argumentum populum would not apply here, since that one is based off populous opinion and I'm making the logical guess that since fish is the least used shell of the three we have talked about that it's usage would be proportional as well. This might not be the case obviously which was why I was asking if you did share your scripts. Lack of usage was my biggest reason for bailing on it. I do a lot of script sharing with my friend group and I'm not wanting to have to do everything twice in order to be able to share it.
If you were looking for a argument from fallacy case, your best fallacy would be likely I believe appeal to probability, would be the close but not match as I am assuming most of your friends are not using fish, but I'm also basing it off the knowledge that it has a significantly smaller user base which makes it more likely. or possibly an illicit minor which would be the path of "My friends don't use fish, so it's unlikely that your friends use fish" which would potentially be valid, but again I am questioning the case not stating it as an exact, but since my initial question was based based off statistics and experience, I would go with the first one.
but back to the topic:
Documentation wise? I have read it. The examples are nice don't get me wrong, but its layout needs work, the examples need better real world use cases and struggling to search for how to do something because the makers of the shell in their infinite wisdom decided to make a new keyword for something that was already stupid easy to use is just a hard pass for me (like I said I had already learned bash prior to this, whereas you had not learned bash). Not to mention with bash or zsh, I run into an issue I can just search the issue. What would take me 2 minutes to search for a problem with a script using zsh took me 10-15 minutes of research with fish and sometimes it wouldn't even solve the issue at hand and required just rewriting it completely. Usually my path of research would require me to look up the issue using fish, find no solution so look up the issue using bash, then have to convert it to fish. Sometimes the issue would work fully in bash, and just not in fish. I came to the conclusion that if I was having to convert parts of it to bash anyway in order to research issues with it, I might as well do it in bash to begin with.
I agree with you, the more people using it the more examples and documentation will be available as a result, but I'm not going to be a spearhead for it, I don't want to have to exert more energy than necessary, and I found the gains I got using fish didn't outweigh the losses. Like I said I might revisit the shell some day, maybe if it ever becomes super popular, but for now I have removed it and ported my scripts back to bash again.
ammendum: btw LLM's do not like fish shell for bash to fish conversion. I had tried it a handful of times resolving an issue (deepseek had the most success of them) but it was almost always a try 3 or 4 times, get something that has nothing to do with it, or uses something that fish shell doesn't support, and then have to clear the context or find another path for resolution.
ammendum2: also fair warning, the last 2 topics on fish shell that appeared in this community got nuked after 2 days, so it's possible this entire thread will disappear as well(hopefully not but it seems to be a reoccurring issue).