Friday Facts #297 - New resource icons

[eradicator]

This is just an OCD thing I think, but ever since I started playing Factorio, it always bothered me that the colors of Iron Ore & Copper Ore weren't as accurate as they're IRL. As I said, noting major, just kinda OCD related, tho maybe if anyone has the time for it?

https://mods.factorio.com/mod/xander-mod-th

[Agamemnon]

These new ore icons look really good on the belt.
This is not a change I needed, but a change I appreciate.

While on the subject of nitpicky optics changes... Can you make the items on the belts rotate properly, too? With some things (namely flat things like concrete, plates, circuit boards) it becomes noticeable they also pop over each other (I mean reverse z drawing order) when going around corners.

Edit: Clarification: It wouldn't make sense to do this for everything. Only stuff that is perceived as transported while lying flat on the belt. Standing-upright-stuff is not affected that way.

[Unknow0059]

the new icons look like the stuff of dreams. I had been wondering all this time if it just wasn't possible.

[plepper1]

I love the new ores, except the closeness of the coal/iron. Coal used to be more black, Iron used to be more blue. Now they both average out to a greyish color without the prior contrast - They don't "pop" as being different on the same belt.
They are still different colors, but only when I look - not from a quick skim as I would fly by in map view.

Edit: The screenshots looked weird. In the game, they look great!

[ignatio]

Thanks for the inserter fixes. Inserter orientation still makes a difference in some cases though. I notice it in particular for long-handed and burner inserters moving items from chest to chest or chest to belt:



In these cases horisontal inserters are a little more than 2% faster than vertical ones. This is a regression since 0.17.42, and I only see it for long-handed and burner inserters. (Burner inserters are also affected by refueling, but I've eliminated that error source.) I think this could be worth looking into, as these cases used to be entirely consistent before.

Here's another regression:

0.17.42: 0.17.45:

Although inserters generally are slightly quicker in 0.17.45, in this particular case it gets out of whack and it looks like it only catches every 2nd item, resulting in a 43% slowdown compared to 0.17.42. This particular case is limited to sideloaded underground entrances, so it's perhaps not terribly important. But there might be more regressions that I haven't noticed yet.

How do I know these things? Well, I've taken an interest in exploring inserter throughputs for my calculator spreadsheet, and I eventually ended up doing a fairly comprehensive test of inserter setups (the above are three out of some 1600+ test rigs). All measurements for 0.17.45 are in this spreadsheet. For comparison here are the 0.17.42 measurements.

Unrelated to above, I wonder if the issue with inserters not picking items up is fixed now!?

It's better but not gone. There are only a few cases that I'm aware of: Basic and burner inserters taking items from the outer lane of an express belt turn, and even then it works more than half the time.

There's also this one:



Guess what the difference is between the two clips? It's not that I've cut it down to a few frames to the right to make it look stuck - it actually is. Rather, the difference is that the left belt lane has been released 1 tick later in the right case. This is so rare I regard it a curiosity rather than a problem - I've only seen it in this setup with an underground entrance being sideloaded in one lane exactly like that, and even then it happens only for a handful of belt timings. It goes to show though that tiny one tick timing differences can have quite an impact on inserter throughput.

[Twinsen]

Thanks for the detailed post. I'll look into it.

Like I said, some situations might end up being slower because of how things are timed, so one slowdown might be fine compared to the many other speedups. But I'll take a look at all of these and see what's the best change. There's a situation where the inserter still can't pick up one lane.

Inserters will never be very consistent as long as we are doing the hand movement simulation and item tracking. But that's something we don't plan to let go of.

EDIT: the rotational asymmetry for long hand inserters(and possibly others) is fixed in Version: 0.17.46.

[Ormy]

New ores look great, can't wait to use them. I'm using the "better ore icons" mod (which looks quite similar to the new icons?) at the moment but will switch to the new vanilla ones when they are implemented.

[V453000]

New ores look great, can't wait to use them. I'm using the "better ore icons" mod (which looks quite similar to the new icons?) at the moment but will switch to the new vanilla ones when they are implemented.

Yeah there's a little story I omitted for the FFF, basically when we tried to make the icons for 0.15 I accidentally pushed them and they got released, and by some people reused. They are the cut out versions from the resource entity spritesheets as mentioned in the FFF. That's why they are similar, they come from the same Blender materials and the same source of random geometry.

[ignatio]

Thanks for the detailed post. I'll look into it.

Cool, thanks!

Like I said, some situations might end up being slower because of how things are timed, so one slowdown might be fine compared to the many other speedups.

Sure, there aren't a lot of big slowdowns like that afaics. I was more concerned about some previously perfectly consistent cases getting inconsistent, and that there still are cases when inserter direction makes a difference. I'm hoping the latter can be fixed entirely.

But I'll take a look at all of these and see what's the best change. There's a situation where the inserter still can't pick up one lane.

In that case, as well as in the example in the FFF, I think the issue is simply that the speed of express belts cause items to have different phase shifts on the successive belt tiles:

An express belt advances items 24 positions per tick. If an item starts at position 0 on a belt tile it will per tick advance to position 24, 48, 72 etc. There are 256 positions on a tile, so when it enters the next tile it'll start on position 11*24-256=8, i.e. shifted 8 positions from the first tile. So the item will never be on exactly the same positions on those two tiles. On the next tile it will start on position 16, and on the fourth tile it is back to position 0.

This phase shift between 0, 8 and 16 means that inserters can behave differently depending on where they are placed, but they will repeat the same arm cycles when placed on even multiples of 3 tiles. Only the express belt has this behaviour, as the other two with their speed of 0 and 16 positions per tick divide 256 evenly.

