More Variable Weather??!!

Subtype Zero · 01-20-08, 01:10 PM

Blacklight--

I have forwarded your observations to Stiebler and am awaiting his comments. Have you been able to confirm anything via testing yet? However, I suspect it will be very difficult to confirm anything because so many variables and random variation are involved. Good luck, though!

difool2 · 01-22-08, 10:31 PM

Well I think I just found a very big piece of the puzzle.

Subtype Zero: "Theoretically, if you make sure to save no later than every 32 hours, your weather pattern would never change from winds of 5 m/sec with partly cloudy skies, no rain, and light fog."

You are running WeatherRndInterval at 2, correct? Well, I run it at 1, and I just had a weather change after an in-patrol reload after...yeah you guessed it 16 hours! I also remember earlier in the patrol that it changed after about that time on a couple of occasions, but never any quicker than that (despite what the editor might say-that may have been typed in by some ignorant copy writer for all we know). So subject to further tests, it would appear that you may get changes after every:

WeatherRndInterval x 16 hours.

The chance of it actually changing at that time still needs more data to pin down, but now that I know the actual interval determining that after a sufficient number of trials should be a piece of cake. I can also see if I can catch a change "in the act", so to speak.

The default game has that variable at 5, so if you are running vanilla you will never see any changes if you are reloading after every three days (80 hours). No wonder the weather seems to get "stuck" for what seems like forever! I wonder what would happen if you used a non-integer for that value, like .1-changes every 2-4 hours? In any event the rates of change I do see now seem to be about spot-on: a storm for a few days, then clearing and/or low winds for awhile, then another storm. I do wish that "partly cloudy" encompassed a whole range of cloud types, from drifting scuds to puffy cumulus to an altostratus "mackeral" sky, and in fact partly cloudy seems to be the least common of the 3 kinds of cloud cover (clear and overcast being the other two of course).

I will continue my patrol some more and check and see if all of this holds.

difool2 · 01-23-08, 02:09 PM

Quick followup: apparently there is a little randomization occurring with the change window-i.e. it isn't always going to be 16 hours on the dot, but 16-18 hours I've found. Or perhaps 17 hours +/- 5% * 17 (or +/- an hour each way).' I also caught the beginning of a change-rain slacked off, then stopped, visibility improved from socked in to overcast, then about a hour later winds died down and it became partly cloudy.

01-20-08, 01:10 PM	#1
Subtype Zero Lieutenant Join Date: Mar 2007 Location: 19,998 Leagues under the sea Posts: 262 Downloads: 7 Uploads: 0	Blacklight-- I have forwarded your observations to Stiebler and am awaiting his comments. Have you been able to confirm anything via testing yet? However, I suspect it will be very difficult to confirm anything because so many variables and random variation are involved. Good luck, though! __________________

01-22-08, 10:31 PM	#2
difool2 Commander Join Date: Apr 2005 Posts: 459 Downloads: 31 Uploads: 0	Well I think I just found a very big piece of the puzzle. Subtype Zero: "Theoretically, if you make sure to save no later than every 32 hours, your weather pattern would never change from winds of 5 m/sec with partly cloudy skies, no rain, and light fog." You are running WeatherRndInterval at 2, correct? Well, I run it at 1, and I just had a weather change after an in-patrol reload after...yeah you guessed it 16 hours! I also remember earlier in the patrol that it changed after about that time on a couple of occasions, but never any quicker than that (despite what the editor might say-that may have been typed in by some ignorant copy writer for all we know). So subject to further tests, it would appear that you may get changes after every: WeatherRndInterval x 16 hours. The chance of it actually changing at that time still needs more data to pin down, but now that I know the actual interval determining that after a sufficient number of trials should be a piece of cake. I can also see if I can catch a change "in the act", so to speak. The default game has that variable at 5, so if you are running vanilla you will never see any changes if you are reloading after every three days (80 hours). No wonder the weather seems to get "stuck" for what seems like forever! I wonder what would happen if you used a non-integer for that value, like .1-changes every 2-4 hours? In any event the rates of change I do see now seem to be about spot-on: a storm for a few days, then clearing and/or low winds for awhile, then another storm. I do wish that "partly cloudy" encompassed a whole range of cloud types, from drifting scuds to puffy cumulus to an altostratus "mackeral" sky, and in fact partly cloudy seems to be the least common of the 3 kinds of cloud cover (clear and overcast being the other two of course). I will continue my patrol some more and check and see if all of this holds.

01-23-08, 02:09 PM	#3
difool2 Commander Join Date: Apr 2005 Posts: 459 Downloads: 31 Uploads: 0	Quick followup: apparently there is a little randomization occurring with the change window-i.e. it isn't always going to be 16 hours on the dot, but 16-18 hours I've found. Or perhaps 17 hours +/- 5% * 17 (or +/- an hour each way).' I also caught the beginning of a change-rain slacked off, then stopped, visibility improved from socked in to overcast, then about a hour later winds died down and it became partly cloudy.