Best way to defend

[NotABiter]

You guys might be interested in how to make solar reliable for laser defence.

That seems too complicated, wasteful, and flawed. It's wasteful because you're burning up a lot of power just for a sensor (though IIRC later in the thread someone has a better version of it). It's flawed because (if I'm understanding it right, and note that using a defense power system one doesn't understand is also dangerous) it gives you "just enough to get through the night", but that means you have approximately 0 power in reserve if you get attacked shortly before dawn. A simpler and more robust way to go is to reserve X% of your accumulator capacity for your defenses - it just requires one wire from one accumulator to one switch which powers your non-defense stuff (and of course keeping defense and non-defense on separate electric networks).

no drain lasers

That seems cheaty to me. The drain is supposed to be the cost of having that laser ready to go. If this cheaty solution becomes popular, then the devs will probably fix it (perhaps by making vanilla lasers a bit like the Yuki ones - try to power up a whole bank of Yuki lasers at once and the massive initial charge they draw will take out your whole power grid; or perhaps by having a long "initialization period" after being powered up, during which the lasers simply won't work).

[Qon]

You guys might be interested in how to make solar reliable for laser defence.

That seems too complicated, wasteful, and flawed. It's wasteful because you're burning up a lot of power just for a sensor (though IIRC later in the thread someone has a better version of it).

That really isn't much power at all for a base big enough to power your lasers and your base by solar. 7 radars for the biggest version is just 2.1 MW. If you have a 100MW base then using this to guarantee that your defence has power that isn't really a bad deal. 2.1MW is nothinhg. And you don't "burn" any energy, solar power is endless. There is only the initial cost. And I made the big version just so that the principle could be demonstrated easily without complex combinator magic that no one would understand. And as you said, I did suggest a cheaper 1 panel and accumulator solution. But maybe you should postpone relying on solar+laser for defence until the power drain of radars isn't a big deal.

It's flawed because (if I'm understanding it right, and note that using a defense power system one doesn't understand is also dangerous) it gives you "just enough to get through the night", but that means you have approximately 0 power in reserve if you get attacked shortly before dawn.

Well you didn't understand it right then. The accumulators will not run out shortly before dawn. They will only be allowed to be depleted by your low priority network when the sun is high enough that you get exactly the game day average of 42kW of energy/panel so that they can provide all the power you need. The system is perfect

A simpler and more robust way to go is to reserve X% of your accumulator capacity for your defenses - it just requires one wire from one accumulator to one switch which powers your non-defense stuff (and of course keeping defense and non-defense on separate electric networks).

It's extremely simple. One system of 50 panels, 42 accumulators and 7 radars in one grid. A switch that turns off your low priority network when your main grid has less accumulator charge % than the sensor system. That's it. Takes a few seconds to build and understand. Your system will allow your main base to drain your accumulators to the "reserved amount" during the day, before dusk even begins! So you need an extreme amount of accumulators to last the night with that system. Which solution do you think is cheaper then? Mine costs 42 accumulators maximum no matter how big your power network is. With your system, if you reserve say 50% for laser defence, then you have to make make your accumulator field 2 times bigger for those 50% to be enough for the night.

no drain lasers

That seems cheaty to me. The drain is supposed to be the cost of having that laser ready to go. If this cheaty solution becomes popular, then the devs will probably fix it (perhaps by making vanilla lasers a bit like the Yuki ones - try to power up a whole bank of Yuki lasers at once and the massive initial charge they draw will take out your whole power grid; or perhaps by having a long "initialization period" after being powered up, during which the lasers simply won't work).

They just added the switch for us to use it. If you can build smart factories with the circuit network then you should be rewarded for your ingenuity. The circuit system has been useless until now, finally the good players can actually improve their base with it and not just use it for novelty stuff like text displays. This is a step forward for factorio.

[Fahrradkette]

Usually I go the other way round: cutting power supply for the miners and even some production like assemblers mk3. Using gophers S-R latch doesn't even give me flickers.

[Qon]

Usually I go the other way round: cutting power supply for the miners and even some production like assemblers mk3. Using gophers S-R latch doesn't even give me flickers.

What do you mean "The other way around"?

If you specify AM3, does that mean you use AM2 also? Why? AM3 is cheap enough fairly quickly after you unlock it if you just automate production of it (you aren't a naughty handcrafter, right?). And since it has more module slots and uses less energy/craft even without modules there is no reason to ever use AM2 after you unlocked AM3.

