buescher

Just have an AI make a video out of it, I guess.

Because it told you so!

I am reminded of the passage in Umberto Eco's _Foucault's Pendulum_ where they reconstruct a particularly wild weekend for a Templar based on the rule of the Templars and the idea that a rule exists because it's something people do.

Fresh produce is pretty inexpensive here for the most part. But fresh herbs (like thyme, not a euphemism) are sold in small quantities (15g-20g) in plastic clamshell packaging in groceries for $2-$4 each.

You don’t need a $900 gadget to grow fresh herbs, either, of course. But that’s one way you could think of it “recouping” the capital.

You can turn almost all of those buttons off in the settings and save it as a template. The only complaint I ever got was from somone who wanted to use the highlighter instead of the built-in comment management system.

"Can" does not mean "always will without fail". On the flip side, you will never catch things in review if you don't review.

Having a specific review for new parts helps to keep footprint errors from "passing through". I've never seen a general layout review catch a footprint error: if you're used to using internal library parts, no one thinks to check them.

There's another updated version of that prolog here along with some links, including an archived article from Microsoft Research on how it was (once upon a time) used in Windows NT network configuration: https://github.com/opless/small-prolog

I thought so. I'm a fan of Glider.

I personally reach for LTspice from Analog Devices first even if I have better available - it is quick, easy enough to use, and someone has ported the model you need to it already. There is a Mac version on their web site but I have never used it.

Micro-Cap is free-as-in-free-beer but windows-native and will do stability analysis (gain margin/phase margin).

Qucs-S + the backend simulator of your choice is probably your best bet for free software on Mac. Original Qucs is still developed (checkins last month on their github, if you want to try building from source) but hasn't had a release in almost ten years. QucsStudio is now uSimmics and is closed-source and Windows-native. They all have their peculiar uses. I haven't used any of them much.

Online you can use PartsQuest Explore for free for public designs and short simulation runs. I have not used the online version much but it is basically the very powerful Siemens (Mentor) VHDL-AMS simulator that has gone by a number of names (Mentor never commits to names) on the backend. I have used that: it came with PADS Standard Plus. It surprisingly does just about everything at least usably well. It can use spice models or VHDL-A(MS) models. It can use RF s-parameter models. It can use control-theory laplace-transform block models. It will do stability analysis and monte carlo analysis. It can run VHDL models. You can model just about anything you can write the ODEs for in VHDL-AMS: try viewing source on their motor model for example. You won't get the filter synthesis or analysis tools you'll find in a dedicated RF simulator like Keysight Pathwave. You won't get the controls design and analysis tools in the Matlab/Simulink toolkits. But it does a whole heck of a lot.

You can catch it with a footprint review. List all your new footprints or old footprints assigned to new parts, and have someone else check them against the datasheets. Or if you ever have the good fortune to work with a very good layout designer, they will check things like that for you just to catch you out.

Altium has taken over a lot of small to medium sized shops. Mostly because the price is right for its capability. It also has a history of being the least bad compromise between the odd mixtures of excellence and user-hostility Cadence and Mentor tend to come up with, going back to the Protel days, and they've done a good job in the last decade+ of marketing it to those shops. Cadence and Siemens nee Mentor (and maybe Zuken? I've never seen Zuken in the wild, but it always makes these lists) have been neglecting the entry level and smaller organizations and aggressively trying to move their customers to their higher tier offerings during that time. But while it's Altium's flagship product, it is not top tier. It is really entry-level for a professional PCB-level design package, like PADS and OrCAD as opposed to Xpedition and Allegro.

You essentially can practice the classical art of memory on a 2D representation (think why you can remember who was at lunch by thinking about where they sat), and it's useful to have them at different levels. Almost any nontrivial schematic should have an associated block diagram (by this I do not mean it necessarily should be captured at the block diagram level in a "hierarchical design"; these can be more trouble than they're worth). There are other types of diagrams that can be useful, too.

I haven't done much HDL work so I don't have strong opinions there, but let's think about graphical representations source code and software intent for a moment. I like to see some block-level description of firmware even if it's just a few rectangles representing the main loop and each ISR other other process with text indicating what they do. I like to see statecharts. If someone has written an implicit flags-and-conditionals state machine, I'll draw the statechart and start probing for bugs and debugging from there. Tools like Scitools Understand are drastically underrated for figuring out other people's code, incidentally.

