PyBlock (alpha)

pyanodon's mods are here

Moderator: pyanodon

Post Reply
kingarthur
Smart Inserter
Smart Inserter
Posts: 1459
Joined: Sun Jun 15, 2014 11:39 am
Contact:

Re: PyBlock (alpha)

Post by kingarthur »

reapersms wrote:
Tue Feb 12, 2019 4:49 pm
Last time I did base pymods, it was before raw ores. The pain points then were finding the ores in the first place, and long build times for most buildings.

I recall getting the basics of circuit production started a lot sooner, but tended to bottleneck on wood and steel for scaling up urea production.

Contrasting with a seablock run I took to the middle of the spacex endgame:

I'll have trains a lot sooner in pyblock. Belt spaghetti can work for angelbob, but there are too many byproducts and loops to wrap my brain around here.

Power is not an issue in pyblock, at all. Biggest panic I've had so far was a hiccup or two when coal gas shut down because the coke was backed up. Seablocks power loop doesn't really stop being a gigantic headache until you're on the cusp of nuclear. On py, that will likely never, ever be an issue, if only because dropping a couple of fawogae plantations and some HPFs to convert them to coke is ridiculously energy positive. There's also so much coke and coal in the main production loops that one would never be in the seablock situation of needing to remember to always be foraging cellulose fiber by hand, because that 1.1 MW it's worth stays a significant portion of your power production until hour 50...

Ore supply seems like it will be more straightforward in pyblock, late game seablock I had a lot of issues getting sludge pumped around
while i personally haven't payed seablock ive seen a few lets plays and so so many Reddit posts on it. i defiantly wanted to avoid recreating that whole power is constantly on the verge of failure that seablock seems to have. if you have to handcraft nonstop just to keep going doesnt feel like itd be fun for every long

reapersms
Long Handed Inserter
Long Handed Inserter
Posts: 69
Joined: Sun Jun 17, 2018 4:55 am
Contact:

Re: PyBlock (alpha)

Post by reapersms »

Things there get a lot better once you find a tree or garden bush -- those give you arboretums for direct raw wood production, and farm products that can be processed into cellulose and oil. Trees are easier, but you only get a set number of arboretums per tree you find, whereas farm capacity has a net positive feedback loop in the seed generation.

Before then it's a nightmare of algae farms, with cycle time and production similar to ralesia or kicalk (albeit without the inputs), but the power usage of a soil generator. The processing steps along the way from algae -> fiber -> wood bricks -> charcoal/carbon -> co2 was a bit messy as well, I recall the first carbon recipe being slightly net negative on power.

That said, the scale needed was quaint in comparison to most pymods sequences. I think a dozen arboretums and 20 farms produced enough wood and liquid fuel to generate about 600 MW or so before nuclear kicked in. The usual issue at that point was general demand for charcoal & carbon for either CO2 production or smelting inputs. Not too long after that I had a working blueprint I could stamp down for a 480 MW 2x2 nuke plant.

Pymods in contrast looks like it will involve a lot more space, and the handy bot mall I had is going to be a lot larger and messier this time around -- automated factories don't pack very well, and angelbobs building ingredients tended to be more along the lines of "previous tier plus 2-3 plates, pipes, or gears". That will be mitigated a bit by having 2 fewer tiers of belts/unders/splitters/stackers/loaders, and one fewer tier of inserters.

The sheer scale of circuit production is a bit intimidating, but I'm hoping the waste water -> urea recipe that got added not too long ago can do a good number on that.

kingarthur
Smart Inserter
Smart Inserter
Posts: 1459
Joined: Sun Jun 15, 2014 11:39 am
Contact:

Re: PyBlock (alpha)

Post by kingarthur »

reapersms wrote:
Tue Feb 12, 2019 6:26 pm

Pymods in contrast looks like it will involve a lot more space, and the handy bot mall I had is going to be a lot larger and messier this time around -- automated factories don't pack very well, and angelbobs building ingredients tended to be more along the lines of "previous tier plus 2-3 plates, pipes, or gears". That will be mitigated a bit by having 2 fewer tiers of belts/unders/splitters/stackers/loaders, and one fewer tier of inserters.

