Factorio Forums

I've been seriously saddened by the changes to belt compression. Before, we had three ways to compress a belt: splitters, undergrounds, and side-loading. Now we have only one: splitters. I feel like this is limiting our options very severely: stifling creativity and hugely restricting the number of solutions. (Further limiting the "best" solutions that many people currently use).

While the effect of undergrounds was unintentional, it was a creative and interesting way to solve problems. It worked in situations where splitters might not fit, etc. It also made logical sense, since the implied diagonal length of belt that led from below to the surface had more space, and therefore inserters could place items there.

I'd love to see underground belts have their compression functionality back, or perhaps a separate item, a "belt compressor" -- perhaps a 1x1 or 1x2 belt-related item that would allow items to be fed in from the top by inserters, and it would compress the output, basically giving the same functionality as the old underground.

(I've heard that side-loading may come back, and I'm very much looking forward to that!)

I agree with KatherineOfSky above that only allowing splitters to compress belts is to limiting. This is a game of options, everyone does things differently and that should be encouraged, not limited.

The undergrounds not compressing anymore seems fair enough because it felt pretty silly, but I do think sideloading compression is a must.
I would presonally also prefer inserters to be able to compress on normal belt and also onto undergrounds (since the inserter is the source of the compression and not the underground belt).
And something that I would prefer to compress is direct export buildings like mines (cant think of the top of my head if there are any other buildings).

My 2 cents

Well, the underground thing i can understand. If it hasn´t been for some youtubers, i wouldnt even have thought about it.
Ok, exploit gone, its fine, i never liked it anyway.
Even the inserters not being able i can live with. As someone already explained, it not exactly an inserter job to place something at an exact point.

But sideloading from belt to belt not working is just stupid, a belt should be a able to throw bulk goods onto another one.
Even when i used inserters to unload Burner Miners (/facepalm) back in the noob days, even I figured out how to prioritize belts, with sideloading the less important one. Its intuitive, unlike the uderground exploit

Its still possible, but only if you never go Blue Belt, because you need just one bit faster than your standard belt to make it possible again. Luckily i only go blue when transfering to Mega and thats when bots start to really work.
The standard 2flask/sec which brings me up to infinite researches is easily done with reds.

I can live with smelting colums needing a splitter somewhere in their middle, but not being able to prioritize one belt over another with sideloading, i dont like. A splitter array that prioritizes one over another might be possible, but it would be a spagetthi monster i guess, I dont even wanna think about how to build one. So i wont ever know if its possible

Having read all replies, it seems most people are either for: letting everything compress, or at least letting sideloading compress again.

Personally I think, going from 4 methods (splitters, inserting on undergrounds, sideloading, and circuit timed insertion) to compress a belt to only splitters is kinda sad.
I filed it as a bug right when I found out, and for me it takes out half the fun of the game. I like spaghetti, use mixed belts a lot, and make combinator contraptions for fun, most of my gameplay is tinkering with contraptions. The sideloading is an essential trick, and it's used for many things, from giving priority to certain belts, or lanes, to compress belts, etc.

Insertion on an underground belt is kind of a weird issue, I've never found that trick particularly engaging, but it was necessary for belt-based mining. Smelting you can get around it pretty much, but for mining you needed it to densely pack miners on large patches.

Although I wouldn't choose to make everything compress, since I do think compression was a fun thing to discover in Factorio, I do think miners and sideloading should both compress.

More importantly, I'd just like to see all parts of the belt mechanics behave more consistently, especially spltiters. I would seriously consider removing the "try to push X items back into a lane that was previously backed up" mechanic. It's an arbitrary mechanic, tied to a magic value, defeating the entire purpose of a splitter (to evenly split when not backed up, even if it was previously backed up). It's a real burden to work around with in any scenario where you want precise control with combinators.

PLEASE make the belt compression work again. Miners, splitters and side loading should work in my opinion.

I'd just like to say that belt compression bothers me not at all. I may be the only person with this opinion on the planet but there you go.

I'd hate to see UPS performance degrade if steps were taken to bring compression back.

Sorry World..

I have never particularly needed my factories to be frame perfect, but recognize the use in compression, particularly for mining. Mining a patch feels weird and uneven without a good way to compress things.

That being said, I have no objections to seeing compression go. I agree that underneathies should not be able to compress belts. I am not sure abput sideloading. I feel like it would be neat if sodeloading compressed belts, but not at the cost of performance.

I'm OK with removing the underground trick but side loading should compress the belt also minors are basically side loading as well so they should too. Not sure about inserters maybe fast inserters can compress but not regular or have a research for each belt speed for an inserter to fit the product in the tiny space. Like I said I'm OK with removing the underground trick but not happy with no alternative other than radically altering our smelting columns and arrays as well as adding more spitters at ore patches. Loved everything else in .16 I didn't think the game could get any better but it did but sad about compression.

