Great work! The video does state this clearly that it was about the journey first and foremost and that's great, but yet to me it feels unfinished when it ends as soon as we get to the really fun stuff, so it's complete in the sense of it being well-produced, publishable content, but it's uploaded as soon as it's publishable, and I'm left with "what, that's it?", as I've mostly been looking at milling and some coating. I get this often with similar videos today. Either it's just me (entirely possible) or it's a sign of the times.
It takes some sifting to find some really good “making” channels on YT. I’ve watched this video and while I applaud author’s efforts, I don’t consider this type of content “good enough” to be subscribing. It felt overproduced and with too epic tone, while giving too little detail on the process, the experimentation, the actual solution (he said ratios are important, but what ratios did he use) and no thorough explanation of what is happening.
The golden standard is Applied Science channel, of course, but there are some smaller channels with similar vibe.
I'm not sure how similar these are to what you're looking for, but:
- https://www.youtube.com/@primitivetechnology9550 - Primitive Technology, with John Plant. Non-narrated, but subtitled, videos of him building houses & other useful things with just clay, wood & stone. It's not a recreation of how people lived, but of what people might have done - he does research and tries to apply what he's learned to the materials available.
- https://www.youtube.com/@TechnologyConnections - Technology Connections. Less making, and more explaining, this has deep dives into (usually) older technology. There's something like six hours explaining how a particular pinball machine works, and I think his most recent video about VHS-C has already made it to the top of HN earlier this week.
Chris from Clickspring, the canonical YouTube machinist who has been slowly but accurately reconstructing the Antikythera mechanism for about one decade now.
Second Primitive Technology (don't forget to turn on the captions). Don't recommend Technology Connections to be honest (a lot of talk to the camera, I prefer videos that show things that can't be conveyed via text).
Here's the channels I like, in no particular order:
- https://www.youtube.com/@TechIngredients Thumbnails and titles are clickbaity, but don't let that fool you. One of the most thorough channels. Polymath like Applied Science.
A lot of honest projects are going to end this way, with a sort of half-failure. YouTube channels which show it anyway are more credible than the ones who seem to always succeed at whatever they're trying.
Yeah, I was hoping he'd get it at least to a somewhat usable state where you can at least load a small file (maybe with some file system sector fiddling).
It's interesting how HN crowd are mostly text (and text with low formatting too!) consumers. Compared to other social media, and even old school forums...
Are we mostly l33t developers here, in love with CLI and Vim? Ha!
Same reason I never click on YouTube links from friends: I find it annoying to have to wait to find out if I'm interested or not. On the other hand, I'll click on any non-video link someone sends me to have a look.
While it is a great video, it doesn't seem like he actually made a viable floppy disk in the end. Even if he didn't though, it would have been great to say what was actually achieved in the end: what write density was achieved? Could we write and recover even 1KiB of data?
I watched the video when it made the rounds last week. I was impressed with the work and the results. I did wonder, though, if a 5 1/4" disk would have been an easier initial goal, seeing as how the outer envelope is a lot less involved than a 3 1/2".
I was expecting a 5 1/4" or maybe 8". But the video was sponsored by a CNC machine company, so 3 1/2" hard shell form factor (the only popular one that can be CNC'd) makes sense. :)
I don't know how he got it, but if I were faced with that problem myself, I'd try this:
1. dissolve a bunch of rust in hardware-store hydrochloric acid,
2. dilute it in a lot of water,
3. into a similar quantity of water, mix an large excess of baking soda to neutralize the acid,
4. rapidly mix the two solutions together to precipitate a very fine iron hydroxide powder,
5. decant the powder and/or filter it with coffee filters,
6. rinse it to remove the remaining salt and sodium carbonate,
7. heat it to convert it to Fe₂O₃, and
8. heat the Fe₂O₃ in a sealed container with enough carbon to reduce it to Fe₃O₄.
