If I'm not mistaken, you've bootstrapped a substantial fraction of an industrial economy. Now that is epic. To me, a software development environment is small potatoes compared to your windmill.
Here's a question I've often thought about that you might have insight on: what would be the shortest path from raw materials and an unlimited amount of information to, say, a reasonable bicycle?
For example: bronze, despite its scarcity, is easier to make from raw materials than steel, but I don't know if it's hard enough for ball bearing races, and I think you need ball bearings for a ridable bicycle. You can probably make a tiny blast furnace with leather bellows and ceramic (or can you?), but then you need to form the bearing balls and the races. Would you cut grindstones from stone? Perhaps electrolyze seawater to get magnesium oxychloride to bind together quartz sand?
The hard part I think would be the mining. Most if not all of the easily accessible resources have been depleted, so bronze age people had an easier time of it than we do!
Making a bicycle out of bronze would be an interesting exercise. I'd not use ball bearings, but rather babbit bearings and instead of a chain drive I'd use a belt drive (made from cloth).
Bamboo rather than steel or bronze for the frame to keep the amount of metal to a minimum, spokes would be tricky, wheels could be made out of wood.
Inner tubes made out of intestine, outer tubes made out of cloth, bulky but workable.
Something like that?
Now that you have me thinking about this I realize that I'm actually sorely tempted to try this.
Another hard part would be to not (unwittingly) cheat and to work off existing tools and/or raw materials that aren't truly raw.
edit: more thinking about this, the high stress points in the bike would be the spot where the pedals attach (even on my regular bike I can see it flex there), the joint where the tube from the handlebars goes through to the fork, the fork itself and the saddle joint.
Those would probably have to be cast out of bronze and then filed (with what?!) to their precise dimensions, or, alternatively you could first bootstrap a primitive lathe (which would require at least one piece of hardened steel for the bit).
Primitive blast furnaces are fairly easily constructed out of clay, I've seen some in NL that were absolutely doable, depending on what ores you can get your hands at in the end steel might even be easier than bronze.
The belt drive, intestine inner tubes, and bamboo frame are excellent ideas.
That's a good point about the bronze. I think there's still a basically unlimited amount of small deposits of rich iron ore out there (e.g. magnetite sands) and, of course, aluminum, but that's a lot harder to smelt.
I had the impression that Babbitt bearings were too inefficient to make bicycles practical, which is why the bicycle boom didn't happen until the invention of the ball bearing. But I could be mistaken about that. Certainly you could do Babbitt bearings without particularly great metallurgy. (But you need a wider variety of metals.)
I know some people here with bamboo bikes, except for the joints, of course.
Spokes could possibly be made of glass or bamboo, but I'm not sure how you'd pull them tight without being able to thread them. Maybe you could lace the wheel like a circular shoe instead.
I don't think cloth tires will wear for long enough to be useful. You'll probably need at least buna reinforced with carbon black or fumed silica.
As for the working off existing tools: maybe you could make it an iterative process, starting with Dave Gingery tools and materials and working your way down with successive bikes?
Do you think you could do a lathe with a non-metal bit? Corundum, carborundum, diamond, or maybe topaz? I've never turned anything on a lathe, so this might be a stupid question.
I think you can grind some things to precise dimensions rather than filing them, and that might require less materials technology. Knife sharpeners use a particular kind of flint for some of their work.
I definitely, definitely want to know if you end up doing some of this!
Modern lathe bits are ceramic or sintered metal, so it should be possible to use other materials.
Grinding / sanding are good ways to get near perfect shapes (it is typically the last pass after the lathe has done its work if you want extreme precision). Ball bearings are made using an interesting process (using a vibrating drum with abrasive, the resulting minimal surface elements are perfectly spherical by definition...).
Babbit bearings have a higher friction coefficient, but we're talking about making a working bicycle, not the most efficient one, after that it is incremental improvements and of course, ball bearings are better but they may not be the shortest path to success.
You've really done a bad thing here, I can't get this out of my head... :)
Here's a question I've often thought about that you might have insight on: what would be the shortest path from raw materials and an unlimited amount of information to, say, a reasonable bicycle?
For example: bronze, despite its scarcity, is easier to make from raw materials than steel, but I don't know if it's hard enough for ball bearing races, and I think you need ball bearings for a ridable bicycle. You can probably make a tiny blast furnace with leather bellows and ceramic (or can you?), but then you need to form the bearing balls and the races. Would you cut grindstones from stone? Perhaps electrolyze seawater to get magnesium oxychloride to bind together quartz sand?