Article note: They're rolling out smart meters slightly faster than the comms protocol they chose is sunsetting. That could only be more Internet-of-Shit if they did something goofy - like turned off utility service, or started trying to connect to open WiFi which can then control them - when the connection went down.
Public Accounts Committee demands timetable for replacements, because things have run so smoothly so far...
A gaggle of MPs are calling for government to put together a timetable for the replacement of millions upon millions of smart meters that will be defunct when 2G and 3G mobile networks are switched off.…
Article note: There was a 2012 interview with Bill Joy (Sun, vi, etc.) and Danny Hillis (Thinking Machines, Long Now, etc.) where they talk about this idea as "The Coming Entanglement" and it keeps being...right.
https://www.scientificamerican.com/podcast/episode/the-coming-entanglement-bill-joy-an-12-02-15/
It's certainly the case that from an educational standpoint, the computing folks that grew up in the 2nd half of the 20th century mostly have pretty deep constructed models of computers because they encountered computers they could understand at an age when they had the time and plasticity to understand them, and _had_ to understand if they wanted to do things (Want sound? Go 3 rounds with your IRQ setup to get that stupid sound card to work! Want multiplayer games? Guess who's building out a network and setting up port forwarding if you're lucky enough to have broadband...). The under 30 set came up with computer-as-magic-mirror, where the interfaces were carefully constructed to not expose any model of its internal function, and the internals are so complicated that even experts can't remotely fit the whole thing in their head. AND the set of students is much less tilted toward enthusiasts, many of the CS/CPE/EE students arriving at universities are there because they were told they're expected to go obtain a college degree and those disciplines seem like a relatively easy way to land a high paying career, not because of any preexisting interest or proclivity.
....and it causes a lot of problems NOW because the people doing the curriculum design started from a much richer model and higher proportion of preexisting enthusiasts than the current crop of students, and LATER because we don't have enough engineers with enough skills and perspective to see us through the next set of problems.
Article note: One of those issues that periodically comes up in the ECE department is that UK's EE/CPE students don't typically get any exposure to PLCs, and a lot of potential employers would like it if they did. The ones that take power classes get a little bit of the programmable controllers power people weirdly call relays even though they contain neither the contacts no any electromagnetic actuators, but that's not quite the same, and those are _all_ super vendor specific.
This particular Arduino one doesn't look very compelling, but it made me look around. Since PLC stuff is standardized under IEC 61131, an open toolchain (like the autonomylogic OpenPLC editor and runtime) would make a perfectly good teaching tool and not cost an arm and a leg for devices and hook us to some vendored environment forever.
On the surface, a programmable logic controller (PLC) might seem like nothing more than a generic microcontroller, perhaps outfitted to operate in industrial settings with things like high temperatures or harsh vibrations. While this is true to some extent, PLCs also have an international standard for their architecture and programming languages. This standard is maintained by the International Electrotechnical Commission, making it so that any device built under these specifications will be recognizable to control engineers and maintenance personnel worldwide. And, if you use this standard when working with certain Arduinos, this common platform can become a standard-compliant PLC as well.
The IDE itself supports programming ladder diagrams, functional block diagrams, and other programming systems covered under the IEC 61131-3 standard. Not only that, it allows the combination of these types of PLC programming with Arduino sketches. The system offers many of the perks of PLC programming alongside the familiar Arduino platform, and supports a number of protocols as well including CANOpen, Modbus RTU, and Modbus TCP. It can also be used for monitoring a PLC system, essentially adding IoT capabilities to existing systems, enabling continuous monitoring, debugging, and program updates.
While not every Arduino is a great platform to build a PLC around, there are a few available for those looking for a system a little less proprietary and a little more user-friendly than typical PLC systems tend to be. There’s a reason that PLCs are built around an international standard and generally have certain hardware in mind to run it, though, and this comparison of a Raspberry Pi with an off-the-shelf PLC goes into detail about why certain components aren’t good choices for PLCs.
Article note: Interesting.
I love how after decades of "Doubles required for srs bsns" (even though it's rarely the case that doubles will save you if your operation stack-up breaks on singles) we're getting all these ultra-low-precision formats for approximate bullshitting.
Curious that everything except E5M2 has no defined infinities.
I think it does cover pretty much all the sane uses of bits in the narrow sizes.
Article note: I've used a bunch of student's iPads with styluses, and they are broadly the least-shitty stylus experience I've ever had, but ...
The Gen1 pencils are ALWAYS missing their cap and/or have connector issues, because that was a terrible design.
The double tap to swap tools feature is both necessary for it to not be frustrating to edit things, and annoyingly inconsistent.
I can't imagine the experience being half as good without pressure sensitivity or wirelessly charging when docked.
The Gen2 pencil is the _only_ one that doesn't have some kind of bewildering design failure to it.
