Google searched worked really well for reddit if trying to find an answer to a technical question or the like. How well will this translate for Lemmy? I assume it would be a lot more difficult as the post could reside on any number of smaller instances.
Since all posts/comments are public, all you need is some kind of indexer to organize them and a list of all the instances. There are people working on it right now. But in the meantime you can just add
(intext:"modlog" & "instances" & "docs" & "code" & "join lemmy")to your search queries to search all Lemmy instances.I can imagine there will be improvements for SEO in the future, but for right now, being in beta, it kinda makes sense that it would not want to be so very visible.
In the meantime, there is also Search-Lemmy.
Thanks for the callout. For anyone using that site, I have a list of bugs I’m hoping to get fixed in the next couple days that should fix a lot of the major issues (broken links, etc…).
lemmy-ui dynamically populates content from a JSON fetch in Javascript, so I expect it will be poor for search engines. In both Chrome and FIrefox on desktop, I can’t even save a post or comments with the normal browser ‘save’ feature.
Doesn’t new Reddit and many other sites do this as well? I suspect this isn’t as big of a problem as it seems.
There’s certainly workarounds that are possible. I saw something the other day on a Google help site that provides SEO tips for dynamically generated JavaScript websites. But I’d imagine it hasn’t been a priority for the Lemmy devs yet.
They are very low in the search rank but I have noticed that when I search lemmy specific topics I do get “posts” from various instances in the results.
So it seems they can be indexed at least by google but I think the project could benefit from some SEO work to help community discovery from search.
