Mastodon does not offer federated full text search on what people posted, right? Compared to Twitter that makes it much harder to find people on niche topics.
@FroehlichMarcel freemo is someone who could help connect you with other scientists.
@AmpBenzScientist Thanks @AmpBenzScientist my question is not so much about me. I’d like to convince more people to switch from Twitter to M. But without additional search capabilities is will be hard to reproduce the experience they find useful.
@AmpBenzScientist Hashtags seem to be key both on profile and in posts for discoverability.
@freemo when I / others use lots of hashtags, what are features that make use of it beyond local search?
@freemo @AmpBenzScientist For example yesterday I published a thread on biology and cryptography. Not exactly a popular topic. In the vast fediverse there should a few people interested in this. How do I properly make my content/profile discoverable. And how can I actively search those? #Mastodon
@AmpBenzScientist @freemo I plan to publish many more such science threads that I already accumulated on Twitter, but first I want to get it right to avoid posting into the void.
@FroehlichMarcel @freemo Post and people will see it. You will attract a following. Searching is frowned upon due to privacy and spiders constantly pulling bandwidth.
🇳🇱
@FroehlichMarcel @AmpBenzScientist Sorry been away for a while now, just got back (didnt have internet connection at sea).
Can I still help with something?
@AmpBenzScientist @freemo Thank you @freemo, probably not. Just had to learn about the boundaries Mastodon sets on indexing across servers.
Each individual server decides if and how to index. If its vanilla mastodon not much indexing happens, if it is pleorma or another flavor of server in the fediverse your full text will be indexed in most cases regardless of what flavor of server your on.
Our local instance enables full text indexing and can search with some restrictions which we may wind up opening up a bit more (has little to do with indexing and more to do with allowing some advanced searches in the UI).
Also, and perhaps more importantly, each individual account on qoto can device if they want to turn webcrawlers on or off in your account settings. This will essentially tell well-behaved search indexers (like google or even fediverse specific searchings such as those in clients) whether you allow indexing of your content or not.
@AmpBenzScientist @freemo the ideal would be to allow fedi crawlers and reject the rest. But that granularity is probably not on the menu so far.
While that is doable code wise the hard part is convincing other serves on the fedi to turn on text indexing.
@freemo @FroehlichMarcel I'm fairly certain that it would have to be scrubbed of any personally identifiable information and heavily encrypted. It would seem to be against the nature of the service. I do understand how much of an inconvenience that this is and how valid of a reason you have. It would still require reaching out to the other instances and asking if they would want indexing for the Sciences. Some will not cooperate but you would likely get other scientists interested in this. You might very well contribute to the verse if it is accepted and approved.
@FroehlichMarcel @freemo I would recommend using social networking. I will find people who will be interested in the posts. They will tell others and that's how your following will grow. I like the approach better than the alternatives.
@FroehlichMarcel@qoto.org 1) Afaik : yes, by design.
But discoverability in itself isn’t sufficient, you would also need people looking to discover such content. 😉 Which can be a common problem on the #Fediverse unfortunately...
#NetworkEffect #CriticalMass #PositiveFeedbackLoop #FediverseProblem
https://en.m.wikipedia.org/wiki/Network_effect
2) Hashtags
a) in posts :
- people can follow specific #hashtags . (On #Misskey and #Friendica , not sure about other networks.)
- people can search for posts, e.g.: https://mastodon.social/tags/computation
b) in profiles : useful to filter the users directory.
E.g.: https://misskey.de/explore/tags/computation
@FroehlichMarcel@qoto.org (old)
@FroehlichMarcel@mastodon.social (new)
@yaksha Right, same here.
@FroehlichMarcel@qoto.org
(Caveat : I haven’t read the whole thread. 😇)
Right. 😢 Indeed. 😢
For details see :
Full-text search - Mastodon documentation
https://docs.joinmastodon.org/admin/optional/elasticsearch/
—
#FullTextSearch #Mastodon #Fediverse #Search #NicheTopics #LongTail #MastodonDocumentation #ElasticSearch
So this is the chicken egg problem for Mastodon for long tail topics. Obvious stuff like open source works well, but specialized science topics are hard. But I understand that it is an intentional design decision that saves people from the mob and from trolls.