Three Pencils were given to the iPads. | Image: Apple
A third, cheaper Apple Pencil now sits in Apple’s lineup of styluses, giving iPad owners more choice than ever. And yet, that choice is fraught with compromises and caveats. There’s still not one Apple Pencil to rule them all, and that’s a problem for shoppers.
Let’s run down what was announced today. The new Apple Pencil is $79, can be magnetically attached to the side of your iPad, and no longer needs to be plugged directly into the iPad’s charging port for power, like the 2015 original. On the newest Pro models, it also supports the hover feature that shows where your Pencil is before you actually touch the screen. These are all good things!
But Apple made some strange omissions, too. Pressure sensitivity, a headlining feature of both...
Input conditioning is one of those things that snares novice designers, causes late-process changes that overrun expectations on cost and board area, and traditionally isn’t terribly well taught to EE/CPE students.
It’s on my mind because next week is the point in the semester where I drag UK’s current crop of EE/CPE sophomores through a lab exercise I designed about 5 years ago to drive home gate delays, static hazards, switch bounce, etc.
While I was thinking about it, an upperclassman who regularly digs up neat stuff sent me the cleverest input conditioning circuit I’ve ever encountered.
The circuit comes from the late, great Don Lancaster of TV Typewriter fame, who in addition to his published designs, wrote and self-published a number of instructional/reference books. He had a well-deserved reputation for clever, cheap, robust circuit designs, and this particular trick is the highest wizardry.
Here’s the whole circuit diagram from the text:
The design comes from his CMOS Cookbook (PDF, link to his own hosted copy of the 2nd ED), on p.317 amid a discussion of Flip-Flops and Clocks. It is presented as “An Alternate-Action Push Button” which is entirely correct but really undersells how clever it is, and has apparently been in there since the 1st edition in 1977.
The fundamental trick is that it’s a master-slave Flip-Flop where the capacitor is the master storage element, and the pair of feedback-coupled inverters is the slave. The cap tanks the next state based on the output of the first inverter when the switch is open, and induces it on the inverters on switch close. This means, in addition to latching/toggling, it de-bounces, because the capacitor sets the time constant for hysteresis. It conditions, because the load sees the output of the second inverter. No race conditions or potential oscillations, because the cap can’t charge/discharge while the switch is held. No charge is moving inside the mechanism at steady state, so it’s not leaking power. It’s brilliant.
It is only suitable for relatively slow human-scale edges, so probably not a good method for encoders or the like. You can manipulate the time constant for the de-bounce by changing the value of the capacitor, but only down to a few 10s of nF (depending on what kind of inverter you use) before it gets marginal because it doesn’t have the charge to reliably throw the input of the first inverter.
Not only is it ridiculously cheap and simple as presented, which I think intends a 4067 or 74HC04, you can built it out of anything. Any inverting CMOS gate will work. Any inverting TTL gate will work. Ridiculous old RTL or DTL inverters work. A pair of N-Channel FETs (another CircuitJS link, has an extra transistor on the output for integrity reasons) with pullups to build your own cruddy NMOS inverters works. As would P-channels with pull-downs, or BJTs with resistors for constructed RTL (though doing it that way is leaky), or various other assemblages of tiny mass produced minimum cost components to make it even more minimal (though maybe not cheaper in a modern context).
I appreciate a clever domain-crossing design, and this is the highest form.
Article note: I went to find documentation about the current state of nommu Linux systems the other day and things are _rough_ documentation-wise and mostly alarmingly out of date. This (and its sister article) actually strings together all the details in a comprehensible way, which is super nice.
In the vast majority of cases, running a Linux-based operating system involves a pretty powerful processor with a lot of memory on hand, and perhaps most importantly, a memory management unit, or MMU. This is a piece of hardware which manages virtual memory, seamlessly giving each process its own memory sandbox in which it shouldn’t be able to rain on its neighbours’ parade. If there’s no MMU all is not lost though, and [Uros Popovic] gives us a complete guide to building the MMU-less μClinux on a RISC-V microcontroller.
The result is something of a Linux-from-scratch for this platform and kernel flavour, but it’s so much more than that aside from its step-by-step explanation. It’s probable that most of us have heard something of μClinux but have little direct knowledge of it, and he leads us through its workings as well as its limitations. As examples, standard ELF binaries aren’t suitable for these systems, and programmers need to use memory-safe techniques.
Whether or not any of you will run with this guide and build a tiny MMU-less Linux system, anything which expands our knowledge on the subject has to be a good thing. it’s not the first time we’ve seen a RISC-V microcontroller turned to this task, with a nifty trick to get round the limitations of a particular architecture.
Article note: Automated background metadata scanning is a huge source of vulnerabilities in general.
It's parsers. Hand-coded in C. That the caller probably didn't look at, because it's in some library from the dark ages. For old poorly structured formats. Running unattended in system processes.
I had nothing to do with this but I’m super excited. There has been a long-standing bug in the CPU emulation in Basilisk II (a 68K Macintosh emulator) that could be reliably triggered when *inhales* Bruno the shark killed you … Continue reading →