I don't know if this would actually work, because my entire education in chemistry consists of watching NileRed videos in which the primary lesson is that nothing works the way you think it will. Wikipedia has some more-promising-sounding approaches that require materials I don't have: https://en.wikipedia.org/wiki/Iron(II,III)_oxide#Preparation
> use ammonia to promote chemical co-precipitation from the iron chlorides: first mix solutions of 0.1 M FeCl₃·6H₂O and FeCl₂·4H₂O with vigorous stirring at about 2000 rpm. The molar ratio of the FeCl₃:FeCl₂ should be about 2:1. Heat the mix to 70 °C, then raise the speed of stirring to about 7500 rpm and quickly add a solution of NH₄OH (10 volume %). A dark precipitate of nanoparticles of magnetite forms immediately.[9]
You can also buy it as a pottery pigment or as a black "ferrite" pigment for mixing into whitewash to make black paint, but if the particles are too coarse, you probably can't mechanically grind them down to be small enough.
You can get ferrous sulfate from the garden store as a fertilizer, and if you get it wet it likes to oxidize to ferric sulfate with the air. Or you can encourage it with hydrogen peroxide. I wouldn't be surprised if that would work as a replacement for the ferrous and ferric chloride mix in the Wikipedia recipe.
I would be very interested to know what material was used and how these magnetic films were created back in the 80s. Please share any info that you have!
In March 1998, CU Amiga magazine gave away the Amiga port of Doom. It was three DSDD disks, even accounting for the Amiga's larger 880 KB rather than 720 KB capacity. It was also only the shareware levels.
I'm sure that modern compression algorithms could do a better job than what Doom was using for its images. It appears that original Doom was basically using a vertically-oriented image format which indicated vertical strips of raw bytes, or transparent areas. It's much cheaper to skip drawing transparent areas.
Would obviously need some decode time to decompress the images, and memory to store the decompressed images.
A 5.25" single-density disk would be literally an order of magnitude easier to make. 4x larger magnetic domains. Larger tracks mean wobble matters less. No tight-tolerance shell. Thicker substrate.
Hehe, very nice to see something outside the scope of software or PCBs with this level of useless enthusiasm.
Obviously "from scratch" is a bit of a stretch here, but this is the material we come to Hacker News for.
Thanks for sharing!
Edit: sigh, I should probably run my comments through ChatGPT to avoid being downvoted. I like this, I share my enthusiasm. I like the uselessness of it, meaning the uselessness of making a floppy disk in 2025, not the lack of educational value. Sheesh.
He uses quite a bit of tooling, including lasers. It's not like he would be able to get this far in the middle of nowhere :)
In a way it is somewhat similar to people writing demos for old computers using emulators. Still great fun, but using these tools it doesn't take a village to make one floppy disk. With modern hardware you are apparently able to pull this off on your own. That would have been almost impossible in the 1980s, when these floppy disks were popular.
I probably worded it badly, but I really enjoy these efforts, and I would never be able to do this myself, even if I had a shed with all those tools!
For anyone that hasn't seen it yet, there's the YouTube channel "Primitive Technology", where a guy does this literally - in a jungle, with no tools apart from what he makes himself. He gets as far as smelting a tiny amount of iron.
I couldn't find an answer on google - how is a goat farm precursor for music? It's an activity that needs herding and shepherds started playing songs for fun? Or..? :o
It's an old story about using a drum machine and feeling like it's not real music, replacing the synthesized samples with real drum samples, then getting rid of the machine playing the drums yourself, then making your own drums, then finally farming your own goats for leather to make drumheads out of.
> How can someone call themselves a programmer when they don’t even mine for silicon!
To be fair, after three or four Tinder dates I realized that it was mostly silicon to be found there. It's not a stretch to say that a programmer going out on Tinder dates is mining for silicone!
Great work! The video does state this clearly that it was about the journey first and foremost and that's great, but yet to me it feels unfinished when it ends as soon as we get to the really fun stuff, so it's complete in the sense of it being well-produced, publishable content, but it's uploaded as soon as it's publishable, and I'm left with "what, that's it?", as I've mostly been looking at milling and some coating. I get this often with similar videos today. Either it's just me (entirely possible) or it's a sign of the times.
It takes some sifting to find some really good “making” channels on YT. I’ve watched this video and while I applaud author’s efforts, I don’t consider this type of content “good enough” to be subscribing. It felt overproduced and with too epic tone, while giving too little detail on the process, the experimentation, the actual solution (he said ratios are important, but what ratios did he use) and no thorough explanation of what is happening.
The golden standard is Applied Science channel, of course, but there are some smaller channels with similar vibe.
Please share the smaller channels if you have them handy! I’m very interested.