My system will give you flickers, but only in case your solar system is not big enough to power your base (and your lasers will not flicker, it will stay on to keep you safe). Consider it your warning. It's better than having all your base, including lasers, to power off permanently until dawn.

[NotABiter]

And you don't "burn" any energy, solar power is endless.

You have 50 solar panels there producing energy, and you burn all of that energy by sending it into radars. I look at that setup and I see 50 solar panels that "aren't connected right" - they should be powering the grid. Ditto for those wasted accumulators.

There is only the initial cost.

Well, there's also an ongoing cost - they take up space. In my current base that's still an issue.

But maybe you should postpone relying on solar+laser for defence until the power drain of radars isn't a big deal.

The base I posted has been on pure solar long enough that (as I already stated) it triggered the "Solaris" achievement. But 50 solar panels and 42 accumulators would be ~10% of what I have, crippling my power system if I pulled out that many and used them for this sensor. So no, being wasteful with giant power-sucking "daylight sensors" is not a good reason to postpone going solar.

Well you didn't understand it right then. The accumulators will not run out shortly before dawn.

I didn't say they run out, I said "approximately 0 power in reserve".

But in fact, they do run out, at least as far as I can tell (the numbers move very fast when the power in the 42 accumulators is < 100kW). I just plopped down the 50/42/7 setup, and that's what happens:

Code: Select all

energy   light level (solar panel power)
------   -------------------------------
210 MJ   100%
138 MJ     0% (start of dark period)
~60 MJ     0% (end of dark period)
~39 MJ    10%
~27 MJ    20%
~16 MJ    30%
 ~8 MJ    40%
 ~3.8 MJ  50%
 ~0.6 MJ  60%
 ~0 MJ    67%

Consider my current base which has about 500 solar panels. Even at 100% daylight, that's only enough power to fire 12 laser turrets (because they've got no help from accumulators because your system just let them run down to 0). If a biter group comes running into a wall, that may mean I don't have enough power to even fire all of the turrets that are in range - and that's just one biter group, heaven help me if two groups attack at once in two different places. It's even worse than that though, because when your system lets my accumulators run down to 0% I only have 67% light, so my 500 panels are only running at 67% so I've now only got enough power to run 8 lasers. Oh, and it's even worse than that because I allocated 50 solar panels to your sensor so my main grid only has 450, so now I'm down to just 7 lasers working.

And it would appear it's even much worse than that. The above paragraph is based on a laser requiring 2.4 MW, which is what the floaty text says when I mouse over the laser in the build menu. But when I plop a laser down and look at its power requirement, it says 6.2 MW! That's (presumably - my calculations support this theory) because I have all of the laser shooting speed upgrades, and the laser draws more power when it shoots faster. So now instead of 7 lasers firing, I'm down to just 2 lasers able to fire!

The system is perfect

More like horribly flawed.

Only when you've got so many solar panels that you can easily cope with all attacks (including simultaneous attacks) purely on 2/3 of their nominal power output (with no accumulator help) would I even consider your system "usable". E.g., if you want to be able to have 40 (upgraded) lasers fire, then you need 6,200 solar panels. That's more than 12 times the number of solar panels I have. I'm safely running on solar now, I don't want to wait until I produce another 6K panels before doing so. (In fact, as I stated in my original post, my current game is pretty much over - the biters are dominated, so I'll probably be moving on soon. If I had to wait until I had 6.2K solar panels, I might never switch to solar. Not everyone's into building "mega bases" - some of us want the classical succeed/fail challenge that games can provide, which I also essentially stated in my original post - I only played this map because someone else was getting killed on it and I wanted to see if I could succeed where they had failed.)

Your system will allow your main base to drain your accumulators to the "reserved amount" during the day, before dusk even begins!

Now you're just being silly. My solar panels (once I've switched from steam to solar) are never going to be sized so small that they can't even run the base in full daylight. I've been playing the game for over a year now, and it's never happened. I don't design systems around stuff that never happens.

Which solution do you think is cheaper then?

For starters, the one that doesn't get my base destroyed. With your system the only thing that prevents total base destruction is biters being too stupid to go for critical power structures after they've punched through my lasers which were gimped by your system.

With your system, if you reserve say 50% for laser defence, then you have to make make your accumulator field 2 times bigger for those 50% to be enough for the night.

