Explanation on the changes to wind mills

    Quote from Rick

    You are claiming windgens produce 30eu/tick now that is silly.


    I never claimed anything of the sort. Reading Comprehension FTW.


    Quote

    Measuring after its transformed is VERY INACCURATE (PLS READ THAT NOW). Even if only 1 solar was connected to it it will be possible to measure 500eu/tick they way you do it. Dont you understand that?


    I didn't need an "accurate" measurement. What I did was sufficient to substantiate that things had changed in 1.337, and changed back in 1.337b. Can't YOU understand that? If you can't, check the update notes.
    Lofty minds understand concepts, while petty minds nitpick the details. You can't see the forest for the trees.

    (Things like this make me check my address bar to make sure I'm not on the Minecraft Forum...)


    I've since thoroughly clarified what I meant by that statement, and you're the only one that seems to care more about the literal (mis)interpretation of it than of the corroboration of the changes.


    But if it's THAT important to you to be right, I won't deny it to you. It's obvious that you have a deep, unsatisfied need to be correct about something, and since this doesn't matter to anyone but you, you can have it.


    There, do you feel better now?

    Quote from HeadHunter67

    But if it's THAT important to you to be right, I won't deny it to you. It's obvious that you have a deep, unsatisfied need to be correct about something, and since this doesn't matter to anyone but you, you can have it.

    Trying to make me look like the wrong one eh? As i can remember you measured the eu/tick in a very wrong way and tried to base conclusions on that. Even when i told you why they are so inacurate you still claimed to be correct.


    Now can someone from the devs or someone else who knows it clarify those changes so we can end this discusion? Obliviously the results headhunter got are useless.

    Quote from Rick

    Trying to make me look like the wrong one eh?


    You're not even content with getting to be right, are you? You gotta flog this dead horse until you're satisfied.
    I have some news for you, guess you missed it from my previous references: A change was made in 1.337 (per the update notes), and reversed in 1.337b (per the update notes). I don't know how much more "official" you can get.
    Asking for raw numbers is meaningless as the change was reversed.
    My observation was not made for your benefit - so if you're not satisfied with my method or my result, do it yourself. But for Frak's sake, stop harping on it already. You're starting to sound like a broken record. It's obvious that you're the only one who cares about your interpretation of what I meant by my original post.


    In the words of one of my favorite philosophers: "If a man is offered a fact which goes against his instincts, he will scrutinize it closely, and unless the evidence is overwhelming, he will refuse to believe it. If, on the other hand, he is offered something which affords a reason for acting in accordance to his instincts, he will accept it even on the slightest evidence."

    I don't mean to interrupt here, but if you have 1 6 windmills each making 2 eu/t you're going to have 32 eu/t f you measure it before it his a batbox/mfe/mfs or any of the converters. However if you run this into a mfe or mv converter and have it output 512 ticks you can read 500 from the mfe or mv convert, but that won't be per tick. It will take multiple ticks to reach the 500 eu before the mfe or mv converter sends the tick out. So saying 16 windmils makes even "up to 500 eu/tick" isn't accurate. You may be able to get an accturate reading however by takign a reading and then letting a few thousand ticks to by and then taking the reading again, but to get the best reading you should do it before it hits anything that stores or converts the eus.

    Quote from thorbo

    However if you run this into a mfe or mv converter and have it output 512 ticks you can read 500 from the mfe or mv convert, but that won't be per tick. It will take multiple ticks to reach the 500 eu before the mfe or mv converter sends the tick out. So saying 16 windmils makes even "up to 500 eu/tick" isn't accurate. .


    Once more... I said the reading is up to 500 EU/t. I'm well aware that this is not the actual power generated each tick - and despite what anyone says, you won't get a consistent reading on that anywhere with wind power, due to its fluctuating nature.


    I wasn't concerned with the individual or collective actual output per tick - just with the best reading I could get. It was an informal benchmark that supported the implications in the patch notes.

    If you want a reading that means anything, you measure the amount of time between 512eu pulses over a longer period of time with a stopwatch. Get an empty mfe, hook it up, measure the time it takes to fill it up to some amount, divide that by the time in seconds, divide that by the number of ticks in a second. Then you'll have average ticks per second, which is a useful measurement. Since it's random you do it a lot, or over a longer period of time to help average it out.
    What you measured won't tell anything very meaningful. I get what you're saying you did, but it's not very useful.