The sheer scale of circuit production is a bit intimidating, but I'm hoping the waste water -> urea recipe that got added not too long ago can do a good number on that.
ya but now you've got to deal with way more individual buildings need a lot of different stuff compared to angels that i dont think the fewer tiers of belt and things really offset the difference but idk ive just started a full pyblock play through so ill find out soon enough.

Szentigrade
Fast Inserter
Fast Inserter
Posts: 107
Joined: Mon Mar 28, 2016 7:27 am
Contact:

Re: PyBlock (alpha)

Post by Szentigrade »

got an eta of an update to 0.17

kingarthur
Smart Inserter
Smart Inserter
Posts: 1459
Joined: Sun Jun 15, 2014 11:39 am
Contact:

Re: PyBlock (alpha)

Post by kingarthur »

Szentigrade wrote:
Tue Mar 12, 2019 1:39 am
got an eta of an update to 0.17
couple of days. just managed to find a very hack solution to map gen in .17. got to tweak a few more things to take advantage of new things like fluid energy sources and then do a bit of balance testing on the changed fuel valves. if you want to try it before the update is out the most recent work is available on my github. https://github.com/kingarthur91/PyBlock_0.0.5. its functional though balance hasn't been tested yet.

Szentigrade
Fast Inserter
Fast Inserter
Posts: 107
Joined: Mon Mar 28, 2016 7:27 am
Contact:

Re: PyBlock (alpha)

Post by Szentigrade »

kingarthur wrote:
Tue Mar 12, 2019 1:45 am
Szentigrade wrote:
Tue Mar 12, 2019 1:39 am
got an eta of an update to 0.17
couple of days. just managed to find a very hack solution to map gen in .17. got to tweak a few more things to take advantage of new things like fluid energy sources and then do a bit of balance testing on the changed fuel valves. if you want to try it before the update is out the most recent work is available on my github. https://github.com/kingarthur91/PyBlock_0.0.5. its functional though balance hasn't been tested yet.
always keen to test, watching your pyblock playthrough on youtube atm.. makes me want to play it

kingarthur
Smart Inserter
Smart Inserter
Posts: 1459
Joined: Sun Jun 15, 2014 11:39 am
Contact:

Re: PyBlock (alpha)

Post by kingarthur »

Szentigrade wrote:
Tue Mar 12, 2019 1:46 am

always keen to test, watching your pyblock playthrough on youtube atm.. makes me want to play it
Well your free to try it. But starting out I'd recommend waiting a day I dont think you can play a new game as theres not enough fuel available to not cuase a dead lock but I've been awake to long. Like 23 hours atm. I'm going to get some sleep and get back on it when I wake up. Sometime Tuesday afternoon I should have it updated completely and ready for a new game and to resume recording.

reapersms
Long Handed Inserter
Long Handed Inserter
Posts: 69
Joined: Sun Jun 17, 2018 4:55 am
Contact:

Re: PyBlock (alpha)

Post by reapersms »

Still on .16, a balance observation from helmod fiddling:

2 circuit-2/sec needs 12.81 phosphine gas/s for light N silicon
the only source of phosphine gas is wodd + phosphorous acid
the only source of phosphorous acid is powdered phosphate rock

if you don't find one of the skeletons or rocks out in the endless ocean, you can synthesize phosphate rock from collagen, but that needs skin, lots of it, ~160/s to be specific.

That much skin needs 142 autofac 1's to slaughter carcasses, and 640 zipir reefs to breed them.

sadly, in the area I've scanned so far, I've only been blessed with iron, nexelit, and 2 uranium deposits.

an alternative source of either phosphate rock or a more efficient recipe for collagen would get things back to something reasonable... though for now I suppose I shall have to do some exploratory landfilling

kingarthur
Smart Inserter
Smart Inserter
Posts: 1459
Joined: Sun Jun 15, 2014 11:39 am
Contact:

Re: PyBlock (alpha)

Post by kingarthur »

reapersms wrote:
Wed Mar 13, 2019 5:57 am
Still on .16, a balance observation from helmod fiddling:

2 circuit-2/sec needs 12.81 phosphine gas/s for light N silicon
the only source of phosphine gas is wodd + phosphorous acid
the only source of phosphorous acid is powdered phosphate rock

if you don't find one of the skeletons or rocks out in the endless ocean, you can synthesize phosphate rock from collagen, but that needs skin, lots of it, ~160/s to be specific.