Nice strawman.
What I would much more likely do is set that % to be just a bit higher than the % at which the steam engines kick in, so shutting the factory down is just an attempt to keep defenses strong without producing pollution, and if that doesn't work then steam kicks in (while accumulators still have enough juice to fire more lasers than I'll ever need at once). I might even set it to the same (or even lower) % -- really the main purpose would be to ensure that, in a biter-attack emergency, my factory isn't overpowering my steam engines when that power needs to be going into accumulators to be ready for the next attack.

Now if I were designing the "perfect" system, it would not do what your creation does. Instead, it would constantly calculate how much energy idle lasers will consume that solar panels will not cover (due to complete or total darkness). It would then add to that how much stored energy I want to have in reserve in case of an "attack burst" (attacks that are simultaneous or too close together for accumulators to recharge between them). It would then cut off the factory anytime that amount was >= the total amount of energy currently stored in the accumulators. That then allows the factory to run as much as possible, but when accumulators start getting low it shuts the factory down early enough to avoid steam kicking in IF there's no attacks - i.e., if the only power consumed is laser idle power. (Optionally, a higher reserve could be set if you want to handle some attacks without steam coming on.) This scheme is a valid strategy for all solar powered bases, big and small. (It could also be adapted to non-solar use -- even if you're running off steam you may want to cut power to the factory if accumulators get dangerously low.)

Your thinking would appear to be skewed towards very large bases where laser idle draw is the major factor and laser firing power is a non-issue. For more modestly sized bases, e.g. bases made only to beat the game - not to just build for the sake of building - laser idle draw isn't as bad, and laser firing power is an issue that one ignores at their own peril. It takes 258 lasers sitting idle to equal the power draw of just 1 laser firing, 10,333 idle lasers to match the power draw of 40 lasers firing. I don't have 10,000 lasers. I don't believe I've ever had 10,000 lasers in a game of Factorio.

They just added the switch for us to use it. If you can build smart factories with the circuit network then you should be rewarded for your ingenuity.

They just added <<feature X>> for us to use it. If you can <<exploit feature X in game-reality-breaking fashion>> then you should be rewarded for your <<"ingenuity">>.
Your position requires me to believe that lasers don't actually do anything with that drain power. They just have a giant bleeder resistor in there for no reason. Sigh, if only the engineers that designed the laser turret would have been smart enough to leave that bleeder resistor out of their design.

[Qon]

It's meant for a base big enough that the solar can easily support your laser turrets. The point is to prevent your base (which can expand to use a lot more and drain all your power) from killing your defence. Pretty much all your arguments boils down to the fact that your base is too small to use solar+lasers reliably and without buffers. If your lasers can drain your accumulators when your base is powered off then you need buffering of power, steam backup or gun turrets instead to help you.

And you don't "burn" any energy, solar power is endless.

You have 50 solar panels there producing energy, and you burn all of that energy by sending it into radars. I look at that setup and I see 50 solar panels that "aren't connected right" - they should be powering the grid. Ditto for those wasted accumulators.

There is only the initial cost.

Well, there's also an ongoing cost - they take up space. In my current base that's still an issue.

Whatever. It's different points of view. The raw resources can become a sensor or they can generate 42kW of energy. You don't think "Those resources I crafted that beacon with could have generated more power instead". If you need both, you craft both. The sensor is worth the cost. 1 panel+accumulator is extremely cheap and is small enough to fit in the smallest stripe worlds.

But maybe you should postpone relying on solar+laser for defence until the power drain of radars isn't a big deal.

The base I posted has been on pure solar long enough that (as I already stated) it triggered the "Solaris" achievement. But 50 solar panels and 42 accumulators would be ~10% of what I have, crippling my power system if I pulled out that many and used them for this sensor. So no, being wasteful with giant power-sucking "daylight sensors" is not a good reason to postpone going solar.

Use 1 panel+accumulator then.

Didn't say that you should postpone solar. I said you should postpone pure solar with pure laser and no backup. Unless the biters are set to low you risk running out of power from lasers fairly quickly with a small base.

Well you didn't understand it right then. The accumulators will not run out shortly before dawn.

I didn't say they run out, I said "approximately 0 power in reserve".