    I didn't need an exact reading - keep in mind that the time between the origination of this thread and the fix was a matter of mere hours. I took an informal reading when I first heard of the "nerf" and what I saw was sufficient to indicate that my power output was not what I'd previously come to expect. After the "b" fix, I did the same thing again and noticed it was basically along the lines of what I had in 1.23.
    If I wanted to quantify exactly how much the change was, I'd have done more thorough testing in the manner you describe - but the issue wasn't really "how much" as "did they change the output?" And by the next day, it had been changed back.


    This is a thread that's carried on long after the obsolescence of the update that spawned it, and it's serving no purpose now except to reiterate what's already been said and gnaw at the parched bones of details that are no longer relevant.

    Am I the only one that understands HeadHunter here? Yes he made a slight slip in the beginning but then he corrected himself and you guys are still badgering him. The main fact is that, like HeadHunter said, there was a nerf then a buff on wind. So he measured in HV, that doesn't really matter.


    Why do I get the feeling I'm going to get flamed for this...

    Quote from Blue_Sogekihei

    Am I the only one that understands HeadHunter here? Yes he made a slight slip in the beginning but then he corrected himself and you guys are still badgering him. The main fact is that, like HeadHunter said, there was a nerf then a buff on wind. So he measured in HV, that doesn't really matter.


    Why do I get the feeling I'm going to get flamed for this...

    When you measure it in EV it wont matter how much eu the windgen farm is making. Its even possible to measure 2k eu/tick while the windgen farm produces only 1 eu/tick. In fact its even possible with a single watergen...This is not just a bit inacurate but its extremely inacurate. Saying you get 500 eu/tick from 16 windgens can confuse newbies who believe it and it doesnt proof anything changed with windgens. So it does matter alot if you measure it in EV instead of LV.


    And yes iam aware windgen mechanics changed in the last versions but this 'test' he did doesnt proof anything and it only confuses ppl if he puts wrong data like that on the forum.

    While I will agree it is /easier/ to measure on the parts of the system directly before storage/use, it is possible to get accurate-enough measurements even on HV links. It's just that the sample-period becomes really annoying. For any given link this is a good rule of thumb for sample accuracy: Measure for the number of energy-ticks that the fame size is for the link (E.G.: HV is 2048 so ~100 seconds or 2 min); If the number is > 10% of the link's packet size then you're accurate enough. If the number is less than that you need to measure for 10 / (the percentage as a number) times longer. (So if it's 1% of 2048 or ~20 eu/t you need to measure for 20 min to get an accurate reading in that spot; 5% would be 4 min).


    However, yes, it's far easier to just measure at the storage device or just before an MassFab.

    Quote from Rick

    He is too stupid to understand it. Dont waste any more time on him. I already tried explaining it to him several times.


    I told you I'll let you be right and you responded with "still trying to make me the wrong one?" Sounds like someone is "too stupid to understand" English.
    If "right" means "wrong" in your world, then it's no wonder I can't understand you.


    You don't have the sense to know when to drop it and you don't have the grace to be content with "being right". You're what most people would call a "sore winner".
    All things considered, can you see, then, why your opinion of me (let alone anything else) is of no importance to me?


    Are you so insecure that you can't be content with just making your point and leave it at that? Why do you always feel you have something to prove?

    So, if I understand this correctly, Headhunter measured a HV current over a large period of time (Note the key "large period of time", this is an accurate method of measuring current, should the windmills be the only thing feeding into it) in one version, then noted a sharp decrease in the next, followed by an increase back to the previous amount (roughly) in the hotfix (Note that relative measuring, as in a measurement before, during, and after something, is better then exact measurement). Cool, they're approximately back to normal. Does that mean we can lock this trainwreck of a thread yet?

    He measured it for 4 ticks. Thats 0.2 seconds...Else you wont get 500 eu/tick with eu reader from 16 windgens.
    1th tick: nothing
    2th tick: nothing
    3th tick: nothing
    4th tick: 2k eu


    Thus average eu/tick was 500 but thats very inacurate this way. In order to get a accurate result you would have to wait a really long time like MJevans already pointed out.

    This is pretty much the end, and could've ended several posts and replys ago. The caculations are correct. There is not much else to be said here, and this problem is already resolved. IF I could close this thread I would. The only people i could see posting past this would be either trolls, flamers, or someone closing this thread. That is all i have to say. Except for one thing, what is the timezone of this forum, it was a little over 7:00 EST when i posted this and its saying past 2:00PM?

    This will NOT be the last time you see me adventurer. Good luck.

