Against The Cloud

February 19, 2022 by Elliot Williams 58 Comments

One of our writers is working on an article about hosting your own (project) website on your own iron, instead of doing it the modern, cloudy-servicey way. Already, this has caused quite a bit of hubbub in the Hackaday Headquarters. Who would run their own server in 2022, and why?

The arguments against DIY are all strong. If you just want to spin up a static website, you can do it for free in a bazillion different places. GitHub’s Pages is super convenient, and your content is version controlled as a side benefit. If you want an IoT-type data-logging and presentation service, there are tons of those as well — I don’t have a favorite. If you want e-mail, well, I don’t have to tell you that a large American search monopoly offers free accounts, for the low price of slurping up all of your behavioral data. Whatever your need, chances are very good that there’s a service for you out there somewhere in the cloud.

And that’s awesome if you only want the service provided. But what if you want to play around? Or learn how it all works under the hood? This is Hackaday!

For instance, you could run your own mail server just for your friends and family. The aforementioned search monopolist will probably flag all of your e-mail as spam, partly because they don’t trust small e-mail providers, and partly because that’s the “m” in monopoly. But if you can get folks to whitelist the addresses, you’ll be in business. And then you open up a world of fun and foolery. You can write hooks to automatically handle mail, or you can create an infinite number of mail accounts, even on the fly as per Spamgourmet, the most awesome anti-spam tool of the last 30 years. Or you can invent your own. Run a mailing list for your relatives. Or do something stupid.

I used to run a service where, when a particular account received an e-mail, the attached photo was pushed up to a website with the subject line as the caption. Instant photo-blog, of the strangest and least secure sort. Getting it running was a few lines of Bash scripting, and an afternoon of fun. Is there a service that does this, already existing in the cloud? Probably. One that allows you a little privacy and doesn’t track your every move? Maybe. But even if there is, would I have learned about sendmail by using this service? Nope!

I hear you saying “security” under your breath, and you’re right. This system was secured by lock made of purest obscurity. But still, in seven years of running the service, nobody guessed the magic e-mail address, not once. Knowledge of the e-mail address was essentially a password, but if I needed extra security I probably could have implemented it in a few lines of Bash anyway. The webpage itself was static HTML, so good luck with that, Hackerman! (The site’s been down for a while now, so you missed your chance.)

If you just want a service, you can be served. But if you want to be a server, a first-class Internet citizen, with your own cloud in the sky, nothing’s stopping you either. And in contrast to using someone else’s computers, running your own is an invitation to play. It’s a big, Internet-connected sandbox. There are an infinity of funny ideas out there that you can implement on your own box, and a lot to learn. If you hack on someone else’s box, it’s a crime. If you hack on your own, it’s a pleasure.

I know it’s anachronistic, but give it a try. (PDF, obscenity, uncorrected typos.) Be your own cloud.

The Real World Strikes Back

February 12, 2022 by Elliot Williams 17 Comments

My son was into “Secret Coders“, a graphic novel series wherein a pair of kids discover and thwart a plot to take over the world by learning to program in the LOGO computer language. When I told him that these “turtle bots” were originally actually real physical things, he wanted one. So we built one out of some nice geared DC motors I had lying around.

A turtle bot has essentially three jobs: move forward in a straight line a controlled distance, turn a given number of degrees, and raise and lower a pen. If you’re already screaming “use stepper motors!” at your screen, well, you’re probably right. But I had these nice Faulhaber/Micromo geared motors with encoders that were just collecting dust in the closet, so I used ’em. And because of that, the robot stumbles on two of its three goals in life — the servo pen lifter works just fine.

Perfectly matched DC motors don’t exist. Of course I knew this, because I’ve built bots with DC motors before. But they’ve all had complex control mechanisms and/or feedback that made it moot. Not here. This bot needs to drive perfectly straight without any lines to guide it or more interesting navigation algorithms.

We spent a good half hour driving it around in not-quite-but-almost squares, tweaking each side’s PWMs, running the motors backwards for short bursts to brake the wheels, and generally trying to map degrees of rotation to milliseconds of motor drive. And you know what, my son enjoyed it. The concepts were simple enough for a second grader, and guessing the right PWM values was like a game. When we finally got it good enough, there was a small celebration.

Of course I know that what it really needs is encoder feedback. I installed those encoder gearmotors on purpose after all. But dealing with quadrature and probably a PID loop to control and sync the two sides is not for my son, at least not for another couple years. (They do learn closed-loop control theory in fourth grade these days, right?) I’ll have to do that all offline some night while he’s sleeping.

