To open the post on your instance you can go to [email protected] and see the recent top posts, or use an app/frontend/ browser extension (ex. [email protected])
I also wanted to share this tip for how you can filter for Lemmy posts when searching:
Search using site:home_instance. So if I wanted to find recommended phones, I could go site:lemmy.ca recommended phones. Since every instance has its own collection of posts, you will be getting results from all over Lemmy. The limitation is that you won't see content from instances that aren't federated with yours, but you probably didn't want to see that stuff anyway since you picked your instance for a reason. You can also put any instance into the search if you wanted different results.
Question to everyone, what does Lemmy need to make it easier for people to find content? What are the implications of the Fediverse on how people might find content in the future?
One thing is that people are more likely to get posts from the larger instances, likely because more people are linking to them and opening those links? Another thought was the common complaint about how our post links aren't community specific. While I can search for posts using the method above, I can't search within a specific community like I can with Reddit (ex. I can't search site:lemmy.ca/c/Vancouver recommended restaurants
EDIT: The issues for it are here, looks like the devs are good with it now and someone just needs to implement it:
spez wants to remove Reddit from Google actively, so at least for this specifically I sincerely doubt he minds. Of course, the main reason he wants thatis only that he wants to inflate Reddit's perceived investor value, so not sure what that even helps. But eh.
Reddit is talking about hiding Reddit from Google. I hope they do that because it will let Lemmy start to replace Reddit as the go to source for non-SEO, real-human answers.
Except each instance has its own URL meaning ranking for ANYTHING is extremely hard since each domain’s rank will always be weak in the sea of others. Each is even being penalised by the algorithm if there are duplicate content mirrored between different URLs. It’s the weakness of the fediverse if we are to follow how search engines have worked in the last decades. Maybe it will lead to new search engines (I hope so) but right now it is not going to work well to replace for example Reddit … or rank well in general at all.
I wonder if there’s a way around this that we can create, instead of doing nothing or hoping google adapts.
Like a dummy instance that catalogues everything on all instances (but also links to the original posts) for the purpose of showing up on google search.
Since this instance isn’t for posting but for search engine indexing, there may be some otherwise undesirable micro-optimizations that can help improve its chances of showing up.
Each post refers to the poster's home domain as the canonical URL, regardless of which instance you're viewing it on specifically to avoid duplication SEO concerns
One of the main reasons reddit mega turned to shit was due to far too many people joining and using it, granted this is due to mobile phones but is it really worth it to attract more and more people? These instances are run by average people not corps with money they can easily collapse under tuw burden of to many
The cool thing about the fediverse is that if grows to much where and the moderation turns to crap on some instances, you can always defederate from those problem instances and avoid the trouble makers entirely. Instance admins can always turn off sign ups when they feel like they have reached their limits for moderation as well. And users can always self host their own instances and only let the people they trust sign up. The fediverse is more resilient than traditional corporate run social media
I use duckduckgo and lately have been adding Lemmy at the beginning (like I did Reddit) and it really works. Search results come back if there is a discussion going on about the topic.
I've not tinkered around with it too much yet, but how has your experience been with actually viewing the results? I would imagine that most results are statically likely to be for an instance other than the one you have an account on, requiring a few extra steps to load the post in your home instance if you wanted to vote/comment on it.
This is one of those UX things that I think is still holding Lemmy back from more mainstream adoption, imo.
You can make that a bit less painful with browser extensions. 'Instance Assistant for Lemmy & kbin', for example, will always show a button to open a page in your home server. Not sure if there's anything exactly like it here yet but FediAct for Mastodon lets you like/boost/follow/comment on other servers directly, more or less eliminating the other-servers problem.
It's hit and miss most of the time. I just tried Lemmy memes and it redirects to lemmy.ml and the memes community.
Then I tried Lemmy how do I wash my kitten and there was an answer. Again for Lemmy ml. I think it depends on whether there is a post answering the question and if people commented on it.
I have been regularly sharing shit with friends that I see in Lemmy, and they always said to me why my links always have weird names and domains and shit... so I proceed to explain and we get to nowhere.
Anyway this is people that weren't even into Reddit, so that people are the harder to get, IMHO.
I’ve had success describing it as “imagine you owned the server your Facebook info was stored on but could still interact with all other Facebook servers”. It’s a little simplified, but it usually gets the point across.
Sounds like a "them" problem, you keep doing what you're doing. Maybe they'll eventually get it, maybe not. Unless you give up sharing content with them entirely, of course, but that's your choice.
I have a question, I have posted with my lemmy.zip acccount to a community hosted on feddit.ro, and when I search the title only lemmy.world appears, even tho i didnt post it there, any idea why?
Every instance stores what are essentially copies of everything it's users are subscribed to. So when you post "to" feddit.ro, you're posting to the copy on lemmy.zip. Similarly, Lemmy.world has their own copy (that for whatever reason is ranking higher in search) because someone there is presumeably subscribed to the community.
The post should exist on all 3 instances (as well as any other instance federated with feddit.ro). However lemmy.world is bigger and so more people are likely linking to /from that instance, which mean Google is indexing it with higher priority.
The other instances should probably show up over time?
Hello from kbin... (federated here too)
But this poses an interesting dilemma for Google, potentially to top 100 results could end up just being the same post observed on many Lemmy and kbin instances.
That's cool. I seem to have issues finding this post, or the top post of the day (https://sh.itjust.works/post/8365139) by searching:
site:sh.itjust.works After watching the 2nd episode of 11th season of Futurama
And I'm not getting the correct result. Am I doing it right, but there's something else affecting the results (top of day for lemmy isn't as popular as needed to show up on google, bing, nor DDG)? or am I making a mistake somewhere?
I don't think the instance is the issue like the other comment is suggesting (although I don't quite understand the specifics of how it works). I'm playing around with it myself right now
If I search site:sh.itjust.works Hello and set it to 24 hours, I do see posts. So timing should be ok too
Update: So I think what's happening is that the post needs to go through a few stages when it's on a different instance
One big issue that Lemmy has because it's a distributed service is the dilution of results.
For example, there is only one Reddit domain (that people use to access the service) but there are hundreds/thousands of Lemmy domains and the dilution will continue to increase as Lemmy's popularity increases. It's either that or there will only be a couple of Lemmy instances that will dominate all of Lemmy.
technically most crawlers abide by robots.txt so if you really wanted you could forbid the google crawler but it would be silly to since it's one of the most popular engines