In the case you pointed to, the difference is that sideloading and dropping items directly on the belt make them start at different positions on the horisontal belt. They may be an even multiple of 8 apart, in which case the perfect pick-up can be restored by moving the picking inserter one or two tiles along the belt. Or it may not be, and then that might not help.

I doubt this can be fully fixed just by making inserters pick up faster - there will always be interference between the hand movement and the different item positions on the belt that makes the inserter arm cycles play out differently. Imo this is one of the finer details of the game mechanics that makes it interesting - there's no bug to fix, only understanding to be had.

If one is so inclined, it is possible to calculate the belt positions precisely and plan the belt accordingly (the belt physics page has a table with exact lengths).

Inserters will never be very consistent as long as we are doing the hand movement simulation and item tracking. But that's something we don't plan to let go of.

Glad to hear that, I think the complex and sometimes chaotic behaviour of the inserters bring a lot of depth to the game. Wouldn't have it any other way.

[Azzinoth]

I think randomizing the resource icons is a great idea, but to be honest it still looks very repetitive to me.

The reason is that one ore item is always represented by 4 ore chunks of almost equal size and then placed at the corners of a square. In the last picture it is most visible for copper and stone. I'm pretty sure if i would make a spatial fourier transform of these belts I would get very localized spikes. What you need to do is randomize the spatial frequency a bit more.

To do that i would make the first icon out of 4 medium sized ore chunks, the next icon out of one very big chunk, the next icon out of one bigger chunk and one smaller chunk etc. That would give you more variation of sizes and distances between visible chunks, even tough the icons all have the same distance from each other.

I personally would also sprinkle some dirt over the chunks and the space between them. They are just coming out of the ground and look too clean for my taste...

[Spatial]

Looks really good. Not something I thought could be improved so much!

[catma]

seems like the next step to help the devs is to make a color lamp / numbers display type base that crunches all the numbers and displays them in a simple array. Every inserter in every situation you can imagine can be automagically tested for rebellion against engineering...

...and test a few extreme inserters from mods, because a lot of little local fixes on the vanilla inserters means the underlying math is still broken in a way that is sometimes spectacular and unexpectable in mods. Modders will also want to have the all the fixes' source code for this reason!!!!

[NelsonSKA]

I really like the new icons for the ores, It's a little change that makes see more natural the raw materials. It won't look only like a icon!
I know that It's in WIP, but only are there 2 sprite for ore type? or will have more variations?
Greetings

[catma]

I also think SOME people can't stand the randomized ores and I know 2 who will scream bloody murder if they don't have a checkbox to turn the randomness off. Having to resort to a mod for something that is so VISCERALLY VERY WRONG will not do for either of the 2 people I talked to.

I'm pretty sure it's a minority opinion, but I've only talked to 7 people about factorio in RL, so from my sample it's not a completely insignificant percentage (also neither of the 2 complain about factorio for anything else).

[Mxdanger]

That video at the end looks amazing! Like a real life ore belt - can't wait to see a huge smelting set up with these textures...

Exactly! It just blew me away.

[Goose]

There's a lot on your plate already, but is there opportunity for a few other items to have alternate belt icons? Uranium 238 and Uranium 235 have so much potential for this kind of stuff, especially with belted Koverex enrichment designs!

[Lubricus]

I also think SOME people can't stand the randomized ores and I know 2 who will scream bloody murder if they don't have a checkbox to turn the randomness off. Having to resort to a mod for something that is so VISCERALLY VERY WRONG will not do for either of the 2 people I talked to.

I'm pretty sure it's a minority opinion, but I've only talked to 7 people about factorio in RL, so from my sample it's not a completely insignificant percentage (also neither of the 2 complain about factorio for anything else).

Relax someone will make a mod.

[Goose]

I also think SOME people can't stand the randomized ores and I know 2 who will scream bloody murder if they don't have a checkbox to turn the randomness off. Having to resort to a mod for something that is so VISCERALLY VERY WRONG will not do for either of the 2 people I talked to.

I'm pretty sure it's a minority opinion, but I've only talked to 7 people about factorio in RL, so from my sample it's not a completely insignificant percentage (also neither of the 2 complain about factorio for anything else).

Relax someone will make a mod.

I could if you want

[eradicator]

Having to resort to a mod for something that is so VISCERALLY VERY WRONG will not do for either of the 2 people I talked to.

If they're important to you you should start your brainwashing efforts now. So they can embrace the fact that modded factorio is the only true way. Before it's too late. All hail Yuu Peyes the Alaccelerating!

[Cadde]

On one hand, it's nice to have options in game. On the other, you'd definitely end up with waaay to many options if everything was customizable.
Just look at OTTD, especially modded OTTD with their "advanced settings" dialog.

Personally i love all the options OTTD presents to the user at runtime. But imagine how users would quickly ignore all the things in there and make a fuss on the forums about how the game is too complex and they want their money back.
I dunno, it's a balance act and some things really do belong in mods.
Which by the way OTTD doesn't really support in that respect. And "modding" for OTTD is hella complex compared to most other games. Writing a NEWGRF in itself is like learning a new programming language, especially if you are intent on writing them with "assembly instructions" rather than the high level alternatives. And making changes to the source is just a small hop away from NEWGRF.

I asked for a setting to remove/change the timeout of temporary train stops and a dev responded he added that to a prototype. Again, it would have been nice to have it in the game menu but it's a setting one really only changes once and it's a very specific request for a very limited subset of users.
As such, it belongs in mods.
The new resource icons fall in that category, a very limited subset of users will want the old graphics. The majority will either love it or have never known there was a worse looking alternative. Imagine them installing a "classic" mod, they would complain about the resource icons looking god awful.