What's the point of the markdown filter?

[jvoisin]

Miniflux supports converting feeds from markdown to html via github.com/yuin/goldmark, but I fail to see a usecase, as I don't know a single rss/atom feed that spits markdown and not html.

If it has no use, it' be nice to remove it, as reducing the amount of external dependencies is almost always a win, packaging-wise, security-wise, reliability-wise… but also size-wise, as it adds adds a lot of code to the miniflux binary:

$ goweight | grep goldmark
  2.5 MB github.com/yuin/goldmark/ast
  2.4 MB github.com/yuin/goldmark/util
  1.7 MB github.com/yuin/goldmark/parser
  448 kB github.com/yuin/goldmark/renderer/html
  364 kB github.com/yuin/goldmark/text
  139 kB github.com/yuin/goldmark
  114 kB github.com/yuin/goldmark/renderer
$

• [x] I have read this document: https://miniflux.app/opinionated.html#feature-request

[fguillot]

It was introduced in this PR https://github.com/miniflux/v2/pull/1513

This PR adds a new rewrite rule that parses markdown. Parsing is done with a new dependency yuin/goldmark. This PR also configures the parse markdown rule by default for blog.laravel.com.

Removing this feature would be a breaking change for the few people that use it. I'm not aware of any lightweight alternative Markdown parser library in Go.

[jvoisin]

I don't think it'll break anything, as nobody should be using it. The reason it was added, blog.laravel.com, doesn't publish markdown.