But I hope he’ll remember the lessons learned from stabbing at it the naive way. Abstractions are great, but no two motors are ever perfectly alike. You’d think you could just calibrate it out, but the motors differ in driven and coasting behavior, so you’ve got a lot more calibrating to do than you think at first. The real world is tough, and although it’s important to have theory and ideas and abstractions to guide you, you’re going to have to tweak to make it work when the wheels hit the floor. But also that it’s fun to do so, and super rewarding when it finally draws a wonky square.

The Weirdest Hack

February 5, 2022 by Elliot Williams 31 Comments

I was on the FLOSS podcast (for the Episode of the Beast no less!) and we were talking all about Hackaday. One of the hosts, secretly Hackaday’s own Jonathan Bennett in disguise, asked me what the weirdest hack I’d ever seen on Hackaday was. Weird?!?!

I was caught like a deer in headlights. None of our hacks are weird! Or maybe all of them are? I dunno, it certainly depends on your perspective. Is it weird to build a box that makes periodic meowing noises to hid in a friend’s closet? Is it weird to design new and interesting wheels for acrobats to roll themselves around in? Is it weird to want a rainbow-colored USB DIP switch? Is it weird that these are all posts from the last week?

OK, maybe we are a little bit weird. But that’s the way we like it. Keep it weird and wonderful, Hackaday. You’ve got enough normal stuff to do eight hours a day!

Hacking Is Hacking

January 15, 2022 by Elliot Williams 5 Comments

Tom Nardi and I had a good laugh this week on the Podcast when he compared the ECU hacks that enabled turning a VW with steering assist into a self-driver to a hack last week that modified a water cooler to fill a particular cup. But it’s actually no joke — some of the very same techniques are used in both efforts, although the outcome of one is life-and-death, and the other is just some spilled ice-cold water.

This reminded me of Travis Goodspeed’s now-classic talk “In Praise of Junk Hacking” from way back in 2016. For background, this was a time when IoT devices and their security were in their relative infancy, and some members of the security community were throwing shade on the dissection of “mere” commercial crap. (Looked back on from today, where every other member of a Botnet is an IP camera, that argument didn’t age well.)

Travis’ response was that hacking on junk lets us focus on the process — the hack itself — rather than getting distracted by the outcome. Emotions run high when a security flaw affects millions of individuals, but when it’s a Tamagotchi or a pocket calculator, well, it doesn’t really matter, so you focus on the actual techniques. And as Travis points out, many of these techniques learned on junk will be useful when it counts. He learned about methods to defeat address-space randomization, for instance, from an old hack on the TI-85 calculator, which garbage-collected the variables that needed to be overwritten.

So I had junk hacking in the back of my mind when I was re-watching Hash Salehi’s great talk on his work reverse engineering smart meters. Funnily enough, he started off his reverse engineering journey eleven years ago with work on a robot vacuum cleaner’s LIDAR module. Junk hacking, for sure, but the same techniques taught him to work on devices that are significantly more serious. And in the craziest of Hackaday synergies, he even hat-tipped Travis’ talk in his video! Hacking is hacking!

The End Of The Electromechanical Era

December 18, 2021 by Elliot Williams 141 Comments

When viewed from the far future, the early years of the 21st century will probably be seen as the end of a short era in human technological development. In the beginning of the 20th century, most everything was mechanical. There were certainly some electric devices, but consumer products like gramophone players and “movie” cameras were purely mechanical affairs. You cranked them up, and they ran on springs. Nowadays, almost every bit of consumer gear you buy will be entirely electronic. In between, there was a roughly 50 year period that I’m going to call the Electromechanical Era.

Jenny List’s teardown this week of an old Fuji film movie camera from 1972 captures the middle of this era perfectly. There’s a small PCB and an electric motor, but most of the heavy lifting in the controls was actually put on the shoulders of levers, bearings, and ridiculously clever mechanisms. The electrical and mechanical systems were loosely coupled, with the electrical controlled by the mechanical.

I’m willing to argue the specifics, but I’d preliminarily date the peak of the Electromechanical Era somewhere around 1990. Last year, I had to replace all of the rotted rubber drive belts in a Sony Walkman WM-D6C, a professional portable tape player and recorder produced from 1984-2002.

It’s not a simple tape recorder — the motors are electronically regulated to keep ridiculously constant speed for such a small device, and mine has Dolby B and C noise reduction circuitry packed inside along with some decent mic preamps. But still, when you press the fast-forward button, it physically shoves rubber-coated drive wheels out of the way, and sliding pieces of metal make it change modes of operation by making and breaking electrical contacts. Its precision lies as much in the mechanical assemblies and motors as in the electronics. It’s truly half electronic and half mechanical.

