This is also a typical one in electronics, very few people understand how ferrites work, the frequency bands where they are effective and the difference between common mode vs differential mode.
This equipment was deployed in remote places without any kind of connectivity, sometimes not even cell coverage.
But the real problem is that the frequency of this failure in a single device was much lower than that. There were hundreds of these devices deployed and we never had one particular unit that was triggering all the time, sometimes here, sometimes there. Really a nightmare.
Yeah, for sure if I was perfect I wouldn't make mistakes, but I am not.
However, I have to disagree in one aspect, it is not basic electrical engineering practice to add filter capacitors on every node. Some nodes need it, some don't. This one, by the topology of the circuit didn't, but from the need for immunity did. This is not always obvious, but for sure I could have anticipated.
Well, you're better than me than. Mistakes like those are @$%#ing expensive, so they have me questioning WTF I'm even employed sometimes. I wasn't the only one looking for the issue so maybe you're right, but it feels really bad.
I know exactly what you mean, I've been there myself so many times that I have learned be humble just for this one reason. I have a post about good engineers being humble which is based on this idea.
If you think this way, you are already on a good path. Think that there's a lot of folks that don't even care at all, you do care and want to improve, that sets you apart.
I do use AI tools in my writing process — the ideas, stories, and opinions are mine (20+ years of them), but I use AI to help structure and polish the prose.
I think this is where most technical writing is heading, and I'm still finding the right balance between efficiency and keeping my natural voice. Appreciate the feedback.
I don't believe so. The irony is that this slop-fest will actually make people want to have a controlled part of the internet whereby one must certify and prove a) who they are b) they didn't produce a piece of work purely from a slop-machine.
> I think this is where most technical writing is heading...
Not if you want anyone to actually bother reading it. I want to read what you have to say, flaws and all. Not what comes after the slop machine did a pass on your work.
Love this, will try without any AI assistance and see how it goes. Thanks for the feedback, I'm starting to write after all, who knows, maybe I'm an even bigger slop machine myself...
I kept this one intentional as a thesis statement, but the concrete examples live in the companion pieces — like a bug that took 7 years to find because I kept assuming my mental model was right (https://agilitza.com/blog/the-7-year-bug-that-took-3-minutes...).
That one's the humility and persistence lesson in action.
I did have this situation several times in the past. One thing that has worked well for me is to be a bit of a stubborn mofo in challenging people with arguments and not let go so easily.
I don't think you need to be arrogant, but you do need to know how to argue rationally. At least this works quite well with germanic-origin folks, maybe not so well in other contexts.
Another thing that helps is reputation, you do need to build that and that allows your arguments to be heard. No need really to be arrogant and loud IMO still.
This reputation thing, also brings other problems, because it's not good when people don't challenge you, then you need to find skeptics. This is very necessary when you do high end stuff.
I love when a simple solution fixes something.