Friday Facts #235 - 0.16 stable

[Aardwolf]

Will you please also consider adding a bit more lane management in the game? Splitting two lanes of 1 belt onto two belts with splitter options (instead of slow inserters), and maybe a way to swap two lanes of 1 belt.

That would be a great QoL improvement, especially in earlier gameplay when not needing full belts with same resource on both lanes yet, to manage better what goes on which lane in particular parts of the factory. It would open up more ways of doing things.

Thanks!

[weaknespase]

[[Every bit related to win32]]

When i use CheatEngine to hack out the guts of games i like, i usually go for the code to make it harder, easier, of more fun by changing some in-game mechanics. From time to time it requires some heavy-duty disassembly and a few hours into the code but result usually worth it. In case of Factorio it's easier to do just using Lua, but not everything could be done with mods.

To filter out crashes related to such modifications, you should track at least two things:
1. External vmem allocations (i usually go for code caves, but standard approach is to use VirtualAllocEx to get yourself a piece)
2. Memory page access modifications (to write into read-only code section, technically should be case in every scenario).

Alternatively, for 2, you can substitute by running checksum on image's .text section contents and comparing it to the original. Probably, would be better to store checksum separately, instead of computing it from executable stored on disk.

No one in their right mind touches code from JIT (well, usually), so that should cover most use cases which involve code modifications. No luck with data writes, or, at least, i don't know any _reliable_ way to do it.

[maddoctor]

fuel stations!!!!

[luc]

Possibly:
[*] Spidertron (kovarex, GFX)

Nice try. I'm not falling for it this time!

[Zyrconia]

Great news!

And great plan for 0.17.

But I'd accept a 0.17 with no new features, but a spidertron! Or two spidertrons!

Or a version where you have a train network and a spidertron network, with an army of spiders delivering goods between outposts...

[Pohybel]

Pliss implement waving trees efect viewtopic.php?f=6&t=53640

[impetus maximus]

Will you please also consider adding a bit more lane management in the game? Splitting two lanes of 1 belt onto two belts with splitter options (instead of slow inserters), and maybe a way to swap two lanes of 1 belt.

they have given us the tools.

split

underground.exploit.png (118.8 KiB) Viewed 6933 times

swap

swap.png (44.49 KiB) Viewed 6933 times

[Ghoulish]

Would be great if we can get also a fluid filter like the inserters. Pump with a static filter like buffer chests would be great. This would improve unload and loading trains and filter unwanted mixed pipes ... maybe it allows new designing of oil factories and beacon layouts.
This also make unwanted conditions like "0.7 fluid > 0 " is false deprecated

Hopefully the Fluid Physics rework will make it... because that is one of the things that hasn't been touched in ages... and that is why it feels like it does........ Horrible.
May not be that obvious in casual games... but when pushing the limits with Nuclear Reactors it became quite obvious.

Yup - nuclear stuff shows how strange fluids are currently.
And if you start building things on your own there are quite a few traps that can lead to setups not working properly.

Totally agree. Fluid mechanics is a huge part of Factorio, yet seems to be quite simple with regards to the tools we get to manage how we want fluids to work. Having an item like a fluid splitter is essential, and logical for example, I know you can split fluid 50/50 with a couple of pumps on a storage tank along with 2 wires (or with combinators) But having a entity which can do this natively would be better, and especially useful for new players. I hope the pipes themselves get an upgrade and allow connections in specific directions only.

There's a MOD for that ™



The above is a picture from GotLag's Flow Control, and shows how you can connect pipes in a better, more useful way. The various valves in the MOD are great too, and the express pump is a god send when unloading fluid wagons in a timely manner. These sort of tools should be in vanilla in my most humble opinion.

I've gone through the list again and there is one item missing: "Neuter the bots."
What about the bots and belts topic? Is everything to stay as-is till at least 1.0?