But in fact, they do run out, at least as far as I can tell (the numbers move very fast when the power in the 42 accumulators is < 100kW). I just plopped down the 50/42/7 setup, and that's what happens:

The accumulators in the sensor do run out exactly at the moment where the solar panels give enough energy to support the radars. The combinator does the same for your power grids, lets your buffer run out when your solar can sustain your lasers without the buffer. If your solar power isn't big enough to support your lasers without buffers, rely on something else.

The system is perfect

More like horribly flawed.

Yes, it's not designed for small bases. It's "horribly flawed" for a base that can only support 3 turrets continously. But I wrote about how you could combine this with a buffering system to allow smaller bases to use it. If you remove a radar/limit the radar power further in the 1 panel+accu setup the sensor will never drain completely so your low power grid would never be allowed to drain the accumulators completely. And my design is not perfect in all aspects, but for the purpose it was intended for it is an extremely simple way to give your high power network the same "laser priority" property that steam has, and in that aspect it is perfect. That was how I introduced the concept in my post.

Your system will allow your main base to drain your accumulators to the "reserved amount" during the day, before dusk even begins!

Now you're just being silly. My solar panels (once I've switched from steam to solar) are never going to be sized so small that they can't even run the base in full daylight. I've been playing the game for over a year now, and it's never happened. I don't design systems around stuff that never happens.

I've seen a lot of people on youtube run out of power because they expand their power system after their base. And they use laser+solar. You haven't really done any big production expansions yet, but it still impressive that your keep your power system adequatly sized. But it doesn't have to be so bad that your base can't run at 100% at full daylight. If your power system can't sustain your factory and your lasers at the same time then the system is still useful.

With your system, if you reserve say 50% for laser defence, then you have to make make your accumulator field 2 times bigger for those 50% to be enough for the night.

Nice strawman.
What I would much more likely do is set that % to be just a bit higher than the % at which the steam engines kick in, so shutting the factory down is just an attempt to keep defenses strong without producing pollution, and if that doesn't work then steam kicks in (while accumulators still have enough juice to fire more lasers than I'll ever need at once). I might even set it to the same (or even lower) % -- really the main purpose would be to ensure that, in a biter-attack emergency, my factory isn't overpowering my steam engines when that power needs to be going into accumulators to be ready for the next attack.

If you have a steam backup plant, the my system is fairly useless. My system is designed to give you the same reliability that steam gives you even with pure solar (no steam backup). If you have a steam backup plant then that will take care of anything you throw at it. A steam backup plant can easily supply enough power. Why would you use a system to give you the reliability of steam when you have steam?

Now if I were designing the "perfect" system, it would not do what your creation does. Instead, it would constantly calculate how much energy idle lasers will consume that solar panels will not cover (due to complete or total darkness). It would then add to that how much stored energy I want to have in reserve in case of an "attack burst" (attacks that are simultaneous or too close together for accumulators to recharge between them). It would then cut off the factory anytime that amount was >= the total amount of energy currently stored in the accumulators. That then allows the factory to run as much as possible, but when accumulators start getting low it shuts the factory down early enough to avoid steam kicking in IF there's no attacks - i.e., if the only power consumed is laser idle power. (Optionally, a higher reserve could be set if you want to handle some attacks without steam coming on.) This scheme is a valid strategy for all solar powered bases, big and small. (It could also be adapted to non-solar use -- even if you're running off steam you may want to cut power to the factory if accumulators get dangerously low.)

Please show me an example implementation. Seems interesting.

Your thinking would appear to be skewed towards very large bases where laser idle draw is the major factor and laser firing power is a non-issue. For more modestly sized bases, e.g. bases made only to beat the game - not to just build for the sake of building - laser idle draw isn't as bad, and laser firing power is an issue that one ignores at their own peril. It takes 258 lasers sitting idle to equal the power draw of just 1 laser firing, 10,333 idle lasers to match the power draw of 40 lasers firing. I don't have 10,000 lasers. I don't believe I've ever had 10,000 lasers in a game of Factorio.

Well you don't need 10k lasers for it to be useful, but yes it's meant for big bases where the game day average of your solar panels is enough to power your lasers. Maybe 50MW-100MW and bigger. If you have a single semisparse line of turrets then 10 lasers firing continously for an entire game-day could be supported. If you want more than that you need more power. Or you use my system but combine it with some extra buffering.

They just added the switch for us to use it. If you can build smart factories with the circuit network then you should be rewarded for your ingenuity.

They just added <<feature X>> for us to use it. If you can <<exploit feature X in game-reality-breaking fashion>> then you should be rewarded for your <<"ingenuity">>.