I'm not sure how similar these are to what you're looking for, but:
- https://www.youtube.com/@primitivetechnology9550 - Primitive Technology, with John Plant. Non-narrated, but subtitled, videos of him building houses & other useful things with just clay, wood & stone. It's not a recreation of how people lived, but of what people might have done - he does research and tries to apply what he's learned to the materials available.
- https://www.youtube.com/@TechnologyConnections - Technology Connections. Less making, and more explaining, this has deep dives into (usually) older technology. There's something like six hours explaining how a particular pinball machine works, and I think his most recent video about VHS-C has already made it to the top of HN earlier this week.
A few I like:
https://www.youtube.com/@Blondihacks - A (primarily) model engineering channel with a focus on hobby / home precision machining
https://www.youtube.com/@daliborfarny - A guy working to keep the art of nixie tube manufacturing alive
https://www.youtube.com/@StuffMadeHere - Silly / improbable projects mostly for fun (e.g. basketball hoop that you can't miss a shot)
Chris from Clickspring, the canonical YouTube machinist who has been slowly but accurately reconstructing the Antikythera mechanism for about one decade now.
- https://www.youtube.com/@Clickspring
Second Primitive Technology (don't forget to turn on the captions). Don't recommend Technology Connections to be honest (a lot of talk to the camera, I prefer videos that show things that can't be conveyed via text).
Here's the channels I like, in no particular order:
- https://www.youtube.com/@TechIngredients Thumbnails and titles are clickbaity, but don't let that fool you. One of the most thorough channels. Polymath like Applied Science.
- https://www.youtube.com/@HuygensOptics Optical Systems and connected topics from a veteran of the field
- https://www.youtube.com/@Borgedesigns Designing 3d-printed tools
- https://www.youtube.com/@Nighthawkinlight Like Applied Science, but trying to do stuff with easily acquirable materials
- https://www.youtube.com/@AdvancedTinkering Chemistry and vacuum tech
- https://www.youtube.com/@ExcessiveOverkill Hardware projects, one of the biggest is controlling an industrial robot arm, but others are cool too
- https://www.youtube.com/@SamZeloof Reached home-made semiconductors
- https://www.youtube.com/@projectsinflight Trying to reach home-made semiconductors
- https://www.youtube.com/@christopherhelmke Building industrial 3d-printed parts sorting system
- https://www.youtube.com/@MariusHornberger Most thorough woodworker
- https://www.youtube.com/@BreakingTaps Like Applied Science, but with more free time
- https://www.youtube.com/@benmakeseverything Cool hardware projects
- https://www.youtube.com/@ancientjames Holograms
- https://www.youtube.com/@NileRed More entertaining than educational, but a prominent chemistry channel
- https://www.youtube.com/@BenEater Classic: made computer on a breadboard
- https://www.youtube.com/@theCodyReeder Like Applied Science, but more outdoors type; builds a Martian-like base
- https://www.youtube.com/@colinfurze A welding guy with extremely high energy, builds underground garage
- https://www.youtube.com/tomstantonengineering Hardware projects mostly about flying stuff
- https://www.youtube.com/@mymechanics Machining guy restoring things; currently restores a car by individually handling every nut and bolt (yes)
- https://www.youtube.com/@HyperspacePirate Hardware / Chemistry projects, made liquid nitrogen with disassembled AC units in a long-running series of attempts
Not small but he's a regular amateur: RestoreIt [0]. I love this guys channel.
[0] https://www.youtube.com/c/RestoreIt
A lot of honest projects are going to end this way, with a sort of half-failure. YouTube channels which show it anyway are more credible than the ones who seem to always succeed at whatever they're trying.
Yeah, I was hoping he'd get it at least to a somewhat usable state where you can at least load a small file (maybe with some file system sector fiddling).
TFA is a very short blog post that says you should go watch this YouTube video. Here's a direct link to the video:
https://youtu.be/TBiFGhnXsh8?si=wra84H0R8fy2XCnd
We'll put that link in the top text. Thanks!
Title: "I"
First line: "[YouTuber] PolyMatt"
The article just advertises the video. This post could be just the video.
I for one never ever click on a video link here. I suspect I'm not alone.
You're not. I'll only click on a video, here, after checking the comments
It's interesting how HN crowd are mostly text (and text with low formatting too!) consumers. Compared to other social media, and even old school forums...
Are we mostly l33t developers here, in love with CLI and Vim? Ha!