That much skin needs 142 autofac 1's to slaughter carcasses, and 640 zipir reefs to breed them.

sadly, in the area I've scanned so far, I've only been blessed with iron, nexelit, and 2 uranium deposits.

an alternative source of either phosphate rock or a more efficient recipe for collagen would get things back to something reasonable... though for now I suppose I shall have to do some exploratory landfilling
Alright. I thought phosphate rock was a byproduct of some recipe or other that made a lot of it. Maybe I meant to add it and forgot.

I'll look into it asap but my pc sprung a leak so it's in pieces. itll be a few days before I can really look

reapersms
Long Handed Inserter
Long Handed Inserter
Posts: 69
Joined: Sun Jun 17, 2018 4:55 am
Contact:

Re: PyBlock (alpha)

Post by reapersms »

Given the ridiculous number of other ingredients that need making first, no rush there.

FNEI might be lying to me, but the only recipies I see that give phosphate rock are the bone meal + collagen + sulfuric acid one, and the phosphate mine/big phosphate mine ones

I may hack in a particle accelerator transmutation recipe for it, either 10 sulfur + receiver -> 10 phosphate in 15s, or 10 sulfur + receiver -> 20 phosphate in 10s or 15s probably, depending on whether to model it after iron->chromium, or sand->coal, or copper->nickel. leaning towards sand->coal, since the others are transmuting finished plates...

edit: seems a reasonable approach, would still have to bootstrap via collagen to get the accelerator, but would be able to use that to scale up production. may still need another source to keep molten stainless and phosphate glass running in the far, far future

reapersms
Long Handed Inserter
Long Handed Inserter
Posts: 69
Joined: Sun Jun 17, 2018 4:55 am
Contact:

Re: PyBlock (alpha)

Post by reapersms »

Got circuit 2's produced at around 125h. It's not completely automated yet, still have to move wafers and circuit 1's around to it.

The first things I used them on were the distillation columns cracking wood and coal, as the syngas supply had dried up. That didn't go exactly as planned, as the wood production couldn't keep up anymore. That also gutted tar production, so the quenching towers don't produce nearly as much saline, which slows down the zipir->urea cycle, which slows down collagen->phosphate, which is strongly limiting wafer production. I found a salt mine, but the mining there is still not quite enough.

I haven't tried out the nucleosynthesis recipe for it I added yet (no rayon yet). I think my route out of this mess is getting activated carbon running properly again, supplementing syngas production with the flue gas filtering, and finishing off the logs->wood->coal pipeline upgrades, and changing out all of the combustion turbines for mk2's.

That ought to reduce the syngas consumption by the power generation enough for the borers to kick back on, but the train system is up and running now at least.

reapersms
Long Handed Inserter
Long Handed Inserter
Posts: 69
Joined: Sun Jun 17, 2018 4:55 am
Contact:

Re: PyBlock (alpha)

Post by reapersms »

So everything is a mess still. Lubricant is stalled in two places, one on aromatics, the other on nickel ore. The zinc borer is further down the line from the rest so it gets almost nothing.

I got a trickle feed of urea into the collagen plant so that should unstick wafer production eventually. Circuits now are usually bottlenecked on ferrites needing nichrome and zinc.

Could really use an oil patch on an island, but that is probably out until pyph lands, and patching that in will be a trick.

kingarthur
Smart Inserter
Smart Inserter
Posts: 1459
Joined: Sun Jun 15, 2014 11:39 am
Contact:

Re: PyBlock (alpha)

Post by kingarthur »

reapersms wrote:
Sun Mar 31, 2019 1:32 am
Could really use an oil patch on an island, but that is probably out until pyph lands, and patching that in will be a trick.
That's the plan I'll be patching that in and getting the .17 version done soonish.