But that era is long over. The coming of the CD player signaled the end, although we didn’t see it at the time. Sure, there is a motor, but all the buttons are electronic, and all the “mechanism” is implemented almost entirely in silicon. The digital camera was possibly the last nail in the Electromechanical Era’s coffin: with no need to handle physical film, the last demand for anything mechanical evaporated. Open up a GoPro if you don’t know what I mean.

While I’ll be happy to never have to replace the drive rubber in a cassette recorder again, it’s with a little sadness that I think on the early iPods with their spinning metal hard drives, and how they gave way to the entirely silicon Zoom H5 recorder that I use now. It has a S/N ratio and quiet pre-amps, no wow or flutter, and a quality that would have been literally unbelievable when I bought the WM-D6C.

Still, if you find yourself in the thrift store, and you’ve never done so before, buy and take apart one of these marvels from a bygone era. A cassette recorder, even a cheap one, hides a wealth of electromechanical design.

Ordering prototypes like they were fast food

Has DIY Become Click And Buy?

December 11, 2021 by Elliot Williams 54 Comments

We are living in great times for DIY, although ironically some of that is because of all the steps that we don’t have to do ourselves. PCBs can be ordered out easily and inexpensively, and the mechanical parts of our projects can be ordered conveniently online, fabricated in quantity one for not much more than a song, or 3D printed at home when plastic will do. Is this really DIY if everything is being farmed out? Yes, no, and maybe.

It all depends on where you think the real value of DIY lies. Is it in the idea, the concept, the design? Or in its realization, the manufacturing? I would claim that most of the value actually lies in the former, as much as I personally enjoy the many processes of physically constructing the individual parts of many projects.

For instance, I designed and built a hot-wire CNC foam cutter recently. Or better, I designed a series of improved versions, because I never get anything right on the first try. All along the way, I 3D printed new and improved versions of the plastic parts, ironing as many of the little glitches out as I had patience for. This took probably a good handful of weekends’ time, spread out over a couple months, but in comparison to time spent testing, fixing, and redesigning, very little time or effort was spent in the physical building.

Moreover, I bought most of the parts at the hardware store. The motor controller shield and cheap Arduino clone came from eBay. And even those that I did manufacture myself, the 3D-printed bits, were kind of made by a machine — my experience of the whole process wouldn’t have been any different if I ordered them out.

Of course craftsmanship still exists, and we see that in Hackaday projects all the time. Heck, I’ll admit that I still enjoy a lot of the process of making things with my own hands for its own sake. It’s peaceful. But if there’s one thing that the rapid proliferation of ideas and projects that have been facilitated by 3D printing and cheap short-run PCB services, it’s that the real value of many projects lies in the idea, and the documentation. Which is to say, I gotta get around to writing up that foam cutter…

Separating Ideas From Words

November 6, 2021 by Elliot Williams 11 Comments

We covered Malamud’s General Index this week, and Mike and I were talking about it on the podcast as well. It’s the boldest attempt we’ve seen so far to open up scientific knowledge for everyone, and not just the wealthiest companies and institutions. The trick is how to do that without running afoul of copyright law, because the results of research are locked inside their literary manifestations — the journal articles.

The Index itself is composed of one-to-five-word snippets of 107,233,728 scientific articles. So if you’re looking for everything the world knows about “tincture of iodine”, you can find all the papers that mention it, and then important keywords from the corpus and metadata like the ISBN of the article. It’s like the searchable card catalog of, well, everything. And it’s freely downloadable if you’ve got a couple terabytes of storage to spare. That alone is incredible.

What I think is most remarkable is this makes good on figuring out how to separate scientific ideas from their prison — the words in which they’re written — which are subject to copyright. Indeed, if you look into US copyright law, it’s very explicit about not wanting to harm the free sharing of ideas.

“In no case does copyright protection for an original work of authorship extend to any idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work.”

But this has always been paradoxical. How do you restrict dissemination of the papers without restricting dissemination of the embodied ideas or results? In the olden days, you could tell others about the results, but that just doesn’t scale. Until today, only the richest companies and institutions had access to this bird’s eye view of scientific research — similar datasets gleaned from Google’s book-scanning program have trained their AIs and seeded their search machines, but they only give you a useless and limited peek.

Of course, if you want to read the entirety of particular papers under copyright, you still have to pay for them. And that’s partly the point, because the General Index is not meant to destroy copyrights, but give you access to the underlying knowledge despite the real world constraints on implementing copyright law, and we think that stands to be revolutionary.