Am wondering the same myself, there's 92 pages of debate here on the forums across 2 threads from Friday Facts, it's quite an emotive topic! Originally I though bots were OK, but when kovarex crunched the numbers it's quite clear how overpowered they are in relation to belts. So I now hope that belts are buffed or bots get a nerf - the fallout from a nerfing would no doubt flare up the forums again.. On this topic the bottom line for me is that a player shouldn't be penalised if they choose to use belts in the end game, it's considerably more complicated and requires a lot more forethought in its implementation, the result of which should at the very least be parity with bots.

[eradicator]

Arachnophobia is not a joke, why would any developer introduce spiders to alienate part of potential customers?

Do you even have Arachnophobia? Mine at least doesn't trigger on mechanical spiders. I have to agree though that rideable humanoid mechs might fit the setting better.

Am wondering the same myself, there's 92 pages of debate here on the forums across 2 threads from Friday Facts, it's quite an emotive topic! Originally I though bots were OK, but when kovarex crunched the numbers it's quite clear how overpowered they are in relation to belts. So I now hope that belts are buffed or bots get a nerf - the fallout from a nerfing would no doubt flare up the forums again.. On this topic the bottom line for me is that a player shouldn't be penalised if they choose to use belts in the end game, it's considerably more complicated and requires a lot more forethought in its implementation, the result of which should at the very least be parity with bots.

Belts got UPS optimized so much that they are at least on parity with bots for megafactories. And apart from that...haven't you noticed salami tacticts to slowly buff belts have already started :P?

[bobucles]

I don't think arachnophobia is a real concern here for many reasons:
1) It's late tech
2) It's totally OPTIONAL tech. If you really don't like spodes just don't get it.
3) It's a machine that doesn't really resemble a spider. The more steampunk it gets, the less it resembles the dreaded friendly arachnid.

Players can certainly have a distaste or hatred for spiders and they can certainly not like the idea for that reason alone. But that's very different from breaking down in cold sweats because a spider exists somewhere. A spider could even be under your desk right now!

[Ghoulish]

Am wondering the same myself, there's 92 pages of debate here on the forums across 2 threads from Friday Facts, it's quite an emotive topic! Originally I though bots were OK, but when kovarex crunched the numbers it's quite clear how overpowered they are in relation to belts. So I now hope that belts are buffed or bots get a nerf - the fallout from a nerfing would no doubt flare up the forums again.. On this topic the bottom line for me is that a player shouldn't be penalised if they choose to use belts in the end game, it's considerably more complicated and requires a lot more forethought in its implementation, the result of which should at the very least be parity with bots.

Belts got UPS optimized so much that they are at least on parity with bots for mega factories. And apart from that...haven't you noticed salami tactics to slowly buff belts have already started ?

It's not about UPS though. Wanting higher UPS in big factories is why bots are preferred over belts - as there is no collision involved - they just go straight from A to B - and can then dump an infinite amount of resource into a chest at the same time, you just can not load chests like that if you go the belt and stack inserter route.

From Friday Facts #225 (written by kovarex)

To be able to compare numbers, I made test setup. I created 4 fully compressed express belts and put them against a fully satisfied lane of roboports (also 4 tiles wide). To be able to measure the throughput easily, I just modded the furnaces to be very fast and smelted the result, iron for belts and copper for robots. This is the setup where belts are as strong as they can possibly be are in the bots versus belts balance. In majority cases, bots are going to be relatively much stronger compared to belts.



The result is, that bots did 16.4k per minute while belts only 9.6.



With this in mind, we can safely say, that robots are at least 2 times stronger then express belts, but in real factories, it is much more as belts need lot of other parts and are rarely used as ideally as robots would be, so my private guess would be that robots are currently around 5+ times stronger compared to belts.

---

It's not about UPS (though of course that is important!). Also I am fairly sure that all the recent updates to belts was to address the UPS concerns or compression problems, belts themselves have not been buffed, they have changed nothing to address the underlying belt vs bot debate. I'm quite sure this is the case, though it's quite hard to keep up to date considering how fast the devs push out updates... Correct me if I have missed something!