How are the rocks working so far?

reapersms
Long Handed Inserter
Long Handed Inserter
Posts: 69
Joined: Sun Jun 17, 2018 4:55 am
Contact:

Re: PyBlock (alpha)

Post by reapersms »

Only have two hooked up so far, an early iron and a salt mine. Others in range are not too useful (nexelit, uranium), redundant (another iron), or nice but out of tech range still (nickel). Neither the iron or salt produces gas enough to really kick in yet. Iron serves as a trickle source to keep one steel setup running, salt generates some fill in capacity saline water.

I tossed phosphate and raw coal borer recipes into the local tweak, along with a set that use refayngas for 3x production, but haven't set any up yet.

Medium term route seems to be restructuring the factory around the rail blocks to consolidate some things, and untangle some egregious pipe spaghetti. The middle 3/4 of the factory is a chaotic abyss of insanity.

On raw coal borers, treating it like iron, 15 raw for coal gas is a big net loss, 30 for syngas just about breaks even even with the lubricant, and 45 for reggae would be quite positive. Numbers change a bit if you can go past coarse coal. Base one could be a win if you have a lot of wood to coal gas going on, but the sources of crude probably kill it.

Right now the factory peaks around 5k/min lube, and full production needs about 50. Right now I have a couple desulf units making heavy from aromatics crude, and one fts making it direct from coal gas. The first two are constrained on aromatics, the other on nickel ore.

Power is running on hydrogen with a bit of assistance from acetylene, coal slurry, and refgas. That did solve the coal gas and syngas demand issue at least. Needs another 4 or 5 turbine 2s, but circuit 2s have been pegged for wood and diamond research so far, and then the zinc really ran dry. Excess coal dust into kimberlite should keep things from clogging up there for a while.

Nickel and zinc feel a lot harder to acquire, since there isn't a ezmode quench recipe for them like the first 6. I could pave the factory with chromium I think.

kingarthur
Smart Inserter
Smart Inserter
Posts: 1459
Joined: Sun Jun 15, 2014 11:39 am
Contact:

Re: PyBlock (alpha)

Post by kingarthur »

reapersms wrote:
Sun Mar 31, 2019 2:22 am
I could pave the factory with chromium I think.
Image

if you would send me a copy of that save file id like to take a look at it in detail and the tweaks you have made. would like a look at a later factory and i just havnt had time to play in a while to get as far as you have

reapersms
Long Handed Inserter
Long Handed Inserter
Posts: 69
Joined: Sun Jun 17, 2018 4:55 am
Contact:

Re: PyBlock (alpha)

Post by reapersms »

Sure thing, I'll get it and the tweaks packaged up a bit later.

edit:
Sent some download links along in a PM

reapersms
Long Handed Inserter
Long Handed Inserter
Posts: 69
Joined: Sun Jun 17, 2018 4:55 am
Contact:

Re: PyBlock (alpha)

Post by reapersms »

after a stint poking at space exploration, I cloned an install off to try 0.17 py.

I managed to hack things around to get pyblock to load with space-exploration, but I had to make some hefty changes to the world gen stuff, no clue if the space parts will still work though. Added oil-mk04, tar-patch, and sulfur-patch to the rock generation list, not sure if any of the other oil ones or the natural gas one is suitable.

A potential blocker in the beginning is that PyPH has added small parts to a number of things, so the initial iron may not quite be enough (they eat 3.5 iron plates a piece)

I fiddled around a bit with the chunk generation to give some guaranteed rock resources near the start area, and a guaranteed average density, will see how it works out. The approach I implemented was to generate one rock per 12x12 chunk region, placed in a random chunk within there, and placed randomly within that chunk such that it doesn't overlap (not sure how the chunk generation works there. To guarantee variety, I split the rock types up into 4 subsets, and interleaved those subsets in a checkerboard across the world. Additionally, the rock types for the 4 superchunks around (0,0) and the 8 immediately N, S, E, and W of those have their selection forced. That probably needs some tweaking, and the checkerboard pattern could be randomized to keep, say, iron from always being SE of the origin. To make you work at least a bit, it pushes the rock spawns out to at least 4 chunks away from the origin point.

