@Zergling_man Gopher is a separate protocol? No way. Boy, thanks for telling me, oh random fucking faggot of the interwebs. I would have never known otherwise.
@lks My point is that it's a breaking change: If you want to browse stuff on gopher someone has to put stuff on gopher for you to browse. Using lynx to browse http is not, the stuff is already there, it's just that approximately half of it doesn't work at all. For example, silvie.org, where I was told I could find a comprehensive list of Grand Archive cards. It doesn't work at all but with some combination of lynx, curl and palemoon I managed to discover that it was using github to host the images so I just found the repo and cloned it (also over https), and it even came with metadata. http is actually fine, it already has suitable tools to deal with whatever weird scenario you might encounter, like wanting to memorialise Terry Pratchett: yun-wuxin:[wisknort]:~$ curl -sI rakka.au | grep -i terry X-Clacks-Overhead: GNU Terry Pratchett ... and they get out of the way when you're not using them. Every header (except maybe content-length) is optional, you just have to coordinate server and client into defaulting the same way. If you're already considering coordinating server and client into using gopher, this would be less work. html is actually fine, as long as you don't forget to close your headers, or nest 1000+ divs. css is actually fine as long as your html is not retarded. Maybe gopher still uses html+css, I don't know. I would not argue for a non-breaking change over a breaking change just on the grounds that it's non-breaking.
@lks Gopher is a separate protocol, lynx is a fully-featured world wide web (WWW) client. Maybe lynx supports gopher, I don't know. I don't know anybody using gopher, so I haven't looked into it at all. Probably it is better than gemini.