Your position requires me to believe that cutting power to structures that use it needlessly is "exploitation" of the power switch. Please enlighten me how to use the power switch without doing something naughty like limiting power/fuel usage with it! The switch was given to us to completely turn off structures that don't need it and are just wasting it on drain or not important things at the moment. It is useless if shutting things off is not allowed with it. I think the factorio devs are very pleased with the no drain laser wall and that it is exactly the kind of creativity that they want to see from the community with the new powerful and programmable tools they have released upon us.

If you think a power switch is a reality breaking feature then you are either from the far future where nothing can be unplugged anymore and everything is forcefully on standby, or from the past where electricity wasn't invented yet. Read what you wrote again and realise how silly it sounds.

Your position requires me to believe that lasers don't actually do anything with that drain power. They just have a giant bleeder resistor in there for no reason. Sigh, if only the engineers that designed the laser turret would have been smart enough to leave that bleeder resistor out of their design.

The power is obvously for searching for biters. That is what it IS used for. Why do I have to power every single turret so that every one of them can do it when it's enough with 1 spotter? So the power is used for something, which is exactly why you let 1 spotter turret every ~12 tiles drain energy all the time. Why a bleeder resistor? The engineers of the single turret didn't think of connecting the turrets so that only one spotter was running at a time, but I did. K? I engineered a laser wall more efficient than the naive implementation. The game lets some puzzles be solved by the player. It would just be bad game design to give use the switch and then automagically let turrets drain less when close to other turrets without any need for the player to play with the circuits himself.

[TheSkiGeek]

The power is obvously for searching for biters.

This is not so obvious when gun turrets do the same "searching" with zero electrical power input.

Whether taking advantage "instant-on" laser turrets is an "exploit" or a "feature" is not so clear-cut. It certainly removes a seemingly intended downside of spamming laser turrets everywhere.

[NotABiter]

A steam backup plant can easily supply enough power. Why would you use a system to give you the reliability of steam when you have steam?

You say that as if every steam setup were essentially infinite in size or something. That's just ridiculous - it's not so. Look at the picture of my base I posted. My steam power is only 20 steam engines. That's only enough power for 1 (upgraded) laser. Just like you'd need 10K+ solar panels to power 40 lasers from just solar panels, you'd need 487 steam engines to power 40 lasers just from steam engines. The biggest steam power setup I've ever built (multiple times bigger than any other steam setup I've ever built before or since) only had 240 steam engines.

Having accumulators and keeping enough reserve charge in them at any given time is how I have enough power to fire lasers. Not solar panels. Not steam. As far as laser defense is concerned, solar and steam are effectively only there to charge up accumulators between battles, not to power lasers during battles. It's been that way in every last game of Factorio I've ever played. The judgement calls you are making have never been valid for me, and I would guess have never been valid for the majority of Factorio players. (Even my inadequate-by-your-standards 240 steam engine setup got a "Wow! Very impressive steam engine set up!" comment, suggesting I'm not alone in my thoughts that most people don't build steam big enough to directly power 40 lasers firing at once.)

Please show me an example implementation. Seems interesting.

If/when I get around to making it I'll post it. (Note that I said "if I were designing the "perfect" system" - I haven't actually designed or built it yet.) If you just agree with everything I say then maybe I can spend my "Factorio time" actually playing the game and designing this thing rather than debating here.

Maybe 50MW-100MW and bigger.

40 lasers is 247MW. That's 372MW nominal after accounting for the 67% light @ the 0 MJ point.

Even 40 lasers firing at once isn't that much. That's good enough for base defense, but if you're playing "death world" and turret creeping with robots, 40 is not a lot of lasers. E.g., each segment of this "battle wall" has 70 lasers in it, and when plowing directly into a mass of biter bases (look where I'm headed in the mini-map), it attracts a lot of "attention" from the biters, with attacks coming from 3 sides and potentially lighting up more than one "battle wall" segment at a time. (And of course, occasionally, while I'm doing this there could actually be one or two attacks happening at my base as well, pushing the active laser count even higher). You'd need more solar than this map would support if you tried to play it without proper reserve power in accumulators. (The map string's right there in the thread if you want to try it.)

If you have a single semisparse line of turrets then 10 lasers firing continously for an entire game-day could be supported.