What kills me about the aforementioned style is that it eschews the schematic as a tool for thinking and implicitly sees schematic capture as a layer of friction you have to get through to get to layout. I suspect a lot of designs drawn that way were built up point-to-point on a breadboard and never sketched or even otherwise mentally visualized.

There are many. What's your (computing) platform and what are you trying to simulate? The more professional tools (beginning with freebies like LTspice) are oriented around letting you probe waveforms from simulated inputs and are not as oriented towards interactive probing "does this LED light when I press that button?" as a hobbyist might want.

I mean when they put parts down on the schematic and just attach net names to them. You know, the pull-up resistors in the corner with named nets on them, not shown attached to particular pins or buses. Usually with a dotted box drawn around them. Frequently horizontal.

Also usually totally implicit power routing. And no one that draws schematics that way has ever learned the value of drawing IC symbols to fit the circuit as opposed to the physical pinning of the device.

Better simulators can do parameter sweeps and monte carlo simulations so you can explore tolerance stackup issues. But most of the failure modes you've listed there are just why you shouldn't build your circuits for one on breadboards and for another with dodgy parts sourced outside of distribution channels.

Schematics also help explain your circuit because the idioms of drawing them communicate intent. Which is one of the things the Shenzhen/Ladyada/Sparkfun/SeedStudio crowd don't get with their "schematic as data entry for layout" style. Some of them should know better.

Breadboards make me wince a little especially in a professional setting but I've made my peace with dev boards that come with DIP-style pins for them. Decent breadboards are so cheap now that every dev board get to live on its own permanent breadboard.

Some people like Everycircuit and Falstad for that.

I'm not too familiar with the ESP32 ADCs except I remember they're unusually lightly specified even for microcontroller ADCs. If "the knowledge" involves things you couldn't do - or rely on - in production like careful calibration and characterization, that would answer your question.

It's not surprising that people who write professionally would learn one tool to the point it gets out of the way and then not want to change. It's not just sticking with the first thing they learned - there's a constant churn of "tools for distraction-free writing" that address some of the complaints that people that still use older word processors have about more up-to-date systems.

Once you know the pattern, every so often you'll see a piece about a writer or journalist and the funky software they use and you can just wait for it... it's going to be Wordstar, XyWrite, one of the XEDIT editors, sometimes Wordperfect for DOS. Rarely Word for DOS. Neal Stephenson uses emacs, but he's an outlier in a lot of ways. I think there was a piece linked here recently by a journalist who uses the macOS TextEdit for note-taking, which dates back to NeXSTEP. (not exactly the same thing, but consider)

Probably SharePoint lists, even though they’re even less relational. That’s probably a good constraint on most users ironically.

I increasingly use the rubber-ducking I do with an llm as a sort of engineering notebook or complement to one - I wish they had better search features.

There's a digital code as part of the WWV transmissions (!) but you're right that the typical "atomic" clock doesn't sychcronize to the HF stations.

There are weak wwvb simulators out there as phone apps and such that depend on using EMI to sync your clock. Like the old AM radio bus noise music hack. https://github.com/kangtastic/timestation?tab=readme-ov-file...

>because I simply want the person to shut the fuck up

Because you don't want to think about what they have to say, which is why you choose that approach to ending the conversation.

That’s another one that’s true but particularly irritating when used as a cheap shot. I have responses for that like “Is it good when it breaks? Who owns it when it does?” Sort of like “are you fast if you don’t finish? Are you finished if you have to do things over?”

“All models are wrong, some are useful” is frequently misused or misunderstood this way.

You can find the fcc notices in manuals if you search on espressif’s module grantee code which is 2AC7Z. Espressif is extremely widely used.

Yes. Lapack was the successor to linpack and I seem to recall some of the linpack routines going back much further than the eighties. MATLAB (which existed before the commercial release in 1984) was built on linpack.

At this point, I hoard llm chats and projects, and struggle to turn them into action. There are only so many hours in a day.

It depends on the system's potential failure modes and what's required by your safety standard, not on one engineer's opinion of what's "best".

Interesting. I didn't know that.

Not if you go to the cheap "Asian brands" like you're thinking with micros, plus with your cheap micro you'll need a reset controller. And budget isn't all BOM cost.