WIP: Support source:markdown in RSS feed

pzingg commented

2022-12-05 18:05:35 +00:00

First-time contributor

Dave Winer, author of RSS 2.0, has been reviving his work on federating RSS feeds. One of the things he wants to accomplish is supporting Markdown as the syntax of choice.

You can read his justification here:
http://scripting.com/2022/07/19/152235.html?title=devNotesForMarkdownInRss

He has also submitted an issue on the Glitch-Soc Mastodon repository:
https://github.com/glitch-soc/mastodon/issues/1973

To implement this, RSS 2.0 feed builders, such as Akkoma, only need to keep the original Markdown source in their posts (which Akkoma does), and expose it in a <source:markdown> element in the feed item.

This PR adds the <source:markdown> element to any feed item that has ["source"]["mediaType"] set to "text/markdown".

Dave Winer, author of RSS 2.0, has been reviving his work on federating RSS feeds. One of the things he wants to accomplish is supporting Markdown as the syntax of choice. You can read his justification here: http://scripting.com/2022/07/19/152235.html?title=devNotesForMarkdownInRss He has also submitted an issue on the Glitch-Soc Mastodon repository: https://github.com/glitch-soc/mastodon/issues/1973 To implement this, RSS 2.0 feed builders, such as Akkoma, only need to keep the original Markdown source in their posts (which Akkoma does), and expose it in a `<source:markdown>` element in the feed item. This PR adds the `<source:markdown>` element to any feed item that has `["source"]["mediaType"]` set to "text/markdown".

pzingg added 1 commit 2022-12-05 18:05:35 +00:00

ci/woodpecker/pr/woodpecker Pipeline is pending Details

eed5d7dc76 support source:markdown in RSS feed

nbsp commented

2022-12-07 11:47:46 +00:00

Contributor

this proposal looks poorly defined
what is the "Markdown" that's being proposed for use here? i haven't seen him link to a spec anywhere (saying "staying within a reasonable subset of the spec" doesn't count)
he probably refers to CommonMark (which version?), but that's technically not the original Markdown by Gruber. petty nitpick in informal writing but this is a spec we're talking about

this proposal looks poorly defined what *is* the "Markdown" that's being proposed for use here? i haven't seen him link to a spec anywhere ([saying](http://scripting.com/2022/07/20/135214.html?title=whichMarkdown) "staying within a reasonable subset of the spec" doesn't count) he probably refers to [CommonMark](https://spec.commonmark.org/0.30/) (which version?), but that's *technically* not the original [Markdown](https://daringfireball.net/projects/markdown/syntax) by Gruber. petty nitpick in informal writing but this is a spec we're talking about

pzingg commented

2022-12-07 15:59:44 +00:00

Author

First-time contributor

I agree that there's a general problem with the RSS 2.0 spec, which definitely allows for namespaced extensions such as <source:markdown>, but these are usually vendor-driven and not created through an RFC or published standards, an exception being the media namespace.

In the argument for <source:markdown>, it's a chicken and egg problem. Dave Winer is honestly trying to promulgate the benefits of Markdown-based world for RSS, but I think only he has the generators, readers, aggregators, etc. to support it at the moment, so if this proposal is too early for the main Akkoma distribution to support, I understand.

On a side note, there doesn't event seem to be a consensus on the RSS/XML encoding necessary for RSS 2.0 content (whether or not that content is Markdown):

https://en.wikipedia.org/wiki/List_of_XML_and_HTML_character_entity_references#List_of_predefined_entities_in_XML
https://www.rssboard.org/rss-encoding-examples
http://www.xn--8ws00zhy3a.com/blog/2006/06/encoding-rss-titles
http://dev.fyicenter.com/1000762_Protect_XML_Special_Characters_in_RSS_Feeds.html

I wonder whether it's worth trying to do a retest of James Holderness's 2006 article with today's feed readers.

I agree that there's a general problem with the RSS 2.0 spec, which definitely allows for namespaced extensions such as `<source:markdown>`, but these are usually vendor-driven and not created through an RFC or published standards, an exception being the [media namespace](https://www.rssboard.org/media-rss). In the argument for `<source:markdown>`, it's a chicken and egg problem. Dave Winer is honestly trying to promulgate the benefits of Markdown-based world for RSS, but I think only he has the generators, readers, aggregators, etc. to support it at the moment, so if this proposal is too early for the main Akkoma distribution to support, I understand. On a side note, there doesn't event seem to be a consensus on the RSS/XML encoding necessary for RSS 2.0 content (whether or not that content is Markdown): https://en.wikipedia.org/wiki/List_of_XML_and_HTML_character_entity_references#List_of_predefined_entities_in_XML https://www.rssboard.org/rss-encoding-examples http://www.xn--8ws00zhy3a.com/blog/2006/06/encoding-rss-titles http://dev.fyicenter.com/1000762_Protect_XML_Special_Characters_in_RSS_Feeds.html I wonder whether it's worth trying to do a retest of James Holderness's 2006 article with today's feed readers.

👍 1

pzingg commented

2022-12-17 16:11:42 +00:00

Author

First-time contributor

I'm going to voluntarily close this PR. I'm currently doing a bunch of other work with Akkoma's RSS feeds, and I'll set up issues on them after I have fully tested with a variety of feed readers. If the issues are deemed of interest, I will re-submit this PR and some new ones later.

pzingg closed this pull request

2022-12-17 16:11:42 +00:00

ci/woodpecker/pr/woodpecker Pipeline is pending

Details

Pull request closed

Please reopen this pull request to perform a merge.

WIP: Support source:markdown in RSS feed #343

Pull request closed