I did not play that long on any of my factories to reach a megabase state. I didn't even replaced everything with bots until now.
I just think with the underground trick/bug gone there should be a more compact item then a splitter to compress items on a belt.
This 1x1 compactor item seems like a good idea. We still need to change our commonly used Blueprints because one more part is needed at some places.

The sideloading will be fixed with the compator aswell because it does not need more space and will achiev what is wanted.

Removal of compression really ruined some of the fun with the game, the devs should see this avalanche of comments about it as being a really important issue for the players.

AcolyteOfRocket wrote:I'd just like to say that belt compression bothers me not at all. I may be the only person with this opinion on the planet but there you go.

I'd hate to see UPS performance degrade if steps were taken to bring compression back.

Sorry World..

I agree fully. Your not alone!

brunzenstein wrote:

AcolyteOfRocket wrote:I'd just like to say that belt compression bothers me not at all. I may be the only person with this opinion on the planet but there you go.

I'd hate to see UPS performance degrade if steps were taken to bring compression back.

Sorry World..

I agree fully. Your not alone!

I am pretty sure that while sideloading might have an impact on the belt optimizations it would still be minor as sideloaded belts are pretty rare compared to long chains of belts which are the culprit of UPS drops in megabases. And as sideloading algorithm was stipped to streamline belt optimizations readding it will most likely add some IF case causing some increase in CPU usage but nowhere near 0.15 levels.

Optimisations are good when sacrifices are appropriate.

Kunstduenger wrote:

Jap2.0 wrote:My opinion on belt compression is that splitters and sideloading should work, but not inserters or underground belts.

I completely agree, though I'm also willing to trade sideloading compression for performance, since there are workarounds.

I get the feeling that this design is intentionally overcomplicated:


Here is a design i used even before 0.16 because I never was a fan of the underground belt compression trick

To be honest this method of side-weaving half of the belt on the bottom picture feels even more complicated to me. Sure, you use less inserters, but you have to use even more underground belts than with underground compression trick to make it work and it gets even more tricky with shorter yellow underground belts and small electric poles. Either way, it's surely not intuitive at the early stage of the game when this issue matters the most and for new players. Plus, at least for me it's less aesthetic.

Side-loading should DEFINITELY stay in! I don't understand why it would have to be removed. It's neither a bug, nor illogical that side-loading should compress the belt (I would expect it in a real world situation).

Please keep side loading in. Undergrounds... okay, I understand... although I liked the option to be there. Like KatherineOfSky, I like the idea of a 1x2 sized "Belt Compressor" item you could place on a belt to duplicate the functionality of having back-to-back undergrounds, compressing. But if this has to go... fine.

But side-loading. Side-loading NEEDS to stay in. There's no reason that side loading shouldn't work. If there's a hole, and a belt is providing constant pressure to try fitting things onto another belt, than ANY hole should be filled. It just makes sense!

For inserters... I do agree that from a newbie's perspective, having any and all inserters automatically compress makes sense. But it's more interesting to have an engineering problem to solve, and learning to compress belts (for example, via SIDE LOADING!!!!) is an interesting and fun problem to tackle. Although, again, a 1x2 belt compressor item to duplicate the old back-to-back undergrounds would be nice! But barring that... most inserters shouldn't automatically compress. Most. But STACK inserters should. They are expensive, resource intensive, and they way they work (dropping a constant stream of items onto a belt) really suggests that they should be able to fill those gaps. It's just good gameplay.

I feel strongly about compression.

About Belt Compression: Are you aware of this bad boy: https://gaming.stackexchange.com/questi ... rting-work

How much of that design is intentional?

AcolyteOfRocket wrote:I'd just like to say that belt compression bothers me not at all. I may be the only person with this opinion on the planet but there you go.

I'd hate to see UPS performance degrade if steps were taken to bring compression back.

Sorry World..

AcolyteOfRocket wrote:I'd hate to see UPS performance degrade if steps were taken to bring compression back.

Actually, if I remember older FFF about belt optimalisation, compressed belts are supposed to be more UPS-friendly than non-compressed ones. Unless something changed, memory was supposed to store distances between each itemS chain on belts, not actual item count+position. Or something like that. So, less spaces between items = less UPS drain.

Belt compression can still be done with the splitter.
Anyway I always plan for +10% so if there is little less on the belt it is still fine.
Good enough for me.

Just Please, Please, Please
If adjusting belt mechanic / compression will decrease performance
Don't do it.

Performance gain's are unbelievable, what was unplayable in .15 is now perfectly fine.

Well lack of compression means that in belt based bases we will probably have to use more belts and splitters to achieve the same throughput. So in that respect adding compression might actually help performance of belt based factories. (Yes this depends on how much extra processing is involved to add the compression). For bot based factories, the adding compression shouldn't matter, because with no belts, it shouldn't cost the bot based factories anything