Reminds of a story my father tells - he's a stage designer and one day one of the older actors was complaining that he couldn't see from his position back stage to know when to come on set. The stage manager set up an early cctv system for him (this was 1980something) to help him see.
The actor then complained that he had a hard time seeing the black and white screen so while the stage manager went to see how he could improve his set up the actor asked my father for a penknife, cut a small hole in the set that he could put his eye against, and said "that'll do".
I worked for a PC computer systems supplier in the late 80s and early 1990s. One of our customers designed industrial filters - everything from dialysis stuff to chemical purification meshes for the pharma industry.
They had a range of special 'rare earth metal' mesh filters that would periodically come back for cleaning and refurbishment and as part of the final test, solvents of very specific purity would be used to see how the filters were performing. The rate of filters coming back for refurb was low (a few a month) and there was a small testing room set aside for the purpose.
The problem we were asked to tackle was that because some of the solvents were extremely volatile, corrosive and flammable, no exposed live electrics or 'generic plastics' of any kind were allowed in the room (even though the testing was done in a fume cupboard) - and a PC + CRT was out of the question, so was any form of handheld device. So how to record the test results, which comprised a serial number and a few small fractional numbers?
When we were brought in, we were asked to consider the feasibility of several proposals - from some kind of 'encapsulated' PC and keyboard, to cutting a hole in the wall and having either a working touch screen or 'safe' light pen (in a 'condom'!?) through a glass panel..or 'any other ideas??'
Having carefully considered the volume of work and the small amount of data, we suggested using a pre-printed pad for the results, a pencil, and putting the computer outside the room.
It's a cool project for sure, but i'm not sure how i feel about training people that it's okay to interrupt you while you have your headphones on. That's the only way i get peace in my office.
I had no idea `screen` could do that, and I've been using it for years. Assumed you'd have to use some sort of telnet to communicate with the board. Very cool stuff.
Yup, this is a typical “modern” way to get terminals over serial connections. For instance, my home server is headless, but I can use its onboard serial connection (via GNU screen) to manage it “remotely” even if, for whatever reason, it’s not connected to the network itself, SSH isn’t running, or similar.
One nuisance with cu when used in conjunction with ssh: they both use the same escape character, so I've accidentally terminated my ssh session when I'd meant to disconnect from the serial port many times. But that's what tmux's for. :-)
Tie it into your company's project planning software. Instead of doing this, what it does is anytime someone tries to add something, saying "This should be easy, right?", it adds a month and a half to the gantt chart.
Yeah, i don't think the Dash Buttons were ever hackable other than by intercepting it's network traffic. And there isn't really any way they can stop that.
Cute project. I started getting into working with Arduinos and then ESP8266 devices late last year.
I built a simple "button-thing", but to make it more flexible it just posts a message to an MQ-queue. That way anything on my system can listen for the events, and react.
At the moment it is used to play "alarm.mp3" on my desktop-PC, but in the past it was configured to turn out all the lights in my flat.
I'm planning to do something similar, but mostly to alert me that someone is looking for me at my desk at work. The ESP8266 would connect to our Guest WiFi and sent a request to IFTTT, which would in turn either send me a notification on my phone, or change my blink(1) light to a different color.
It doesn't look like there is a battery in there (and there was no talk of deep sleep..) but at the same time this is not a true USB peripheral, there is a FTDI or similar chip that's talking USB but the ESP is only talking serial with that chip. So there isn't any scope for the ESP taking over its host by pretending to be a keyboard or similar.
But at the same time we are talking about using a 66 MHz processor and WLAN chip to connect all around the world to some Slack data center that then does a bunch of internal handling that is then at some point through another few world roundtrips received at his desk computer, a whopping metre away from the ESP and connected through a serial line. And no shit, all that nonsense is making the simple button press take seconds and for no good reason at all. To relay a hint from people standing a few feet away.
In other words, this fits right in with the IoT crowd!
I don't think he's leaving it plugged into his computer. He's just using the USB to serial connection for programming the board. Once that is done, he's probably plugging the USB cord in for power (although I bet he could have gotten it to last a decent amount of time on battery power).
Looks like it is running a fairly basic react app :shrug: - I'd guess the blog is a side project and doubles as a way to learn new tech. Supporting noscript users is probably prioritized right next to localization and accessibility. Not to say they aren't important, just that implementing that might be a future learning project... as the author's interests / professional growth dictates.
The board in question can be found between $2 and $3. Once you’re at the point of wanting to build "an electronic thing to notify you of something", the actually cost difference in the BOM between the simple approach and the “IoT” approach is basically zero.
EDIT: Sorry, not the exact board, but something that’s basically the same. The usual “value adds” you find on top of the basic wifi board are 1. on-board usb/ttl chip and 2. 5v-3.3v regulator. Still, you can find something with both of those for under $4.
I'd probably do it with a transistor amplifier for fade-in and an RC network for fade-out, because it'd be an easy opportunity to learn how not to suck at analog quite as much as I currently do. But not everyone has parts bins as deep as mine.
Through hole soldering isn't hard, it just takes technique. If you provide the right conditions, the solder will pretty much do what you want it to do.
The actor then complained that he had a hard time seeing the black and white screen so while the stage manager went to see how he could improve his set up the actor asked my father for a penknife, cut a small hole in the set that he could put his eye against, and said "that'll do".