quote:Originally posted by Harley Nuss (teamEIS)

Colos die off in TWGS.


I did not know that. Would you share the circumstances please.

i swear i dun seen whut i seen. but i prolly dun seen long ago and wuz retarded.

and colos die. make a planet, set hazard to 255 and hability to 0.
i bet ya colos will die

I'm not really sure *when* they die, but they do. With stock planets, the average over time is positive growth.

Back in the old days if you filled up org on the H it significally cut down on the number of deaths

OK I did some testing and here's the results:

Control Group:
5 Planets 50M colos in fuel
4 die-offs in 10 days.

Split Population Test:
5 Planets 50M colos in fuel, 12.5M colos in equ
5 die-offs in 10 days.

Organics Test:
5 Planets 50M colos in fuel and 5K units of org on planet
5 die-offs in 10 days.

Equipment Test:
5 Planets 50M colos in fuel and 5K units of equ on planet
5 die-offs in 10 days.

Summary:
There was no significant difference in the number of die-offs for all test groups. The overall chance of a die-off was pretty close to 10%. The die-offs occur when the date changes.

Thank you for the tests, Mantis.

quote:Originally posted by Mantis

OK I did some testing and here's the results:

Control Group:
5 Planets 50M colos in fuel
4 die-offs in 10 days.

Split Population Test:
5 Planets 50M colos in fuel, 12.5M colos in equ
5 die-offs in 10 days.

Organics Test:
5 Planets 50M colos in fuel and 5K units of org on planet
5 die-offs in 10 days.

Equipment Test:
5 Planets 50M colos in fuel and 5K units of equ on planet
5 die-offs in 10 days.

Summary:
There was no significant difference in the number of die-offs for all test groups. The overall chance of a die-off was pretty close to 10%. The die-offs occur when the date changes.


Mantis, thanks for introducing this topic.
Based on your earlier suggestion re system date v running externs, I rechecked the ten (pers) 50M colo H's that I made yesterday, 25 hours after they had been made. I checked them using Tedit, did not log in as a player, scan/land etc.

3/10 planets lost 6% of colos, 7/10 gained 1.25%. I also noted the ore on hand on each planet. BUT here is the peculiarity:

I then logged as a player and did a pers planet scan:

864 . Class H, Volcanic No Citadel
--- (47M) 47T 0 0 53T 0 0 1T 0
864 . Class H, Volcanic No Citadel
--- (51M) 49T 0 0 53T 0 0 1T 0
864 . Class H, Volcanic No Citadel
--- (51M) 49T 0 0 53T 0 0 1T 0
864 . Class H, Volcanic No Citadel
--- (51M) 49T 0 0 53T 0 0 1T 0
864 . Class H, Volcanic No Citadel
--- (51M) 49T 0 0 53T 0 0 1T 0
864 . Class H, Volcanic No Citadel
--- (47M) 47T 0 0 53T 0 0 1T 0
864 . Class H, Volcanic No Citadel
--- (47M) 47T 0 0 53T 0 0 1T 0
864 . Class H, Volcanic No Citadel
--- (51M) 49T 0 0 53T 0 0 1T 0
864 . Class H, Volcanic No Citadel
--- (51M) 49T 0 0 53T 0 0 1T 0
864 . Class H, Volcanic No Citadel
--- (51M) 49T 0 0 53T 0 0 1T 0


All the scan data agreed with tedit data. Note however, that all the planets have the same amount of ore. Colos could not have died local midnight. They were all started at 1600 approx yesterday and checked 25 hours later. Might be that Someguy's theory about "scanning, landing" or even tedit checking, causes the deaths.

Kav

It would appear that colo deaths occur when you check the planets. The amount of the death (or growth) is probably dependant upon how long ago it was last scanned. Planets update when scanned based on the last time stamp.

quote:Originally posted by Harley Nuss (teamEIS)

It would appear that colo deaths occur when you check the planets. The amount of the death (or growth) is probably dependant upon how long ago it was last scanned. Planets update when scanned based on the last time stamp.


That would make sense - for test run I used the personal planet scan to check the population levels rather than TEdit.

This might also help to explain the bug where rogue planets suddenly recharge with fuel. (I just got nailed by this bug again last night. <grr>) If there is no one to perform a planet scan the planets never get updated. The other action that appears to cause planets to update is when someone lands on any planet in the sector. I'm fairly certain that this causes an update - I had drained a rogue L5 H down to about 5k sector damage (5% sector blast) then after landing on my planet in the sector to reload on shields, I warped out and back in and took a full 16.7k blast and a podding.

prolly why rogue planets gain a buncha ore after you try to land
there ya go JP, there is your trigger. fix it
"landing/scanning" doesn't cause the deaths persay, but they probably are triggered after tern or something. the first scan/land of the day updates the planets. (so you can safely land/scan repeatedly w/o killing all your colonists)

quote:Originally posted by Slim Shady

prolly why rogue planets gain a buncha ore after you try to land
there ya go JP, there is your trigger. fix it
"landing/scanning" doesn't cause the deaths persay, but they probably are triggered after tern or something. the first scan/land of the day updates the planets. (so you can safely land/scan repeatedly w/o killing all your colonists)


Landing/scanning/checking using Tedit does trigger the deaths. If you look at the ore levels in my previous post, note that all the planets have 25 hours worth of ore production for 50M colos, including the 3 planets that were down 3M colos.

ya, same thing.
the first time the planet is checked, in game or otherwise.

quote:Originally posted by Kavanagh

Landing/scanning/checking using Tedit does trigger the deaths. If you look at the ore levels in my previous post, note that all the planets have 25 hours worth of ore production for 50M colos, including the 3 planets that were down 3M colos.

I would imagine that the deaths are not caused by landing/scanning/checking but that the colo populations are only updated when the planet is landed on/scanned/checked/whatever. It would be fairly easy to test though. Write a script to continuosly land, and let it run for a few hours, see what it does.

Heh, this sounds remarkably like Schroedinger's Cat... whether it is alive or dead is determined at the moment it is checked.

This reminds me of another thread where Harley explained the time stamp somewhat. It seems that the server doesn't constantly perform updates to things like resource or fig production, but rather calculates them each time the info is needed based on the previous quantities and time stamp. To be really accurate, it would need to perform a compounding calculation, which it doesn't appear to do.

For anyone who is especially industrious, I imagine that there are some anomalies that you might be able to uncover. The idea that the population decrease doesn't occur until you make the server display the population makes me wonder if aborting the display on landing might postpone the population change. And to take that a step further, if you could then change the population level without displaying the planet, thereby updating the time stamp, and eliminate the risk of losing colos. Any takers?

+EP+