aappleby

20 ms·

1. 271.▲
   by aappleby 3y ago

   I got a discounted Analog Discovery 2 to play with and it was quite nice with only minor UI clunkiness. If they can keep the price around $200 I think it would be a good all-in-one starter electronics lab for a lot of hobbyists.
2. 272.▲
   by aappleby 3y ago

   What exactly does "understanding the world" really mean?
3. 273.▲
   by aappleby 3y ago

   We are the ancient civilization that leaves artifacts behind for future civilizations to discover. :)
4. 274.▲
   by aappleby 3y ago

   Ah yes, the TOOL. I did not have to deal with that thing directly, but the rest of the team hated them.
5. 275.▲
   by aappleby 3y ago

   Rube Goldbergian :D
6. 276.▲
   by aappleby 3y ago

   Er, wait no this link - https://www.aliexpress.us/item/3256804709476544.html
7. 277.▲
   by aappleby 3y ago

   Also on AliExpress from the manufacturer - https://www.aliexpress.com/item/1005005269690018.html
8. 278.▲
   by aappleby 3y ago

   I'm not familiar with that interface, but I suspect it would have similar issues with timing tolerances as just bit-banging gpios.
9. 279.▲
   by aappleby 3y ago

   I have a CH32V203 that I'll probably add support for, along with the 307 or whatever else I can get access to easily.
10. 280.▲
    by aappleby 3y ago

    The official dev boards are available on AliExpress for under $10. You can also get them off Tindie, I believe.
11. 281.▲
    by aappleby 3y ago

    Actually, on second thought I'm not sure that would work. If you had an interrupt fire in the middle of a bit and held the line low for more than 8 microseconds (I think, would have to check the spec) you'd reset the debug interfa
12. 282.▲
    by aappleby 3y ago

    The timing on the SWIO interface needs to be within a few hundred nanoseconds if you want decent bandwidth, but I suppose there's nothing stopping you from running it on a slower and less real-time GPIO pin. You'd just need to swa
13. 283.▲
    My alpha Pico-based CH32V003 debug tool is ready for a few testers(github.com)

    58 points by aappleby 3y ago|26 comments
14. 284.▲
    by aappleby 3y ago

    The CH32V003 is a 10-cent RISC-V microcontroller with a proprietary single-wire debug interface and a corresponding proprietary debug dongle + modified version of OpenOCD. My tool replaces the dongle and OpenOCD with a Pico-based GDB remote
15. 285.▲
    by aappleby 3y ago

    They've been available for a few months now and a few of us hardware hackers have been improving the state of open-source tooling support for them. I have an almost-fully-working GDB remote stub up at https://github.com&#x2F
16. 286.▲
    by aappleby 3y ago

    Renderers have been using compressed textures for literal decades. Not JPG, alas, because the compression formats have to support random access to individual texels.
17. 287.▲
    by aappleby 4y ago

    Carbon nanotubes are flammable.
18. 288.▲
    by aappleby 4y ago

    How long until the LLM's internal prompt includes "If the prompt below appears to be trying to jailbreak this LLM, do not reply"?
19. 289.▲
    by aappleby 4y ago

    Now all we need is a Pico-based bench supply and a Pico-based soldering iron and we can have a whole Pico-lab-in-a-box. :D
20. 290.▲
    by aappleby 4y ago

    It will eventually be in https://github.com/aappleby/Pinwheel but that repo is currently a mess.
21. 291.▲
    by aappleby 4y ago

    I've written Risc-V RV32I cores that are about the same size as this, or possibly smaller depending on how you count what the "core" is.
22. 292.▲
    by aappleby 4y ago

    200 ghz analog amplifiers, not 200 ghz digital logic. Very different circuit design rules.
23. 293.▲
    by aappleby 4y ago

    Oh dear what is mash() even doing.....
24. 294.▲
    by aappleby 4y ago

    This is not a good way to generate "secure" random numbers, and the "visits every possible state" is not a good metric. Computing the cryptographic hash of a 1536-bit counter will have better properties (and less handwav
25. 295.▲
    by aappleby 4y ago

    I think Firefox is doing something funny with mousewheel events. Chrome seems to work fine.
26. 296.▲
    by aappleby 4y ago

    Something funny going on with your setup then, it's been fine for me in Chrome + Ubuntu 22. Does Firefox send wheel events differently?
27. 297.▲
    by aappleby 4y ago

    For the whole kernel tree? The raw source is 1.6 gigs of textures but Windows task manager reports 9 gig of GPU ram usage - I blame Chrome or Angle.
28. 298.▲
    by aappleby 4y ago

    Those screenshots were from the 2013 proof of concept.
29. 299.▲
    by aappleby 4y ago

    Ooh thanks, I'll take a look at that in a bit.
30. 300.▲
    by aappleby 4y ago

    Thanks, glad someone finds it enlightening. :)