I personally think the plain text howtos and forums of 1996-2002 were way easier to follow than the video links that come up these days.
Come on, it's absurd to think that we all follow a stereotype. Some of us use emacs.
I don't have eight megs to spare, you insensitive clod!
:wq
You monster.
In the beginning was the command line
The phrase “old school forums” really does a number on me. Forums are a web thing, and the web is newfangled tech.
Usenet, bbs. There are a lot of forums that predate the web.
They weren’t called that, were they? Usenet was just Usenet, or “news.” BBSes were BBSes.
And in context, “forums” was presented opposite plain text, and pre-web stuff tended to be plain text.
Why’s that?
Same reason I never click on YouTube links from friends: I find it annoying to have to wait to find out if I'm interested or not. On the other hand, I'll click on any non-video link someone sends me to have a look.
While it is a great video, it doesn't seem like he actually made a viable floppy disk in the end. Even if he didn't though, it would have been great to say what was actually achieved in the end: what write density was achieved? Could we write and recover even 1KiB of data?
This fascinating engineer [1] has actually achieved making tapes and floppies from scratch [2].
[1] https://www.ninakalinina.com/
[1] https://www.ninakalinina.com/links.htm / "DIY tapes and floppies"
I watched the video when it made the rounds last week. I was impressed with the work and the results. I did wonder, though, if a 5 1/4" disk would have been an easier initial goal, seeing as how the outer envelope is a lot less involved than a 3 1/2".
I was expecting a 5 1/4" or maybe 8". But the video was sponsored by a CNC machine company, so 3 1/2" hard shell form factor (the only popular one that can be CNC'd) makes sense. :)
In the early 80s, a lot of the floppy disks and drives I had to use could have been crafted by cavemen out of a Far Side cartoon.
there is no explanation on how to get the very fine black iron oxide powder in the video, it just appears out of nowhere.
I don't know how he got it, but if I were faced with that problem myself, I'd try this:
1. dissolve a bunch of rust in hardware-store hydrochloric acid,
2. dilute it in a lot of water,
3. into a similar quantity of water, mix an large excess of baking soda to neutralize the acid,
4. rapidly mix the two solutions together to precipitate a very fine iron hydroxide powder,
5. decant the powder and/or filter it with coffee filters,
6. rinse it to remove the remaining salt and sodium carbonate,
7. heat it to convert it to Fe₂O₃, and
8. heat the Fe₂O₃ in a sealed container with enough carbon to reduce it to Fe₃O₄.
I don't know if this would actually work, because my entire education in chemistry consists of watching NileRed videos in which the primary lesson is that nothing works the way you think it will. Wikipedia has some more-promising-sounding approaches that require materials I don't have: https://en.wikipedia.org/wiki/Iron(II,III)_oxide#Preparation
> use ammonia to promote chemical co-precipitation from the iron chlorides: first mix solutions of 0.1 M FeCl₃·6H₂O and FeCl₂·4H₂O with vigorous stirring at about 2000 rpm. The molar ratio of the FeCl₃:FeCl₂ should be about 2:1. Heat the mix to 70 °C, then raise the speed of stirring to about 7500 rpm and quickly add a solution of NH₄OH (10 volume %). A dark precipitate of nanoparticles of magnetite forms immediately.[9]
You can also buy it as a pottery pigment or as a black "ferrite" pigment for mixing into whitewash to make black paint, but if the particles are too coarse, you probably can't mechanically grind them down to be small enough.
You can get ferrous sulfate from the garden store as a fertilizer, and if you get it wet it likes to oxidize to ferric sulfate with the air. Or you can encourage it with hydrogen peroxide. I wouldn't be surprised if that would work as a replacement for the ferrous and ferric chloride mix in the Wikipedia recipe.
I think the device you need for creating a fine powder is a ball mill
https://en.wikipedia.org/wiki/Ball_mill
but yes, you can certainly just buy fine Fe3O4
How beautifully designed was the IBM floppy disk box, visible at the beginning? Great piece of design and branding!
In the new Mission: Impossible film they're tasked with making an 8" disk drive from scratch. That should be his next video :)
8" drives having lower density, I would think that is easier.
Fun fact: I only recently found out that regular 1.44MB floppy disks could be formatted to 32MB.
Iron oxide is not what regular floppy disks use. That's probably what the issue was.