And? What kind of scenario is that? When does that ever happen? Defensive battles are sporadic, not nice and constant like that. You'll have nothing for a while, and then a giant flock comes swooping in all at once. (And if biters are an actual threat, "semisparse" is not how you want your turrets to be.) I don't want 10 lasers to be able to fire all day long, I want 40+ lasers to be able to fire for enough seconds to wipe out the group of biters in the current attack. Only during offense (turret creeping) is a sustained power draw due to laser firing a reality in the game, and in that scenario you the player are in control of how long that power draw lasts - if you've driven the accumulators low by fighting (somewhat) constantly with way more lasers than your baseload power can handle, then you back off and give the accumulators a chance to recharge. (Previously this meant occasionally checking the power grid while creeping, but if I were making that "battle wall" with 0.13, I might make it so the lights would change color to indicate the amount of charge left in the accumulators, maybe flash if it got too low, and then I wouldn't have to bother with clicking on power poles/substations.)

Please enlighten me how to use the power switch without doing something naughty like limiting power/fuel usage with it!

Use it for beacons. Use it to make things not do stuff you don't want them to do (other than just idle-power drain), e.g. turn off the chemical plant if you don't want it making more lubricant because you want heavy oil to go to cracking. Even if you used the switch for lasers, not to avoid the drain but to avoid firing because you didn't want it to fire (e.g. maybe you switch to pure guns every night and you don't want lasers to fire and thereby use up the accumulator power your factory needs), that would be a non-cheaty use (though such usage would still exhibit the current in-my-opinion-flawed reaction to power on/off events, at least until the devs fix it -- but in this case most of that idle draw reduction would be OK as it's over a very long period of time and the turn on in the morning is at a fixed time and doesn't really exploit the instant-on behavior).

If you think a power switch is a reality breaking feature then you are either from the far future where nothing can be unplugged anymore and everything is forcefully on standby, or from the past where electricity wasn't invented yet.

What happens if I walk up to some random desktop computer that's running (e.g. web browser open to Factorio forums), pull the power plug and then some time later plug it back in? Does it just instantly return to the Factorio forums? Or are we now faced with BIOS and OS boot times, possibly a prompt asking you if you want to do a normal boot up (because your computer just went down and the OS doesn't know what happened), and once it boots up you're looking at an empty desktop and you have to get back into the forums yourself? Even for complex embedded systems (e.g. aircraft electronic flight displays), they aren't instant-on (though they are much faster at booting than your average desktop - just seconds to start up).

The laser turrets in Factorio obviously have computers in them. They are recognizing and selecting enemy targets. They can tell friend from foe. They never accidentally shoot you or one of your buildings.

The reality-breaking part isn't that you can switch power, it's the (current) instant-on ability of the laser turrets. That's what it seems like you're exploiting - that you can apply power and they start shooting (essentially) immediately, there's no boot time, no target acquisition time. So you're getting the benefit of being powered down when not needed, but not paying the consequences I would expect there to be for doing so. That doesn't mean powering down lasers shouldn't be an option, or even a good idea (at least in the right situation), but those are not incompatible with there being some consequences.

I mean, in Factorio-land just subtracting two numbers takes 1/60th of a second. Is it not reasonable to expect that booting up an advanced vision recognition system (which needless to say is a lot more complex than a subtractor) in that same game world might take a second or two?

The power is obvously for searching for biters.

That sounds plausible to a degree, but the instant-on part still seems wrong. "Drain power" is the price I pay to keep my desktop available 24/7. One could (correctly) say that drain power is powering the display output, checking for keyboard input, etc., but the only reason I pay the drain power cost is because of the lack of instant-on. (Sadly, I've yet to have a computer where standby mode works right, and even standby modes make you chose between power drain or recovery times, never neither, though if you have an SSD it's closer.) Due to the way my place is wired, I actually have my computer on a wall switch next to various light switches. But I have taped that switch so it doesn't accidentally get flipped because my computer doesn't behave as well as Factorio laser turrets when you interrupt its power.

It would just be bad game design to give use the switch and then automagically let turrets drain less when close to other turrets without any need for the player to play with the circuits himself.

Note that I never suggested the game do that. In fact I made no actual suggestions at all. What I did was predict that if this particular usage becomes common the devs might fix it, and I described two ways they might do so, neither of which is anything like what you are now describing.

