Tweets
Replying to @josecastillo
Still, y’know: never waste a rev. There’s always something you can learn from it. Since I routed my board so precisely wrong — the exact opposite of what I should have done — it was easy to flip the LCD 180 degrees. Now I can get a head start on writing driver code for it.
(original)
Replying to @josecastillo
“Drones tell people to stay home. ‘Control your soul’s desire for freedom,’ it says.” it’s actually decent writing, which is more than I can say for a lot of our recent dystopia.
(original)
My sis sent me this video clip, and now I’m trying to imagine what I would have imagined as a child if told this was a news report from the future.
“Robots patrol some streets reminding people to wear masks.”(original)
Replying to @josecastillo
this was an embarrassment. Wordle 297 6/6*
🟨🟨⬛⬛⬛
🟩🟨⬛🟨⬛
🟩🟩🟨⬛⬛
🟩🟩⬛🟩🟩
🟩🟩⬛🟩🟩
🟩🟩🟩🟩🟩(original)
Replying to @josecastillo
backed the right horse today. Wordle 296 3/6*
🟨🟨🟨⬛⬛
🟨🟨🟩⬛⬛
🟩🟩🟩🟩🟩(original)
Replying to @bitshiftmask
move fast and make thigns
(original)
Replying to @josecastillo
roast me, I can take it
(original)
joey starter pack https://twitter.com/durandal_exe/status/1512391291430219784
(original)
Replying to @mycoliza
many such cases
(original)
Replying to @josecastillo
The things we can do.
(original)
Hilarious fail with the LCD FeatherWing project: first samples came in today, and I realized I’d laid out the LCD pins precisely backwards so all it displays is gibberish. No worries though: @oshpark Super Swift to the rescue; I’ll have a corrected board by next week.
(original)
Replying to @willianpaixaoo
“It’s possible he decided the most efficient way to get rid of all the bugs was to get rid of all the software.”
Fun fact: this works for unit tests too.
(original)
Replying to @femtoduino
you know I can think of at least one load-bearing comment in my codebase, and it does kind of make me think of how far we have strayed from the light.
(original)
Replying to @josecastillo
i have a deeply held instinct that all code is bad, so anything i can do to have less of it is fine by me.
(original)
you love to see it: 17 changed files with 0 additions and 275 deletions. https://github.com/joeycastillo/Sensor-Watch/commit/6bdaff5d2abaeb614e98de2c9e5d0a02b439fa8f
(original)
Replying to @josecastillo
Still waffling between two start words, but neither would have helped me today. Wordle 295 5/6*
🟨⬛⬛⬛⬛
⬛⬛🟩⬛⬛
⬛🟩🟩⬛⬛
⬛🟩🟩🟨⬛
🟩🟩🟩🟩🟩(original)
Replying to @deshipu
For sure, that’s the plan. Design files are up on GitHub, tho I want to test some more and make one last tweak before putting it up on Kitspace.
(original)
Replying to @josecastillo
I got excited the other day and broke the thread, but the work continues here: https://twitter.com/josecastillo/status/1512114638116249606
(original)
Replying to @josecastillo
this spreadsheet estimates based on a run of 100 (which coincidentally is about the number of SAMD51 chips I could get before the year of the rabbit), and I sense it’s actually a low estimate; the PCBA is dense and double sided, and this doesn’t include costs of logistics. Alas.
(original)
Replying to @josecastillo
ugh it pains me to do the math on this. Earlier I thought hey, what if I could make a small run of these? But the cost of making just one (not including my labor) is at least $41 — too high for it to make sense. I think the final gadget would be too expensive to justify itself.
(original)
Replying to @josecastillo
periodic reminder that you can mute this thread if you don’t care about my wordle scores; i won’t be offended. Wordle 294 3/6*
🟨🟨🟨🟨⬛
🟩🟨🟨🟩⬛
🟩🟩🟩🟩🟩(original)
I am decidedly not a game programmer, but I did spend some time this morning porting @BlitzCityDIY’s Blinka Jump game to the PyPen M4! I had a sense when I saw the aspect ratio of the display that it would be great for this style of game :) https://learn.adafruit.com/blinka-jump-pybadge-game
(original)
RT @FishesPretzel: hi! in light of recent bullshit:
i am a transgender boy living in alabama. i am making this thread in a desperate attemp…(original)
Replying to @josecastillo
Once again the major fail in this rev is battery-related. It’s very difficult to insert the flex cable when the battery is soldered into place. Probably need to move those pads to the far side of the board. (also this battery is v dead so I’m off to Aliexpress to get some more)
(original)
Replying to @bob123m
I kind of regret the state that it’s in tbh; I spent last year on the watch instead of the book and as such everything’s in kind of a work-in-progress state. I want to apply all the lessons learned this year to the book project, but it’s not front and center right now.
(original)
Replying to @bob123m
Which version of the Abridged Edition? That one is still a work in progress, C1-00 has all the parts on the board, C1-01 involves a castellated module (C2-01) that I had done as a PCBA job in the hopes of making it easier for folks to build, but I never made them available…
(original)
Replying to @MakerBlock
IDK tho! I wonder if I can avoid words like “patch” where one letter has so many possibilities (match, catch, watch etc), in favor of words with fewer options? I suppose at some point you have to start dialing it in, but this may not have been a wise word for a second guess.