[eradicator]

@Ghoulish:
That test by kovarex was "transfer speed per tile". I don't see how belts can ever be as fast as (current!) bots are, as bots effectively have infinite extra space by using non-colliding air transfer, and as you said, making bots worse (even if there was a good way to do it - i'm not aware of any) is not really desirable. But for "transfer speed per cpu cycle" belts are now supposedly as good if not slightly better as bots, though ofc you will need more space to place all the belts.
As for belt buffs, the filter splitter update was a straight forward belt buff, and i was just speculating that more subtle updates like that might be coming.
_____________________

Also yay! A built-in recipe viewer you say? Finally. The current mods all have horrible interfaces. (And my own attempt overtaxed the engine :P). Even better that it's mentioned as just the beginning to a built-in wiki.

[xng]

So.. 1.0 in Summer 2019 maybe?

I would think that's reasonable considering history. Each minor update takes approximately 9 months to develop and 3-4 months to stabilize. So, if 0.18 is only polish and stabilization + a few months of playing around with optimizing for fun, it could be possible. I hope.

[yohannc]

Great news, can't wait for the new GUI !
But nothing about sound modifications ?
I mean, actually, sound could be wayyyy better.
You should work on the sound engine, for example, we should hear explosions or shot from far distance (less volume, less treble, low pitched, echoes, latency, etc). Sound ambiance based on locations (as already said on others topics, leaf sound when in forest).
I don't think all theses changes take too much time. I know, it take time, but it won't make regression on the rest of the game, it's not like changing belt engine (wich take time and wich is not visible by 100% of the players).
And why not add weather effects like rain/thunder ? (fog could be cool, but could be annoying or should not be present in polluted area, where we are building).
Yes it's like a "icing on the cake", but it could add a lot more immersion.
It's also a better impression for new players in their first 30 minutes of playing. They would have less the feeling that the game is just a sandbox, an "empty flat world where you build puzzles.".

[bobucles]

That test by kovarex was "transfer speed per tile". I don't see how belts can ever be as fast as (current!) bots are, as bots effectively have infinite extra space by using non-colliding air transfer, and as you said, making bots worse (even if there was a good way to do it - i'm not aware of any) is not really desirable. But for "transfer speed per cpu cycle" belts are now supposedly as good if not slightly better as bots, though ofc you will need more space to place all the belts.

Bots pay a minimum fixed amount of energy to cross tiles, and they take time to recharge. Fortunately it all boils down to very simple math. A very nice poster has already worked it out and found some nice hard numbers:
- Belts have 40 item moving capacity per tile.
- The charging power of roboports(with highly upgraded bots) gives about 81 item moving capacity per tile.
He was even kind enough to test out those values in game! The Factorio Creative mod was very nice in providing the item spawning, loaders that never throttled, and super fast furnaces (60/sec) that never got saturated.
The test setup isn't a perfect straight line for the bots, but it gives a decent comparison of 4 lanes of belts vs. a solid line of roboports (it's 4 tiles thick). The results definitely bode well for the raw theory.

factorio test2.jpg (917.28 KiB) Viewed 6839 times

factorio test1.jpg (83.2 KiB) Viewed 6839 times

The end result that bots managed to move about 100% more items than the belts. It's one thing to be close, but this is nearly spot on! Kudos to whoever figured it out.

But wait, you cheated! The bot bus is a fraction of tiles shorter than the belt bus! Hacker! FAKE NEWS

Well, golly gee I sure am busted. It's not like the distance is pretty irrelevant or anything as long as the roboport "bus" is a solid line.

factorio test 3.jpg (555.99 KiB) Viewed 6831 times

factorio test 4.jpg (189.03 KiB) Viewed 6831 times

Oh no! The bots somehow ended up even more powerful than before! What is this witchcraft?

[Omnifarious]

they have given us the tools.

Both of those could be even simpler with the filtering balancers introduced in 0.16.

[Serenity]

What about map generation, so we can get proper oceans and continents again and not just an endless mess of lakes with tiny land bridges between them?

[Rinin]

So, no any plans about trains?

[eradicator]

@bobucles:
If there was any point you wanted to make by quoting me without my name then you forgot to make it.

[ske]

So, no any plans about trains?

I really would welcome some kind of train tunnel as that would make the rail network much more powerful and prevent a lot of situations where jams might occur.