    EDIT: TL;DR


    A wind generator will on average generate 2.5 EU/t (*) but can generate as much as 7.75 and as little as 0.00266666667 EU/t (it will only send whole EU packets though. Packet sizes will be between 1 and 8 EU and will be sent when the fractions of an EU exceed 1.0)
    During a thunder storm, production increases by 50%.
    During rain (not a thunder storm and not sunny), production increases by 20%


    There is a VERY VERY small chance that a wind generator breaks (replacing it with a standard generator and dropping 1 to 4 iton ingots for the blades) if it's production should exceed 5.0 EU/t. At most this chance is 0.0075%. Every 6.4 seconds.


    * - The average is slightly higher than that. Read the full post if you want to understand why!


    ...
    This is out of interest for another thread, but here goes...


    Currently (as of 1.337b) wind generators operate as following:


    each tick (20 times per second) the generator gains "fuel". (internal workings)



    Each 128th tick the generator updates it's production values. (every 6.4 seconds, this)
    Every 128th tick the generator updates its obscured block count. (except for ticks 1024, 2048, 3072, 4096 and so on. This is most likely a bug! It should be every 1024 ticks or 51.2 seconds. I have made a bug report about this)


    When the generator updates it sets it's "subproduction" (temporary) value to:


    subproduction = windStrength * (height - obscuredBlockCount) / 750


    Wind strength is a random number between 10 and 19 initially. Wind strength CAN change every 128th tick (6.4 seconds) and can go as high as 31 and as low as -1. (Yes, this means subproduction can reach -0.1666)
    If wind strength is less than or equal to 20 then it has a flat 11% chance of increasing by 1. For each strength above 20 this chance is reduced by 1%. At strength 31 there is a 0% chance of wind speed increasing.
    If wind strength is more than or equal to 10 then it has an 11% chance of decreasing by 1. For each strenth below 10 this chance decreases by 1. Even at a strength of 0 there is still a 1% chance of it decreasing by 1 to -1)
    Do note however, IF wind speed increases then there is no chance of wind speed decreasing. So there is a 11% chance that wind speed will not decrease alongside the 11% chance of it increasing. I'm not good with such math but someone else probably is.


    If subproduction is less than or equal to 0, "fuel" gain fails. The generator will produce as much power as it did when it was last updated. (thus, a wind generator will never produce negative energy even with a negative windStrength)
    However, if it is positive then we proceed to make some additional adjustments to subproduction.


    If it's thundering then we multiply subproduction by 1.5.
    if it's raining then we mutliply subproduction by 1.2. (Even though it's raining during a thunder storm, in game code it is ONLY raining when it's raining and not a thunder storm. thus, you cannot multiply it by both 1.5 and 1.2 during a thunder storm!)


    At this stage, if subproduction is greater than 5 there is a ((subproduction - 4) / 5000) * 100 percent chance that the wind generator will break and drop 1 to 4 iron ingots.
    For example, say you have your wind generator at the top of the map (height 127) and the wind strength is 31, during a thunderstorm. You only have 2 cables going to the generator through the obstruction area. The math is:


    subproduction = (31 * (127 - 2) / 750) * 1.5 = 7.75


    You run a 0.075% chance of breaking your wind generator. That is, at THAT production level, one wind generator will break on average every 8533.33 seconds. (~2 hours 22 minutes 13 seconds)
    Pretty small chance but it does exist.


    Now, if all has passed (subproduction is greater than 0 and your generator hasn't broken from overproduction), multiply subproduction by 100 and divide it by 100... In other words, leave it as is! (The reason i mention this is because it is in the source code. A constant value called energyGeneratorWind exists that is set to 100 but has probably been another value in the past.)


    Right, you now have your subproduction temporary value. Your generators energy production is the integer part of that value.
    Let's say the IC2 wind strength is an average 15, your generator is at height 127 and has 2 cables going into it and all other blocks around it is air blocks. The weather is sunny and clear.


    subproduction = 15 * (127 - 2) / 750 = 2.5



    Your production every even tick is 2. Your production every odd tick is 3. Your average production is 2.5.
    The reason it works like this is because EU's cannot be split in decimals. But the wind generator will compensate for that by holding on to the fractions of an EU for you.


    As far as production goes, this is all you need to know about wind generators in IC2 1.337b.


    ---------------


    Note: Alblaka might ask me about the chance percentages, how they can be correct... Reason is simple! You are using <= instead of just < in code. This makes it include 0 as a valid chance. F.E: random.nextInt(10) gives you a value between 0 and 9, not 1 and 10 or 0 and 10. If you check for random.nextInt(100) <= 10, there are 11 chances that the condition may be true. (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 are all <= 10)


    LATE EDIT: Clarified that when a wind gen breaks, it is replaced by a normal generator and 1 to 4 iron ingots are dropped in it's place to simulate the blades breaking off.