Ruby should look at how the PHP ecosystem was modernized. Sure the syntax has always been awful and is even more degraded now, but the ecosystem is globally in a much better place.
TIOBE is for the most part crap, but the tendency is also not completely fabricated. Ruby is at rank #25 with 0.67% right. Again, those numbers aren't that relevant, and they fluctuate WAY too much in suspicious ways - TIOBE has many issues, but ruby was doing better in the past there, so something changed. So, not only needs to be an unbiased analysis, but much more importantly so a contingency plan. I feel that in many ways ruby is also way too japanese centric. This is fine for a language that is only used in Japan, but a language should have no real country-focus per se, it should be usable everywhere without constraint. With a contingency plan I mean specific things to do. You can not solve this with single steps - that approach does not work. We saw this with the quest to make ruby faster. Ok, ruby is faster now, that's great, but then why aren't there more users? If ruby being much faster was the number #1 goal, why aren't older users returning for the most part? Why are new users hardly picking up ruby?
I don't want to make this sound too pessimistic per se, mind you. But ruby is now where perl was about 10 or perhaps even 15 years ago. Perl had the problem of perl5 versus perl6, but also python as stronger competitor. Perl5 failed to go against python. Ironically enough perl5 is more active than perl6 - that was also poor planning the perl folks did. (Version changes can be hugely problematic, Guido does not want python4 largely because python2 to python3 transition was problematic.)
Ruby really needs a plan with several items that work. Even more so as matz will sooner than later go into post-design stage (like Guido did with regard to Python though Guido is still somewhat involved with python, just not necessarily as sole decision maker now).
I wrote this a few days ago mostly out of frustration and honestly did not expect it to go anywhere. It is pretty surreal to wake up and see it on HN with so much discussion.
Thank you for reading and for all the comments, messages, and thoughtful critiques.
I am currently looking for roles that sit at the intersection of ML, product, and research. I like open ended work where you figure out what to build as much as how to build it. I am a builder, and I also enjoy PM type work and being close to users and the product. If you are working on something in that space and think I might be a fit, I would love to chat.
Also, thank you to Daniel Han for sending me the link and bringing this to my attention.
In any case, thanks again for reading and for the conversation.