Sjmarf

Mlem dev here! Lemmy has a custom “flavor” of Markdown that is distinct from other social media platforms. Open-source markdown parsers and renderers exist for popular flavors of Markdown (e.g GitHub-flavor), but not for Lemmy-flavor. Most Lemmy clients choose to use an existing GitHub-flavor parser that is close enough to Lemmy’s to be indistinguishable in most cases. Mlem uses swift-markdown-ui to render markdown, which uses cmark-gfm as its parser.

Lemmy’s spoiler format is unique to Lemmy-flavor markdown, so that’s one of the places where use of a third-party markdown parser is noticed by users. Other common parsing errors are _subscript and footnotes.

Adding spoiler support is not particularly easy, unfortunately. You can’t really apply spoiler-parsing logic on-top of another markdown parser - it has to be integrated into the parser itself. This is because the app needs to ignore spoiler markdown in certain situations, such as inside of a code block. The only good option is to write a custom markdown parser from scratch, or modify an existing markdown parser to support Lemmy’s markdown dialect. Both options can be difficult for developers for several reasons:

• cmark-gfm is written in C, which the developer of the Lemmy client may not be familiar with.
• If the app is using a third-party renderer, and not just a parser, that renderer also needs to be rewritten to support the new parsing logic.

This takes a significant amount of time for comparatively little value for users, so most client developers didn’t prioritise it.

In an upcoming Mlem version, we’re replacing our markdown parser renderer with a custom one that can render spoilers and subscripts, but we’ve got a way to go before we achieve full parity with Lemmy. If any developers of other apps are using cmark-gfm, you’re welcome to use our code from that repo under the terms of the licence.

Sorry this is kinda long, I hope this answers your question

Mlem dev here. It shows strikethrough, but not superscript. Lemmy has got a unique dialect of Markdown - in the current version we’re using an open-source Github-dialect parser, which is close to Lemmy’s dialect but not perfect. In the next update we’ve written a custom parser that handles all of Lemmy’s Markdown elements, including superscript and spoilers.

Octopus-Accelerated Machine Learning