wtf thats terrible

the covid chart

I'm always amused when questions assert "[Core feature used by thousands of developers] doesn't work" - gee, you'd think some of those other developers might have noticed if that were the case. Maybe it's just the way you personally are trying to use it? 🙄

The other greatest hit being "I've tried everything" - welp, I guess we shouldn't bother answering, since you've already tried everything including whatever we're going to suggest, so clearly what you're trying to do must be 100% impossible. 😜

It's an expression of frustration. I have some of those experiences too. "Install tsnode, run TypeScript files like nodejs runs JavaScript files", easy. Except "it doesn't work" (on my system, not the way they described it in the documentation) and "I tried everything" (everything I could think of, including various tricks others had success with). While inaccurate communication is bad I'd give them a pass for having had such a bad experience.

Then again I also understand the ... [I forgot the word, I'll replace it with something else] appeal of making fun of them for being obviously wrong. It's just not helpful.

Yeah, that's why I'm venting my own frustration here rather than commenting on the post. 😉

@nwp I think "temptation" was what I failed to remember in the moment.

I need to level up my speech skill. Or stop reading isekais. One of those.

In Designing Games, Tynan has a section about how humans in general are terribly overconfident estimating things. He was specifically addressing planning in game development (how long will X feature take, etc), but I've noticed it applies to a lot of other things as well.

In this case, I've tried everything & X thing is broken both seem to fit. Person had their own personal data & proceeded to extrapolate wildly.

Well, if you have a single data point then whatever the value of that data point is will be your best guess at the true value.

Incidentally, the book is a great read & the chapter about planning & estimating has a somewhat humbling quiz. Pre-test I wondered "how can everyone not score 100% on this"? Which in itself is yet another example of overconfidence in estimating :)

@nwp Yes, but one needs to acknowledge there may be other data points they haven't encountered. The problem in mentality comes when there's a slip from "nothing I've tried works" to "nothing can possibly work".

There is always data you haven't encountered and there are diminishing returns when looking at more data. Obviously a single data point shouldn't be the end, but finding X people who tried to get Y working to get a better estimate at how easy/impossible a thing is is just not feasible.

Also I don't think that slip does much damage. It's technically incorrect, but there is no chance for a misunderstanding, so as far as mistakes go that one arguably isn't even worth fixing.

I'm not saying that you need to find X people. I'm saying that assuming that other data points can't exist tends to exclude the exploration of useful solution spaces.

I don't think anyone thinks other data points can't exist. That barely makes any sense in the context of having tried to get some software to do what you want. Clearly someone has written it and must have tried it at some point.

I don't agree. "I've tried everything" directly contradicts "Core feature used by thousands of developers" . Clearly those developers have done something I haven't tried as they got it to work. Assuming otherwise isn't constructive.

Right, you're hung up on the technicality.

I do agree with you about diminishing returns. That's a sandpit I frequently wander into & try to dig my way out of.

Another issue is that the whole point of using other peoples' software is to save time. If you spend more time trying to understand the library than it would be to write your own then the library is worthless, so it makes sense to give up at some point.

"This library is broken and impossible to use" doesn't mean that it's literally impossible, it's just beyond reasonable to do so.

And the fact that other developers have succeeded where you failed just feels bad and doesn't help.

Not to mention that people use the same libraries for different things. Sometimes a library just doesn't support your use-case but supports most developers' use-case perfectly fine, making the claim "This library doesn't work" just inaccurate because it's missing "for my use-case".

Though you don't even know that without wasting your valuable time on pointless analysis.

Yes, I can relate to that. I've found libs that I probably could have eventually gotten to work, but decided that doing so wasn't likely to save me time compared to doing it myself. That's sometimes a very tough call for me to make.

Especially for math / computational geometry stuff.

And then sometimes people say, for example, that JavaScript is just completely broken and unusable. They have very good arguments and the fact that however many millions of developers use it with some success doesn't mean they are wrong.

In web apps, a Date value isn't a legal parameter. getValues() fails to return data to a web app if the range contains a cell with a Date value. Instead, transform all the values retrieved from the sheet to a supported JavaScript primitive like a Number, Boolean, or String.

Nice. So not just one based index, but one OR zero based indexing depending on context. And data that might not actually be present. And some data that is forbidden.

One the bright side, I can toggle between light & dark theme on the posted code examples in the documentation, so I guess it's all okay in the end. /s