It's strange to me that people are so focused on quote toots (which would be nice but I don't have particularly strong feelings about) when, in my opinion, what Mastodon desperately needs is a functional search bar. There is no better way to find out what is going on with a given topic.
Yeah, I know about hashtags, and I really don't think they are a suitable replacement.
@AbandonedAmerica you may already know this, but searching is a much harder problem to solve. You can install a special search apparatus on an instance - it's a whole other application (ElasticSearch) that's pretty resource intensive. the functionality is still pretty limited, not sure if that's because of the resource requirements
I have a feeling this is a problem that will be solved by third party vendors, not instances
@AbandonedAmerica so, like there might be an ad supported site that you can go to search across many instances. or there might be a service you can pay for (monthly or per search?) that would provide full text indexing and searching.
I think many people would get pretty mad, but I don't think they can really DO anything about such services?
@blt @AbandonedAmerica yeah I could see that. a third party searching site wouldn't actually need to adhere to activitypub protocols at all though, it would just index (like google) and point to posts. it would probably need to adhere to the rules of regular web crawlers, though, not sure if most instances disallow indexing altogether
It's an interesting and quite hard problem!
@mrcompletely
Might be worth noting that based according to the docs, the ES integration specifically doesn't do what I think @AbandonedAmerica is looking for:
"Mastodon supports full-text search when ElasticSearch is available. Mastodon’s full-text search allows logged in users to find results from their own statuses, their mentions, their favourites, and their bookmarks. It deliberately does not allow searching for arbitrary strings in the entire database."
@mrcompletely @AbandonedAmerica
Yeah, I'm sure this is a design choice driven by safety concerns from users who've experienced discoverability as a negative, not on cost/complexity grounds.
I don't experience the harrasment those communities do, and I've only been here just over a month, so I don't feel like I'm well placed to voice an opinion, although personally I don't really notice the lack of full text search.
@AbandonedAmerica We kind of already have an opt in search, the problem is one of expectation brought over from twit (which keys off all words because it needs that aggregate data to drive trending and algorithmic content amplification)
We have hashtags, the use of one marks content findable and the user has to deliberately put them in. It's not the same as a monolithic platform search, but we couldn't have that even if we wanted it.
@AbandonedAmerica This thread by @johnmu discusses the broader topic of web search, but it's a good starting point and a rabbit hole as well:
@AbandonedAmerica Quite a few people seem to be very opposed to search because bad actors might use it to harass or dox people.
Those fears may be legitimate, but I suspect they underestimate how many people are already sitting on full-text indexes of fediverse posts. It looks to be a one-line change for a server admin to remove the current restrictions on ElasticSearch, and other ActivityPub software like Friendica already has full-text search.
@AbandonedAmerica I don't think there's a good way to address those concerns while maintaining a federated network that's open to the public. Bad actors will build search indexes regardless of anyone else's preferences and it's probably impossible to stop them by technical means.
It wasn't a problem when the Fediverse wasn't on their radar, but recent DoS attacks suggest that it is now. I don't have a good answer for people concerned they'll be targeted using search.
Meanwhile the rest of us are deprived of a useful search feature, and anyone who attempts to build one transparently gets bullied.
Been on 2 instances so far and both had full text search. FYI
@zak @AbandonedAmerica
As you say, this is an issue with the content being open. What seems to me to be needed is a way to flag content IN A PAGE as non-indexable...& then get Google/et al to honor that. (Which may be infeasible, I don't know how page crawling works at that scope.)
E.g., if you marked your post 'noindex', then your post isn't indexed.
It'd need to extend to quote-boost implementations, too.
@FeralRobots
There's also a noarchive value.
Let me know if I shall hook you up to SEO wizards here.
@RyunoKi @zak
Just to be clear in order to work in a social media context, the search spider would have to pass over (or the search engine not retain/index) anything in a container that's flagged noarchive or noindex, respectively.
Do you know if major search spiders work that way now? It's been too long since I had to know this stuff for work. This is only slightly more than curiosity for me, though that same may not be true for @AbandonedAmerica.
@FeralRobots @zak @AbandonedAmerica Deferring the question to @matterne
I haven't had to touch SEO in a few years so I'm not ahead of the curve right now.
@FeralRobots
Well, on the web client the time appears to be loaded with AJAX.
So unless there is a special treatment for crawlers they wouldn't see the toots.
Assuming noone queries the API and renders them server-side somewhere.
BornintheBronx51
Abandoned America
Karin!
Sean
Brad T
Mr. Completely
mhamiltonj
☕️🥞 Jelly Doll 🏳️🌈🏳️⚧️
Colby Moorberg
Katzentratschen
🦇 Jennie Rigg 🏳️🌈 (she/her)
Zak
🇨🇭Canniuanos
FeralRobots
Ryuno-Ki
Lacci 🐇
AMS