I know Mastodon does not want I can search the plain-text of other user's toots, but I would like to search for my own toots, respectively find them.
Use case: I knew I tooted something interesting/a link/whatever… some time ago, but I cannot find it again. That is bad…
And scrolling through the whole history is awkward.
I think any user of Mastodon, who had this "problem" once, will benefit.
This is already implemented though
hu? Where/How? If I just enter the search term into the search box, i only get hashtags etc. Not my own toots… (and yes, I've tried it with text I've tooted.)
Note I obviously want the whole toot of mine to be searched, not just hashtags or so…
@rugk i am not sure about that (i am new to mastodon) but i read something about Elasticsearch is needed to enable that feature for your instance, maybe your admin didn't setup it up? (but i might be wrong)
https://github.com/tootsuite/documentation/blob/master/Running-Mastodon/Elasticsearch-guide.md#why-you-might-need-elasticsearch
Oh, yes, maybe that's it.
But wow, just for that "tiny" feature of searching your own toots you need a whole new program. That'S funny… :smile:
Full text search is not exactly simple. With Postgres and other SQL DB servers, it's painfully slow after only a few thousand records. Elasticsearch is designed for full text search, however, and is considerably faster at it. It wouldn't work well for most of the stuff that goes in the database, so switching completely isn't workable at all, but adding it on absolutely is.
I'll add that deploying an instance with Nanobox or Docker (properly configured) will automatically get you all of the software used to provide all the features Mastodon supports, so either is a solid approach.
TL;DR? It's not as tiny behind the scenes as it looks out front, but use Docker (with some configuration tweaks) or Nanobox to deploy and you get it for free.
Most helpful comment
Full text search is not exactly simple. With Postgres and other SQL DB servers, it's painfully slow after only a few thousand records. Elasticsearch is designed for full text search, however, and is considerably faster at it. It wouldn't work well for most of the stuff that goes in the database, so switching completely isn't workable at all, but adding it on absolutely is.
I'll add that deploying an instance with Nanobox or Docker (properly configured) will automatically get you all of the software used to provide all the features Mastodon supports, so either is a solid approach.
TL;DR? It's not as tiny behind the scenes as it looks out front, but use Docker (with some configuration tweaks) or Nanobox to deploy and you get it for free.