer diaries

Development Diary #31 - 3rd of June 2009

Jump to latest Follow Reply
  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Show only dev responses
ElQuapo said:
Perhaps have a check-box before starting the game :

( ) historic weather
( ) random weather

Oh well, perhaps for HOI4 :)
Click to expand...
I think all the people that want "historical weather" should take a step back and count. HoI3 got 15000 provinces (including seazones) and the time resoultion is 1 hour. So you need to find data on the weather in all these locations for 24x365x12= 105120 hours. That means a MS excel sheet of 15000 by 105120 should be enough to define historical weather.

If you are very fast and can find the weather for each province and punch it into the sheet in 10 seconds you are looking at 15'768'000'000 seconds of work or 4.4 million manhours. You better get started to have it ready for HoI7 because Im sure they neither will find it economical to do this research themself nor have computers that like adressing datasheets with 157680 million input points ready before then.


Great job on the dynamic and random system Devs, It looks awesome :)
 
So we went off and spent a couple of days becomes experts in Meteorology (feel the power of the Wikipedia!)...
Click to expand...

If my wife, a meterologist read that, she probably would NOT let me play HOI ever again! lol :eek:

This really looks like another great improvement. This idea may be a bit much, but hurricanes or typhoons capable of damaging ships would be interesting.

http://www.usatoday.com/weather/hurricane/history/typhoons-ww2-navy.htm

http://www.history.navy.mil/faqs/faq102-4.htm

But I realize that the effort to do this must be weighed against the value it would bring, and I'm not sure how much that would be.

It really is starting to look like HOI3 might raise the bar in more than one or two ways. wow... just wow!
 
Nobby said:
I'm quite staggered by this system. Very, very impressive - it goes far beyond what would be normally expected for strategy games. Well done, Paradox.

However, may I be the first to point out you appear to have invented a major Scottish river on the east coast? The Forth (the northern river of your two) is pretty accurate, but the river heading south out of the Firth of Forth didn't exist last time I checked.

And the Clyde, on which a considerable number of Royal Navy ships were built, is absent from the west coast. If we only get two Scottish rivers could we at least have two real ones? ;)
Click to expand...

True enough. If the western river emerged a bit further south, it could pass for the Tweed, but the Tweed isn't a significant obstacle up as far as Peebles the way it is on the map :)
 
As said before, very impressive, even more so because it brings an unexpected level of realism. I need much meteorology for work and I can tell you congratulations, you clearly did your research, and this system models well the weather patterns in mid-latitudes.

You have really already done the hardest part, mid-latitudes. Now, with much less effort, you could make realistic tropical and polar weather. A few tips you may want to consider:

1) keep the fronts out of the regions above 65 degrees or so; or with low probability to move there. You can pretty much stick high pressure there most of the year.

2) Also keep the fronts out of the inter-tropical region, below 30 degrees or so. There are transition latitudes, but you need not complicate your life with them.

3) Weather in the tropics is well modeled by seasonal patterns like in HoI2, if I remember its features, so it's simpler.

4) A neat feature for little coding cost are afternoon thunderstorms in the tropics. You can really set your watch to them in many areas, so they are easy to model. They could have an important limiting effect on bombardments and CAG missions there.

5) Finally, and probably low on the list, you can have hurricanes meander through the tropical oceans and islands, with low probability, between 10 and 30 degrees or so. Given your simulated data and the knowledge and technology of the time, they might as well move at random from east to west in the N. hemisphere, tilting north; then turn into low pressure systems. In the S. hemisphere, they don't behave as neatly, so that's a headache, and they are virtually not present in the S. Atlantic.
 
Are 1 and 2 explained mostly by pressure? In that case they probably have it down, as they rely on a mean pressure/temperature map. King didn't say anything explicitly about a pressure map, but I think it's a safe conclusion from his statements.
 
Alex_brunius said:
I think all the people that want "historical weather" should take a step back and count. HoI3 got 15000 provinces (including seazones) and the time resoultion is 1 hour. So you need to find data on the weather in all these locations for 24x365x12= 105120 hours. That means a MS excel sheet of 15000 by 105120 should be enough to define historical weather.

If you are very fast and can find the weather for each province and punch it into the sheet in 10 seconds you are looking at 15'768'000'000 seconds of work or 4.4 million manhours. You better get started to have it ready for HoI7 because Im sure they neither will find it economical to do this research themself nor have computers that like adressing datasheets with 157680 million input points ready before then.
Click to expand...

Scripts could probably read weather files and directly translate them to game data.

But again, it does not have to be that precise, it could be area based. Like, Eastern Europe had this weather in September 1942, North Africa had this weather in the summer of 1940 etc. etc.
 
Really nice thing. Could it be possible to add very little feature in regards to weather and other map views. There are few buttons above mini map to switch between different views like partisant, weather etc. I would like to be able to hoover over those buttons and general map would show me things im interested in, but only as long, as i have mouse over specific button (like tooltip works). Of course i could always click on that too to switch to that map for longer if that would be needed.

All i am asking is quick prewiev. Its not hard to do i suppose, but could be very handy, especially, if weather would be important at the front. Thx.
 
Meaning of the number of clouds?

I noticed looking at the three maps, that sometimes there is only one "cloud" per province/sea zone, and at other times there are multiple 'storm' clouds; and in one picture there are 5 'rain' "clouds" in a single province!