(original)
Replying to @josecastillo
There’s got to be some way to avoid this. Wordle 293 6/6*
🟨⬛⬛⬛⬛
🟩⬛🟩🟩🟩
🟩⬛🟩🟩🟩
🟩⬛🟩🟩🟩
🟩⬛🟩🟩🟩
🟩🟩🟩🟩🟩(original)
Replying to @the_perigoso
Alas, the SAM L22 doesn’t have an internal temperature sensor. It is an odd omission; every other SAMD and SAML chip I’ve worked with does.
(original)
Replying to @josecastillo
So yeah, electronics twitter: anyone have any advice for me? Do I need an atomic clock, or a frequency counter with a crystal in an oven? What’s the best way to go about measuring frequency drift that’s happening in the range of single-digit parts per million?
(original)
Replying to @josecastillo
Of course I can output the clock on a GPIO pin and measure the frequency with an oscilloscope, but that depends on the accuracy of the reference clock in the scope. I think that what I need is a dedicated frequency counter, and an extremely accurate reference clock.
(original)
Replying to @josecastillo
Since most folks will have a temperature sensor in their watch, I could build a table with the frequency correction to apply for a given temperature, and thus give the watch an even more accurate *temperature compensated* crystal oscillator, all in software.
(original)
Replying to @josecastillo
The gist: Sensor Watch has a 32768 Hz crystal oscillator, and it keeps very good time. But I know that the frequency drifts ever *ever* so slightly based on the ambient temperature, and I’d like to measure that drift.
(original)
Brief advice thread: I think I need a new piece of kit for the shop, and I think it boils down to “Do I need an atomic clock,” or if not, what exactly do I need to accomplish what I’m trying to do here.
(original)
Replying to @GregDavill
omg same! Misread the pin 1 marking for my LED as the marking for the microcontroller (and then spent five minutes trying to figure out which way to orient the LED)
(original)
Replying to @josecastillo
Unfortunately, in a moment of extreme silicon shortage, I fried a precious SAM D51 chip when I plugged it in upside down. Fortunately I had one spare one I’d pulled off of another gadget. Either way, it boots! Now to get a CircuitPython build on here.
(original)
Replying to @cabe_bedlam and @oshpark
In theory, this Sharp Memory Display is transflective. In my testing not a lot of light gets through from the back, but still I figured I’d paint it white this time and see if I can notice a difference.
(original)
Replying to @josecastillo
It’s beautiful! Now let’s see if I can flash the bootloader.
(original)
Replying to @bateskecom and @oshpark
SAM D51, and it’s definitely on my radar screen! The dream would be to train a model that recognizes numbers and arithmetic operators to properly implement its Calcupen inspiration: https://blog.adafruit.com/2022/03/15/calcu-pen-the-electronic-calculator-pen-history-and-demonstration/
(original)
Replying to @josecastillo
update: it has been 0 days since I have placed a microcontroller entirely upside down
(original)
Replying to @josecastillo
Let’s do this.
(original)
Replying to @bateskecom and @oshpark
No gyro or accelerometer on this one; decided I wanted to play with some different kinds of sensor data, so it has a PDM microphone instead! Coupled with the 2MB of QSPI Flash it could probably record voice notes or something.
(original)
Replying to @josecastillo
PHEW. Wordle 292 6/6*
🟨⬛⬛⬛🟨
⬛🟨⬛🟨🟨
🟨🟩🟨⬛⬛
🟨🟩🟩⬛⬛
⬛🟩🟩🟩🟩
🟩🟩🟩🟩🟩(original)
they have arrived
(original)
Replying to @arturo182
I’ve long had a specific game that I want to make and have felt the same way: stuck with a sense that the skills involved are something I’m unlikely to be able to learn. Still I think it’s worth trying, and I hope to someday. I hope you will too!
(original)
Replying to @josecastillo
Wordle 291 5/6*
🟨⬛⬛⬛⬛
⬛🟨⬛⬛⬛
⬛🟩🟨🟩⬛
⬛🟩⬛🟩🟩
🟩🟩🟩🟩🟩Me at guess #4:
(original)
Replying to @josecastillo
I find these tolerances kind of amazing; I’m really curious if I’m going to be able to come up with a 3D printed enclosure for the pen and piezo buzzer that nails this.
(original)
Well if this isn’t the tiniest pogo pin I’ve ever seen! New PyPen boards arrive tomorrow, but parts for it arrived yesterday, including this 2.5mm tall pogo pin that I hope to use to attach a piezo buzzer. Ordered 4 of them and already dropped one, never to be found again. (oops)
(original)
Replying to @orviwan
It’s missing the Casio logo in the top left? Between that and the iffy build quality I would sense that this is it a knockoff, which means its almost certainly not compatible with the board swap. I’ve yet to see a counterfeit that works, because none of them copy the LCD exactly.
(original)
It’s done. Well, the first 90 anyway. Now I have a backer update to write.
(original)