I would be very interested to know what material was used and how these magnetic films were created back in the 80s. Please share any info that you have!
More info here!
https://youtube.com/watch?v=TBiFGhnXsh8&lc=UgyN4fVDoD8lcryPw...
Can you fit Doom on it & play it? Bootable Doom Floppy?
In March 1998, CU Amiga magazine gave away the Amiga port of Doom. It was three DSDD disks, even accounting for the Amiga's larger 880 KB rather than 720 KB capacity. It was also only the shareware levels.
I'm sure that modern compression algorithms could do a better job than what Doom was using for its images. It appears that original Doom was basically using a vertically-oriented image format which indicated vertical strips of raw bytes, or transparent areas. It's much cheaper to skip drawing transparent areas.
Would obviously need some decode time to decompress the images, and memory to store the decompressed images.
A 5.25" single-density disk would be literally an order of magnitude easier to make. 4x larger magnetic domains. Larger tracks mean wobble matters less. No tight-tolerance shell. Thicker substrate.
These kinds of hobbies always teach people more than expected.
He gets surprisingly close to viable storage media. Nicely done =3
Hehe, very nice to see something outside the scope of software or PCBs with this level of useless enthusiasm. Obviously "from scratch" is a bit of a stretch here, but this is the material we come to Hacker News for.
Thanks for sharing!
Edit: sigh, I should probably run my comments through ChatGPT to avoid being downvoted. I like this, I share my enthusiasm. I like the uselessness of it, meaning the uselessness of making a floppy disk in 2025, not the lack of educational value. Sheesh.
Your definition of “from scratch” is pretty unrealistic.
If someone was to say “make a pasta source from scratch” then that wouldn’t mean refining your own copper to make your source pans.
The problem is creating the floppy disk. Not the tooling to create the floppy disk.
Judging from the video, it looks pretty "from scratch" to me. What makes it a "bit of a stretch" to you?
He didn't first create the universe.
"If you wish to make an apple pie from scratch, you must first invent... the universe" Carl Sagan https://www.youtube.com/watch?v=7s664NsLeFM
Related: https://m.youtube.com/watch?v=zSgiXGELjbc
He uses quite a bit of tooling, including lasers. It's not like he would be able to get this far in the middle of nowhere :)
In a way it is somewhat similar to people writing demos for old computers using emulators. Still great fun, but using these tools it doesn't take a village to make one floppy disk. With modern hardware you are apparently able to pull this off on your own. That would have been almost impossible in the 1980s, when these floppy disks were popular.
I probably worded it badly, but I really enjoy these efforts, and I would never be able to do this myself, even if I had a shed with all those tools!
Are you even a musician if you don’t have a goat farm?
How can someone call themselves a programmer when they don’t even mine for silicon!
For anyone that hasn't seen it yet, there's the YouTube channel "Primitive Technology", where a guy does this literally - in a jungle, with no tools apart from what he makes himself. He gets as far as smelting a tiny amount of iron.
I couldn't find an answer on google - how is a goat farm precursor for music? It's an activity that needs herding and shepherds started playing songs for fun? Or..? :o
It's an old story about using a drum machine and feeling like it's not real music, replacing the synthesized samples with real drum samples, then getting rid of the machine playing the drums yourself, then making your own drums, then finally farming your own goats for leather to make drumheads out of.
Stringed instruments use goat or sheep intestines. And drums are from their skin (leather).
> when they don’t even mine for silicon!
Knowing how to design a CPU is quite helpful.
Silicon and silicone are two very different things...
It’s not sili-cone valley, you have to say sili-kawn.
Silicone Valley is in Southern California.
(The San Fernando Valley was central to the porn industry in the late 20th century.)
So now we've got clankers catfishing human singles?
> not making your own plastic monomers from syngas
why even bother
Start with naturally occurring things only.
Mine and refine iron ore to make hub. Mine and refine zinc(?) to plate it.
Drill for and refine oil to make PET for disk and casing. Injection mold casing. Make film for actual disc.
Etc, etc.
I’d be ok using tools that weren’t made from scratch as well, but that’d be bonus ooints.
"If you wish to make an apple pie from scratch, you must first invent the universe" - Carl Sagan
Like pretty much everyone responding, I disagree.
That's it, that's all a downvote means. Don't be afraid of them, it's not worth it.
Oh so OP recreated the Universe?