Is this meant to signify the 'intensity' of the cloud-cover/storm/rain/whatever? Or are there micro climates within a province?

Lastly, I noticed over the islands of Scotland, extra clouds...does the weather change over the smaller islands in an otherwise 'ocean' province?

Love the attention to detail!! :)

Solon
 
Solon said:
I noticed looking at the three maps, that sometimes there is only one "cloud" per province/sea zone, and at other times there are multiple 'storm' clouds; and in one picture there are 5 'rain' "clouds" in a single province!
Click to expand...
Nope, there is always one cloud per province or seazone. Just that at this zoom level the only borders you can makeout are the area borders.

So since each area contain roughly 5 provinces the areas get roughly 5 cloud each.
 
Alex_brunius said:
Nope, there is always one cloud per province or seazone. Just that at this zoom level the only borders you can makeout are the area borders.

So since each area contain roughly 5 provinces the areas get roughly 5 cloud each.
Click to expand...

spot on
 
Alex_brunius said:
I think all the people that want "historical weather" should take a step back and count. HoI3 got 15000 provinces (including seazones) and the time resoultion is 1 hour. So you need to find data on the weather in all these locations for 24x365x12= 105120 hours. That means a MS excel sheet of 15000 by 105120 should be enough to define historical weather.

If you are very fast and can find the weather for each province and punch it into the sheet in 10 seconds you are looking at 15'768'000'000 seconds of work or 4.4 million manhours. You better get started to have it ready for HoI7 because Im sure they neither will find it economical to do this research themself nor have computers that like adressing datasheets with 157680 million input points ready before then.


Great job on the dynamic and random system Devs, It looks awesome :)
Click to expand...


of course you could put it simple and tell the engine which date for winter to start and which date for snow to start melting.

Weather could be, theoretically, be fully analyzed scientifically, but that's never possible in real life anyway, they are affected by too many too complicated factors, you need a Solar system simulator to simulate the whole thing accurate enough (of course the whole earth needed to be simulated...). If you somehow made such simulator we would only need one set of data from the starting dates. Good luck on such projects. It might be till 2109 till we got such detailed games and machines capable of running that.

(actually i think the game made this way would be fun... :D with the whole world simulated...while you are playing the heads of the country.. but i doubt if i would live till such thing appeared. Maybe when HOI 100 comes we got it.)
 
henryjai said:
of course you could put it simple and tell the engine which date for winter to start and which date for snow to start melting.
Click to expand...
You still need to do that for 12years in 10000 provinces or 2000areas, (more if snow melts many times in a single year) and then not have historical results anyway. To simulate only snow. Im just saying that developing a dynamic random model gives much more back because of two things:

1. Its a much efficient way, you can get better results from less work.
2. It can be used over and over again in all other upcomming paradox games.


henryjai said:
(actually i think the game made this way would be fun... with the whole world simulated...while you are playing the heads of the country.. but i doubt if i would live till such thing appeared. Maybe when HOI 100 comes we got it.)
Click to expand...
We never know about that, according to estimates based on exponential growth in computers calculation power a normal PC by 2050 will have more capacity then the entire human spiecis brains have.
 
Jon Sutter said:
If my wife, a meterologist read that, she probably would NOT let me play HOI ever again! lol :eek:

This really looks like another great improvement. This idea may be a bit much, but hurricanes or typhoons capable of damaging ships would be interesting.

http://www.usatoday.com/weather/hurricane/history/typhoons-ww2-navy.htm

http://www.history.navy.mil/faqs/faq102-4.htm

But I realize that the effort to do this must be weighed against the value it would bring, and I'm not sure how much that would be.

It really is starting to look like HOI3 might raise the bar in more than one or two ways. wow... just wow!
Click to expand...

I agree ;) !
 
Really interresting reads about storms there Jon Sutter. Looks only CAGs and Destroyers are vunerable enough to be lost at least, while damage is possible to other ships.

They should all lose a fair deal of organisation though. Especially CAGs and Carriers, imagine trying to defend yourself after sailing out of a storm with no aircraft left on the deck and a mess with fires and damaged aircrafts in the hangar.

Im not sure if there are enough event triggers for this but it would be awesome to make a random event for it with say 1% trigger chance each night a fleet spends in a storm.
 
You know what this game needs at this point?

Animated pictures of your front commanders, who comment on the missions you give to them and how they think of it.

Von Bock when you order him to go Blitz towards the kremlin in December '41: "Zis is impossible, mein Führer, it is snowing from here all ze way to Moscow!"

Yamamoto during an offensive against the Solomon islands in '43: "Your highness, there is a typhoon moving into our area of operations. We should halt the offensive and wait for better weather."

Eisenhower when you give him the order to switch full offensive for the cross-Channel attack in June '44: "Alright, but let's hope this storm is over soon!"

Manstein, under full Blitz orders, fighting against long odds to break through to Bastogne in December '44: "This isn't going to work! We should break off the attack and fall back, before the snowstorm ends!"

Vandengrift when you order him to conduct the nuclear attack on Hiroshima in August '45: "Looking good, skies are clear and I expect no enemy fighters in the target zone!"

Kinda like the talking advisors in Civ2. :cool: I loved them. This kind of weather system absolutely needs some AI voice to add its comments.
 
You must log in or register to reply here.
Top Bottom
Back