What is generally accepted as bad game design (in games with combat) is making any one weapon over-powered/unbalanced compared to the others. Part of the cost to using lasers is idle power draw. If that cost can now be side-stepped, then the balance has shifted, and even from a pure gameplay perspective (ignoring how reality-breaking instant-on is or isn't), the devs may have to do some rebalancing as a result -- not that Factorio has ever actually been balanced.

[Qon]

A steam backup plant can easily supply enough power. Why would you use a system to give you the reliability of steam when you have steam?

You say that as if every steam setup were essentially infinite in size or something. That's just ridiculous - it's not so. Look at the picture of my base I posted. My steam power is only 20 steam engines. That's only enough power for 1 (upgraded) laser. Just like you'd need 10K+ solar panels to power 40 lasers from just solar panels, you'd need 487 steam engines to power 40 lasers just from steam engines. The biggest steam power setup I've ever built (multiple times bigger than any other steam setup I've ever built before or since) only had 240 steam engines.

It's not infinite, it's just that now that steam is blueprintable having steam backup as big (in power output) as your solan panel field is fairly easy. If your backup power is undersized then that's your fault and not really an argument that steam can't handle your power needs. But as I said before, use energy buffers if you have tiny energy needs for your base.

The point of the accumulator syncing to an ideal system was that you could draw your 70% (not 67%) of solar panels max output, your game day average, all the time and no more with the low priority network. So that you get the same predictable and constant power output as steam. If that is not enough, use more energy buffering no matter if you rely on steam or solar. And my system can be combined with buffering as I said.

But if you want as much buffers as possible for laser offence then power down your factory until next dawn with full accumulators and go hunting then or whatever.

Please show me an example implementation. Seems interesting.

If/when I get around to making it I'll post it. (Note that I said "if I were designing the "perfect" system" - I haven't actually designed or built it yet.) If you just agree with everything I say then maybe I can spend my "Factorio time" actually playing the game and designing this thing rather than debating here.

Maybe 50MW-100MW and bigger.

40 lasers is 247MW. That's 372MW nominal after accounting for the 67% light @ the 0 MJ point.

Even 40 lasers firing at once isn't that much. That's good enough for base defense, but if you're playing "death world" and turret creeping with robots, 40 is not a lot of lasers. E.g., each segment of this "battle wall" has 70 lasers in it, and when plowing directly into a mass of biter bases (look where I'm headed in the mini-map), it attracts a lot of "attention" from the biters, with attacks coming from 3 sides and potentially lighting up more than one "battle wall" segment at a time. (And of course, occasionally, while I'm doing this there could actually be one or two attacks happening at my base as well, pushing the active laser count even higher). You'd need more solar than this map would support if you tried to play it without proper reserve power in accumulators. (The map string's right there in the thread if you want to try it.)

Well you are trying to use the system for something it was not designed to do. The point was to allow the game day average for your defence all day and night if needed but still run your base as much as possible. Use buffering on top of that and you have something usable for small bases. I agree with most of what you say, except when you are trying to argument for using a non-buffering system like a buffering system. You can easily use buffering with my system, just limit the power draw in the sensor so that it never goes below your target buffer%. You still get the benefit that it will reserve 100% for dusk so that you have more stored energy than your last reserves when night begins.

If you have a single semisparse line of turrets then 10 lasers firing continously for an entire game-day could be supported.

And? What kind of scenario is that? When does that ever happen? Defensive battles are sporadic, not nice and constant like that. You'll have nothing for a while, and then a giant flock comes swooping in all at once. (And if biters are an actual threat, "semisparse" is not how you want your turrets to be.) I don't want 10 lasers to be able to fire all day long, I want 40+ lasers to be able to fire for enough seconds to wipe out the group of biters in the current attack. Only during offense (turret creeping) is a sustained power draw due to laser firing a reality in the game, and in that scenario you the player are in control of how long that power draw lasts - if you've driven the accumulators low by fighting (somewhat) constantly with way more lasers than your baseload power can handle, then you back off and give the accumulators a chance to recharge. (Previously this meant occasionally checking the power grid while creeping, but if I were making that "battle wall" with 0.13, I might make it so the lights would change color to indicate the amount of charge left in the accumulators, maybe flash if it got too low, and then I wouldn't have to bother with clicking on power poles/substations.)

It's not meant to be a scenario that actually happens. The point was that you always have that power available for lasers at any point in time, even if your solar field is just big enough to support the base or your defence. While running your base in the night if the reserves allow it.