Then I guessed "toad" for what turned out to be "a false toad (Telmatobufo bullocki)" and my head exploded a bit

(New installation work is maybe less interesting, but if you're the sort of developer who enjoys logic-ing their way through understanding why a complicated, undocumented system is behaving the way it's behaving, and why did the last guy decide to connect this to that, a hundred year old house that's been gradually updated from knob-and-tube might be just the thing.)

He wore a full motorcycle-style helmet with a mirrored face visor which he refused to open; it was like talking to a member of Daft Punk. It had fake hologram bullet-holes on it.

More importantly, he skied like an asshole: cutting people off, sudden stops or changes in direction without checking if anyone else was coming, cutting ahead in the lift line, stopping to readjust his gloves or whatever directly in the lift exit ramp, shouting at people who he felt were in his way -- just totally self-absorbed and borderline dangerous.

It turned out to match his management style 1000%.

The original post and discussion was an eye-opener for me; before that I never understood why some people would say "yes" to a request but then act put upon anyway, or would act vaguely like they wanted something but never actually come out and say so. I just thought they expected everyone to be a mind-reader.

Once I understood they were basing things on the premise that putting someone in a position of having to say "no" was rude, it all made a lot more sense, and I was able to adjust my own behavior and expectations to better fit theirs.

https://ask.metafilter.com/55153/Whats-the-middle-ground-bet...

(To me the most interesting thing about the concept is that you can immediately tell from people's reaction to it which category they personally fall into.)

https://myactivity.google.com/u/1/activitycontrols

Disabling a setting asks if you also want to delete your history for that setting; if you say yes it shows a sample of what you're deleting, you confirm it, and it's done. Nothing tried to encourage me to stick around, keep my data, they could easily have let me turn off collection of new data but not mentioned that they still have the existing data around -- I expected trickery but found none.

(Except for one small detail: I can't figure out how one is supposed to discover that this page exists in the first place; the support update we're discussing contains a link to it, but I can't find it anywhere in the YouTube preferences or advanced settings or anywhere else.)

(Oh, and it had no visible effect on my YouTube home screen, I'm not seeing the empty page I hoped for, just the same garbage videos it usually prompts me with for no clear reason. So there's that too.)

* "Do work, but get paid only if we use it" feels potentially exploitative. You do mention a "base pay" rate but I wonder how it compares to the "merged PR" rate?

* Code quality and adherence to internal code standards is challenging for external contributions. Open source projects, of course, are already used to this, so it's not surprising to see that's where the bulk of your current customer base lives; it's more difficult to see how this would work with other types of organization -- I'd be concerned that I'm spending more time in code review and such than I'd gain in contributed lines of code.

* The whole setup seems to fit in an uncomfortable middle ground between just hiring an offshore developer directly or through a traditional outsourcing firm, on the one hand; and hourly piecework on the other. The only advantage I see to this sort of arrangement on the hiring end compared to regular old long-contract outsourcing is that I don't have to pay for work I don't end up using, but that circles me right back to the "exploitative" part.

Note to developer: you may want to add an `onbeforeunload` handler

I finally gave up on that one when it got caught in a loop somehow where it apologized for giving me the wrong line of code for an import, gave an obviously wrong explanation for why it didn't work, then "corrected" it to the exact same line of code.

Another attempt I was asking it to compare different ways for measuring the amount of difference between data trees -- it did give me the names of a couple of different algorithms, and very wordy, plausible-looking descriptions of how each of them worked... neither of which was terribly helpful, because they both boiled down to "recursively examine the tree and tally up the differences."

Asked for an implementation example, it gave me code that expected input arrays of pre-calculated edit costs, and suggested I write my own function to convert the tree data into that array format.

So that one was extra weird, in that it wasn't wrong, just unhelpful, like here I'll do the easy part for you and leave the thing you asked about as an exercise for the reader.

I dunno, maybe with practice I could learn how to drag it towards helpfulness, but for now RTFM still seems easier.