Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

What I learned selling 2,500 MIDI recorders: Hardware is not so hard

chipweinberger.com

Post reply on HN
Print view View original

51–60 of 276 posts

Re: What I learned selling 2,500 MIDI recorders: Hardware is not so hard

#51

Post by alabhyajindal »

embedding-shape wrote: post #37
dgellow wrote: post #35

Wow, congrats, I remember reading your HN post. I hope you will continue writing blog posts on your journey, I expect quite a lot of software engineers would be very interested going though a similar journey now that LLMs removed most of the fun of programming. It’s neat to work on something physical for a change

> now that LLMs removed most of the fun of programming Never understood this, was the "most of the fun of programming" the literal typing for you? For me it was about solving problems and building things, and I get to do that more now than before, and without all the literal typing.

Did solutions to problems just pop into your head when you were programming before LLMs? Did you not pause and think about the code you were writing before?

Re: What I learned selling 2,500 MIDI recorders: Hardware is not so hard

#53

Post by thenthenthen »

swiftcoder wrote: post #36
chipweinberger wrote: post #22

Earlier quoted context omitted.

IMO it's better to ship preprogrammed ICs to the factory, so you can do your secure burning in-house.

Depends on where the ICs originally come from. No point shipping your ICs from China -> US, flashing them, shipping them back US -> China for assembly, and then ship them China -> US again to sell

Can reply to the reply below but yeah, i have done that, lol

Re: What I learned selling 2,500 MIDI recorders: Hardware is not so hard

#54

Post by Hackbraten »

swiftcoder wrote: post #17
stavros wrote: post #13

Earlier quoted context omitted.

Please do, that'd be very interesting. I imagine it's doable to use random per-device I'd when you're hand-flashing, but probably harder on a huge production run.

Almost all commercial devices need a unique ID, private key, etc burned in on the production line. For a big enough production run, you'd likely ship an HSM to the factory

> For a big enough production run, you'd likely ship an HSM to the factory

I figure that the factory is exactly the adversary in the whole threat model? So why give them the keys to the castle so they can moonlight their own genuine batch?

Re: What I learned selling 2,500 MIDI recorders: Hardware is not so hard

#55

Post by chipweinberger »

bonzini wrote: post #50
chipweinberger wrote: post #6

Earlier quoted context omitted.

The app is about 85K lines: device connectivity, first-time setup, syncing, MIDI playback and visualization, library management, audio/video export, and OTA workflows. The firmware is about 110K lines: MIDI recording + streaming, USB, Bluetooth, Wi-Fi, web APIs, storage, OTA updates, diagnostics, security, peripherals, provisioning, and an embedded web interface. There’s a lot required to make a full-featured product…

What is the manufacturing tooling that you mention?

3 main components: flashing, provisioning keys, hardware testing.

Nearly every component gets tested: sdcard, clock, battery, anti-counterfeit IC, led, buzzer, the tactile-switch!

Re: What I learned selling 2,500 MIDI recorders: Hardware is not so hard

#56

Post by TuringNYC »

Congrats. I did want to respond to this

>>> The hardest part of building Jamcorder was still, by far, the software -- roughly 200K lines of code spread across the firmware, app, and manufacturing tooling. It took over 3 years and many long nights in a pre-LLM world.

>>> When compared to that, the hardware was undeniably smooth sailing.

>>> For the record, I don’t think I’m special. It’s just that hardware’s reputation for being difficult is, IMO, overstated.

For me, when I complain about hardware, its actually usually complaining about drivers, compatibility, underdocumented interfaces, firmware etc. If one were to put firmware and manufacturing tooling in a separate bucket, that is indeed the problem!

Re: What I learned selling 2,500 MIDI recorders: Hardware is not so hard

#58

Post by alance »

Ok are you riding with two huge boxes containing all your work on an electric unicycle and I guess the hi-vis top means you're headed into traffic and the box-cords seem to be balancing at the halfway point of the boxes and this is too much anxiety for me. But hey, glad you optimized for minimalism on the PCB ♡

Re: What I learned selling 2,500 MIDI recorders: Hardware is not so hard

#59

Post by etrautmann »

shizcakes wrote: post #11
chipweinberger wrote: post #2

I'll be sticking around in the comments so feel free to ask questions. AMA!

What are some examples of good anti-counterfeit strategies? And, if you’re comfortable, which applied for JamCorder?

+1

Re: What I learned selling 2,500 MIDI recorders: Hardware is not so hard

#60

Post by dgellow »

embedding-shape wrote: post #37
dgellow wrote: post #35

Wow, congrats, I remember reading your HN post. I hope you will continue writing blog posts on your journey, I expect quite a lot of software engineers would be very interested going though a similar journey now that LLMs removed most of the fun of programming. It’s neat to work on something physical for a change

> now that LLMs removed most of the fun of programming Never understood this, was the "most of the fun of programming" the literal typing for you? For me it was about solving problems and building things, and I get to do that more now than before, and without all the literal typing.

I reject the idea that programming is „just typing“, so we are very unlikely to understand each others. Writing the code and thinking about the exact logic you want to apply and all its implications are part of the same problem solving process in my mind. You can for sure plan things at a different layer of abstraction but putting things together in code is a very satisfying and IMHO is the actual building part of the job. It’s full of decision making and thinking.

LLMs move me away from the implementation, which is more a managerial role, something I personally have zero interest in.

Post reply on HN