# I am the Watcher. I am your guide through this vast new twtiverse.
# 
# Usage:
#     https://watcher.sour.is/api/plain/users              View list of users and latest twt date.
#     https://watcher.sour.is/api/plain/twt                View all twts.
#     https://watcher.sour.is/api/plain/mentions?uri=:uri  View all mentions for uri.
#     https://watcher.sour.is/api/plain/conv/:hash         View all twts for a conversation subject.
# 
# Options:
#     uri     Filter to show a specific users twts.
#     offset  Start index for quey.
#     limit   Count of items to return (going back in time).
# 
# twt range = 1 24
# self = https://watcher.sour.is/conv/dfxhevq
@mckinley I didn't thought too much about it.

Personally I think that if a discussion is alive posts will be there, I don't really mind if an old post/page lose its comments.

A solution might be a pod with an option to store twts forever?
@justamoment This is sort of possible, kind of, by tweaking:


$ yarnd --help 2>&1 | grep 'max-cache'
      --max-cache-fetchers int        set maximum numnber of fetchers to use for feed cache updates (default 12)
  -I, --max-cache-items int           maximum cache items (per feed source) of cached twts in memory (default 150)
  -C, --max-cache-ttl duration        maximum cache ttl (time-to-live) of cached twts in memory (default 240h0m0s)


But you would have to supply insanely large values. Better solution would be to improve the search codebase and then incorporate those improvements (hopefully in a reusable way) into yarnd itself so we can index and search the archive.
@justamoment This is sort of possible, kind of, by tweaking:


$ yarnd --help 2>&1 | grep 'max-cache'
      --max-cache-fetchers int        set maximum numnber of fetchers to use for feed cache updates (default 12)
  -I, --max-cache-items int           maximum cache items (per feed source) of cached twts in memory (default 150)
  -C, --max-cache-ttl duration        maximum cache ttl (time-to-live) of cached twts in memory (default 240h0m0s)


But you would have to supply insanely large values. Better solution would be to improve the search codebase and then incorporate those improvements (hopefully in a reusable way) into yarnd itself so we can index and search the archive.
@prologic oh, I see.

I'm not sure how to optimize this but I guess it depends on the demand for such a feature.
@justamoment I'm building an indexer now to indexnand query an archive. So let's see how that goes 🀞
@justamoment I'm building an indexer now to indexnand query an archive. So let's see how that goes 🀞
@prologic I'll leave it to you, "Go" forth my warrior! πŸ˜ŽπŸ‘Œ
@justamoment Haha πŸ˜… The way I see this working is, we present an interface that allows you to query the archive and load up older Twts (or replies).
@justamoment Haha πŸ˜… The way I see this working is, we present an interface that allows you to query the archive and load up older Twts (or replies).
@prologic Great! And that would be the source of the data to show in a potential external page too.
@justamoment Yup πŸ‘Œ
@justamoment Yup πŸ‘Œ
@justamoment
> Personally I think that if a discussion is alive posts will be there, I don’t really mind if an old post/page lose its comments.

I disagree with that. I always enjoy reading what people have to say about blog posts, and it's not uncommon for comments to be months or even years apart. Discussion doesn't have to be "alive" for a comment to be worth reading.
@mckinley True but we're talking about two different cases here πŸ€”
@mckinley True but we're talking about two different cases here πŸ€”
@mckinley You're absolutely right, that's why I kept mentioning it as a personal preference. 😁

When I browse the web I almost entirely look for information with an "expiration" (tech and code related stuff tends to get old fast) and quite often if the topic is too outdated the only useful comments are the ones like "Hey! It's been two years now, here's a better way." with some links to a newer page of some kind.

Of course I'm open to both cases to allow freedom of choice for everyone on how they want to handle that.

Another risk on a persistent comment system is to remember to have a way to moderate the posts, since you need to control what gets on your page and can potentially stay there forever.
@justamoment That's just it...

> Another risk on a persistent comment system is to remember to have a way to moderate the posts, since you need to control what gets on your page and can potentially stay there forever.

This ☝️ There will be zero ways to "moderate" or "filter" such content if one were to use teh soon and upcoming search of any given Yarn.social pod. The only way for a user to filter content or hide things is to have an account on said pod where things like "hidden twts" and "muted feeds" can be taken into account.

It _might_ be possible to do things like "exclusion matches" and such, but now that's getting into some tricky territory. Oh well, let's see how it goes πŸ˜… -- Google already indexes all pods anyway 🀣
@justamoment That's just it...

> Another risk on a persistent comment system is to remember to have a way to moderate the posts, since you need to control what gets on your page and can potentially stay there forever.

This ☝️ There will be zero ways to "moderate" or "filter" such content if one were to use teh soon and upcoming search of any given Yarn.social pod. The only way for a user to filter content or hide things is to have an account on said pod where things like "hidden twts" and "muted feeds" can be taken into account.

It _might_ be possible to do things like "exclusion matches" and such, but now that's getting into some tricky territory. Oh well, let's see how it goes πŸ˜… -- Google already indexes all pods anyway 🀣
@prologic Better not, it'll become a mess.

I think it can be managed with something like a dedicated system that interact with yarn or twtxt in general and act as a specific system for comments and discussion to embed in the hands of the site owner.
@justamoment The _troulbe_ with a website and trying to support some kind of "comment system" or "guestbook" type thing is that Twtxt was designed to be a text format with a 1:N mapping of author to twt(s). In essence a single feed is owned by one author that can contain zero or more twts. It is not really possible (_although @darch is trying and finding various ways to break the spec πŸ˜…_) to have a "centralised" feed that lets N users add to it.

To really support this kind of thing, you'd really have to "dynamically" create a new feed for every user, uniquely identify them in some way so you can do this (e.g: with OAuth, IndieAuth, etc), then have an API or Form that they can enter content to and it write to "their" feed._
@justamoment The _troulbe_ with a website and trying to support some kind of "comment system" or "guestbook" type thing is that Twtxt was designed to be a text format with a 1:N mapping of author to twt(s). In essence a single feed is owned by one author that can contain zero or more twts. It is not really possible (_although @darch is trying and finding various ways to break the spec πŸ˜…_) to have a "centralised" feed that lets N users add to it.

To really support this kind of thing, you'd really have to "dynamically" create a new feed for every user, uniquely identify them in some way so you can do this (e.g: with OAuth, IndieAuth, etc), then have an API or Form that they can enter content to and it write to "their" feed._
I _still believe_ the right way to do this really is for a site owner (for example) to run up a yarnd instance (a Yarn.social pod), embed a widget on the website that uses IndieAuth to your pod, dynamically creates a new feed for the user(s) and the widgets uses microPub to post to their feed on their behalf via some Javascript.

Of course we're still missing microPub support.
I _still believe_ the right way to do this really is for a site owner (for example) to run up a yarnd instance (a Yarn.social pod), embed a widget on the website that uses IndieAuth to your pod, dynamically creates a new feed for the user(s) and the widgets uses microPub to post to their feed on their behalf via some Javascript.

Of course we're still missing microPub support.
@prologic Yes, the way I can think of, by using the @darch way, is to have a feed with a user annotation within the twt and post everything on a single feed controlled by the owner.

It initially was a simple idea thrown there and so much has been built for it, it's amazing. 😲

Also some feedback from other people on the matter might help us find some other great solutions.