πŸ„IdeaImport

Look at the name of this engine. Mycorrhiza. In real world, mycorrhiza is a connection between mycelia and roots of plants which is profitable for both sides. I've chosen this name for a reason: I always wanted wikis running MycorrhizaWiki to connect with the rest of the internet.

The giving away part, the export, is discuss in other hyphae. See feature/feeds, standard/opengraph.

This hypha discusses the ways MycorrhizaWiki may import content in the future.

Also see ../fediverse. This hypha does not discuss federations.

Previews

In many messengers and other social media links have previews. Why not have the same type of previews for rocket links?

Save page in mycomarkup

When creating a hypha, there should be a button Import article from URL. You press it, there's a dialog window where you put the URL. Then there's an attempt to convert it to mycomarkup and save it on the wiki.

Supported protocols could be Gemini, Gopher, HTTP[S]. Converting content into mycomarkup and saving it is easy for the first two. The web protocols are harder. I guess I'll rely on github.com/advancedlogic/GoOse for extracting the main data and then convert HTML into mycomarkup somehow.

What about pictures used in web pages? Should they be fetched and copied for wiki? Or should they not? Maybe an option?

It's easy to guess that web page import will fail often for the same reasons why reader mode in browsers and applications like Pocket fail often too.

Tweet import

Twitter is a mess. Guess what: it's impossible to just download a tweet. You need an API token which is hard to get. Just parsing HTML of the tweet is not easy because Twitter uses a lot of β€œmodern” technologies like React. I think including a browser runtime into a wiki server is a bad idea, thus an alternative way is needed.

nitter.net is a great service which lets you browser Twitter without making your computer fans spin. Mycorrhiza would fetch tweets from there. Since Nitter can also be self-hosted, it makes sense to allow selecting preferred Nitter instance.

Yet again, what to do with pictures?

Telegram import

I often find myself willing to copy something from a Telegram channel. That's why I want to support it. Telegram has a web view for public channels which is good enough for me.

Perhaps importing whole chats using Telegram Desktop's export feature is something to consider too.

Email import

I often find myself copying emails from a mail client to a wiki. Making it more automatic is a good idea.

Other social media import

If needed, it can be considered. VK posts and articles, Facebook posts, Instagram posts.

Importing from social media should be done with the same one button I mentioned earlier. When importing an URL from a social medium that MycorrhizaWiki recognises, it switches to a different algorithm for that site. Of course, the number of explicitly supported sites should be as minimal as possible or this functionality should be generalized.

Import a wikipage from a wiki

It could only be done together with ../interwiki. Since you specify engines sister wikis run on, the engine may have a specific strategy for them ready. For example, fetching an Oddmuse page source is done this way and fetching a MediaWiki page source is done this way. But is this really needed? Why not just fetch with the general method?

Import Markdown files

Drag-n-dropping Markdown file collections would be nice because many people seem to have them.