Factorio Forums

Are you the kind of player that completely redesigns part of your factory when the better upgrades come in? Or do you leave the older stuff as is and just add all the new stuff elsewhere and let the old and new just work in tandem?

I command that you answer this, otherwise - I mean I won't do actually do anything, but I'll still be mildly upset at a later time. Nothing long-term, but it'll still probably happen.

I think I'm a mixed bag? It probably depends on what we're talking about.

For the most part, I try doing my initial base in such a fashion as to minimize redesign, even though that often means higher material costs from everything being spread out. However, as I get further along, I will rework some of it, while other pieces I leave as is.

That said, my biggest issue is finding a new mod that makes me go "Ooooo" and just restart the whole map....

IMO, never tear something down until its replacement is up and running.

You are limited to what has already been produced (and I hat you hand craft) if you tear a production facility/line down before building its replacement. On the other hand, if you build the replacement first, hook it up (and disconnect the original) you do not suffer any real loss in productivity.

I am pretty happy with my mall zone, with sectors for everything (and I do mean everything from iron plates to nuclear reactors) except science packs. How to measure a mall zone? Mine can sustain 20 PM3PM (productivity module three per minute) quite comfortably...
I have a reconstruction zone, with zillions of storage chests and train lines to deliver (and recover for reuse) deconstructed items. I have a flock of spidertrons (is flock the right word for spideys? maybe horde? or band? ooh now I want a band of rock'n'roll playing spideys... I get easily distracted, did I mention that?) for construction, deconstruction, and reconstruction, supported by an artillery train for clearing out biter bases when required for expansion.
Other than that I have rebuilt the rest of the world three times so far. First was to change semi-normal city block to DI (direct insertion) - everything from ore mined directly into trains to science packs being inserted directly from train to research lab. That was a lot of fun especially getting the rocket launches to work properly.
The second was to replace DI with octagons (octagons are the bestagons lol) and mix of belts and bots.
The third and current is what I am calling a wave design - all the sectors are circles, and the tracks make wave designs over and under those circles. This is mostly a bot design, as the circles are tiny, like clouds of little bubbles in map view. It isn't complete, no rocket silos yet, but I am already thinking about ripping it out and replacing it... but with what? Taking a break with a spaghetti world (it is really hard to do spaghetti when you lay down neat rows of stuff like green circuit factories without thinking about it) and trying out IR2 for the first time while I make up my mind...

With the central part of the factory (mall and research), I end up circling around my starting place. I build new stuff to available empty space next door, however water, cliffs and rails limit where I actually build. Later, water and cliffs are no obstacles any more, but for some reason, I refuse to relocate rails.

I don't keep obsolete tech around. If I unlock their electronic equivalents, I deconstruct fueled buildings.

Sometimes, if I need that nice big space in the center, I place down 50 storage chests and deconstruct everything in that space, then put down a prepared blueprint and let the robots do the work. I spend more time in the sandbox to design blueprints than in the actual map. I don't copy factory blueprints from the web, I use only self designed ones, so this is quite time consuming (but more fun than to design blueprints in the actual map). However, I take inspiration from existing blueprints and some best practice small snippets like balancers are used, of course.
I don't build any factory by hand, I only use my blueprints. I even designed absolute starter blueprints and build these manually until I get construction robots.
Part of the fun is to see my carefully crafted perfect blueprints fail in the real map

I use a self designed growing mall that comes with 3 stages following the major research features: up to steel, up to red circuits, up to blue belts/robots (I think). Upgrading this mall during research is a constant process of construction, filtered deconstruction, re-applying of new unlocked recipes on the spot. The storage chests and all the belts for the source materials are what makes the mall as a whole immobile.

I see 4 ingame eras:
1. beginning. Manual crafting to get items for basic automation
2. coal powered smelting
3. electricity powered smelting (as soon as enough solar panels are available); economical module use for single machines where useful, no beacons.
4. build one rocket silo, one rocket, one satellite, then launch and finish.
or as alternative for 4.:
4. nuclear power
4.1. beaconize everything: smelting, production, oil mining and processing.
4.2. build bigger fully beaconized facilities for continuous rocket launches and infinite research

For every stage, I have different blueprints and tear down the previous stuff after new stuff has been built.

I build usually in two phases. At start factory I build everything around messy bus. I know approximately what I need to finish the "official" part of the game so now my starter bases are quite clean if I play without mods.

When I get all finite science I decide how big and what kind of actual base I build and make everything according my aesthetic and engineering mind. I left original base nearby as some kind of abandoned or old "museum" stuff. I like that there is realistic like succession in my base like in real industrial zones instead of perfect ratios everywhere.