After what I went through for that, my hat is off to all of you for putting up with the lua workflow.

edit:

the small parts change is not a blocker, got the full set of starter nursery, burner WP, DDC, and Quench tower set up producing iron & copper with ~85 iron and ~70 copper to spare

kingarthur
Smart Inserter
Smart Inserter
Posts: 1459
Joined: Sun Jun 15, 2014 11:39 am
Contact:

Re: PyBlock (alpha)

Post by kingarthur »

reapersms wrote:
Mon Apr 29, 2019 4:06 am
after a stint poking at space exploration, I cloned an install off to try 0.17 py.

I managed to hack things around to get pyblock to load with space-exploration, but I had to make some hefty changes to the world gen stuff, no clue if the space parts will still work though. Added oil-mk04, tar-patch, and sulfur-patch to the rock generation list, not sure if any of the other oil ones or the natural gas one is suitable.

A potential blocker in the beginning is that PyPH has added small parts to a number of things, so the initial iron may not quite be enough (they eat 3.5 iron plates a piece)

I fiddled around a bit with the chunk generation to give some guaranteed rock resources near the start area, and a guaranteed average density, will see how it works out. The approach I implemented was to generate one rock per 12x12 chunk region, placed in a random chunk within there, and placed randomly within that chunk such that it doesn't overlap (not sure how the chunk generation works there. To guarantee variety, I split the rock types up into 4 subsets, and interleaved those subsets in a checkerboard across the world. Additionally, the rock types for the 4 superchunks around (0,0) and the 8 immediately N, S, E, and W of those have their selection forced. That probably needs some tweaking, and the checkerboard pattern could be randomized to keep, say, iron from always being SE of the origin. To make you work at least a bit, it pushes the rock spawns out to at least 4 chunks away from the origin point.

After what I went through for that, my hat is off to all of you for putting up with the lua workflow.

edit:

the small parts change is not a blocker, got the full set of starter nursery, burner WP, DDC, and Quench tower set up producing iron & copper with ~85 iron and ~70 copper to spare
Ya it's why a .17 version was uploaded yet given the pita it is and issue figureing out new terrain. Going to be working on it soon since pyph is out now

reapersms
Long Handed Inserter
Long Handed Inserter
Posts: 69
Joined: Sun Jun 17, 2018 4:55 am
Contact:

Re: PyBlock (alpha)

Post by reapersms »

Yeah, it took a fair bit of trial and error to get it to generate right, especially with the extra headache of alien biomes from space-exploration. Just turning off autoplace was not enough, had a lot of hilarity when I either missed some tiles from biomes, and some more later when it started using out-of-map. For reference, what seems to have worked in the end was to use seablock as a template, and remove the parts it uses to generate cliff/tree/garden islands, and worm islands.

It does a straight loop over data.raw.tiles wiping out autoplace, then sets some explicit peaks for water and deepwater. After a lot of fiddling, nauvis generates right, but I'll need to fiddle with some creative mode stuff to find out of the space part still works with biomes stripped down.

kingarthur
Smart Inserter
Smart Inserter
Posts: 1459
Joined: Sun Jun 15, 2014 11:39 am
Contact:

Re: PyBlock (alpha)

Post by kingarthur »

reapersms wrote:
Mon Apr 29, 2019 4:02 pm
Yeah, it took a fair bit of trial and error to get it to generate right, especially with the extra headache of alien biomes from space-exploration. Just turning off autoplace was not enough, had a lot of hilarity when I either missed some tiles from biomes, and some more later when it started using out-of-map. For reference, what seems to have worked in the end was to use seablock as a template, and remove the parts it uses to generate cliff/tree/garden islands, and worm islands.

It does a straight loop over data.raw.tiles wiping out autoplace, then sets some explicit peaks for water and deepwater. After a lot of fiddling, nauvis generates right, but I'll need to fiddle with some creative mode stuff to find out of the space part still works with biomes stripped down.
while ill have to look you at least was able and had to specify autoplace for other surfaces so as long as you edit the nauvis autoplaces only it shouldn't break the others. if i cant get the generator to work i can always brute force it to be water only but thats extra work i wanted to avoid. both for me and factorio engine.

Post Reply

Return to “PyMods”