“Easy” Desk Alert Button (ESP8266 and Staples Easy Button) (opens in new tab)

You can buy the IoT Button and get one tailor-made for custom usage. $20 is pretty pricey though. https://www.amazon.com/dp/B01KW6YCIM

IIRC you have to sniff for its ARP probe, it works most of the time but it's kinda finicky. An ESP8266 is a similar price.

Yep: http://kellysutton.com/2017/07/06/building-a-briefcase-that-...

I really like the flic buttons and use them a lot, but they are a bit more expensive.

https://flic.io/

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.

Soldering takes 10 minutes to learn and about a $15 investment in a soldering iron.

Look at the button, it has the word "Easy" on it. It's literally called an Easy Button.

This is beginner level stuff. Not so hard to learn with youtube available.

Well having to pay attention to every Slack notification invites its own sort of distraction.

Just hide the button if you don't want to be interrupted!

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.

It's the only remaining argument for `screen` over `tmux`. I personally use `cu`: `cu -l /dev/ttyUSB0 -s 115200`. (within tmux)

Minicom is also useful for serial coms. Includes xyzmodem transfer as well.

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.

Wow, I couldn't believe it with how minimal the page is, but yeah this is insane, it's just a white page with javascript disabled.

It won't even load at all on the Ice Cream Sandwich system browser.

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.

Because that wouldn't be as much fun.

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.

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.

The canonical solution here is a cue light. If someone offers to buy me a CCTV system, I won't stop them, but that seems rather heavyweight.

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).