ArXiv accepts .ps (PostScript), .tex (LaTeX source), and .pdf (PDF) ScholarlyArticle uploads.
ArXiv docs > Formats for text of submission: https://info.arxiv.org/help/submit/index.html#formats-for-te...
The internet and the web are the most transformative platforms in all of science, though.
When your service is small or not easily visible - while still doing significant good - it's hard to find enough people willing to spend their time and resources helping you sustain it.
When your service becomes big enough to be noticeable - which is the arXiv is in by now - it also becomes attractive to the people looking to subvert it to be something else, to enshittify it, and so the limiting factor in getting help becomes the risk to its integrity.
I have a reason to believe that this is not the first time I heard this argument coming from a boomer physicist.