4 ms·
The idea was of course that someone else, one with working brain could run the example script and discover the underlying issue, which is weird and more complex
by timonoko 9mo ago
The idea was of course that someone else, one with working brain could run the example script and discover the underlying issue, which is weird and more complex.
- Neywiny 9mo agoA great way to get help is to ask for it. I'd imagine the ffmpeg community would be a lot more open to helping you if you didn't start off by saying their tool is broken, though
- timonoko 9mo agoIt is clearly broken and very-very-very deeply so. On organizational level. Quote Gemini: This error is the "smoking gun." It confirms that no matter how we quote the command, your shell is treating the comma as a separator for the FFmpeg command itself rather than a character inside the filter.
- iAMkenough 9mo agoYour reliance on the random word generator is making the point of OP.
- rigrassm 9mo agoSoooooooooo, your SHELL is having issues with quotes, not ffmpeg.
- whateveracct 9mo agothis is what happens when you let people who don't know what they're doing think they do
- nolok 9mo agoYou are completely missing his point with his answer. His point was less about ffmpeg being broken or not, and more about the fact that if you can't have an opinion or input of your own beside parroting a advanced Markov chain generator, then you don't matter in the discussion.
- whateveracct 9mo agoWouldn't be surprised if Gemini is proofing the HN comments too.
- Neywiny 9mo agoAssuming I'm the he, that last part isn't exactly what I meant. I meant that there's not a good faith effort in partaking in the conversation. I disagree with blindly following the LLM but even if you do there are ways to productively get help with it.
- Neywiny 9mo agoMaybe I need to take a step back. If, and heavy on the if, you want to communicate issues with humans, you're doing it wrong. If you want to rely on LLMs (and look where that's getting you), by all means do it that way. With humans, they need to come to that conclusion on their own. An example is like this: You: "I would like to remove blurry frames with ffmpeg. I think I want to use xyz method. Can anybody please help me craft the command?" Them: "yes, here's the command" (and maybe it's exactly what the LLM output) You: "it's giving this error code. Here's a snippet of my input file. Can you please check if that works on your end?" Them: "oh wow you're absolutely right, this doesn't work, I'll put in the issue ticket" This is a lot nicer than "something else told me how to use your tool and it's broken".
- vegavis 9mo agoWith all due respect, you're trusting LLMs to a delusional level... they love talking about "smoking guns", that part made me laugh. Like another user said, sounds like its admitting the issue is in the shell anyways. They dont have intimate documentation knowledge of everything, use it more as a launch pad to test ideas and better shape your understanding of a subject knowing that it gets things wrong alot. Watching the zombifying effects of AI "think" for others in real time is depressing.