This was something I requested ages ago, however, unless it has been added since, there is nothing in the api for this. I have been meaning to add a feature request to piefed for the feature with api support. @wsj018@piefed.social
@blaze@piefed.zip has been asking for this feature for ages and @Skavau@piefed.social made an issue for it on the codeberg already. The reason this isn't in the api yet is because it isn't anywhere in piefed yet. AFAIK, the read/unread status of comments isn't currently tracked anywhere within piefed. Additionally, for this feature, I don't believe we track the timestamp at which you do interact with a post...so doing something like calculating the number of new comments that have occurred since you last opened that post isn't possible...yet.
I don't have a great idea of how to do this in a db-efficient way. Keeping track of when every user opens each post sounds like db-query hell. However, there might be clever caching tricks to make it a bit less computationally intense. @rimu@piefed.social would likely have a better idea of how to do this.
Thanks for the info :) Lemmy 1.0 tracks a readAt timestamp; maybe it's worth looking at how they're doing it.
Can't you track which posts (and when) people view locally, in your app? Then when displaying comments compare the comment time with the last-viewed-post time to determine if it's new.
@wjs018@piefed.social
We could do, but ideally it would be implemented on the backend - some people use PieFed on both mobile and desktop, and storing the last read count locally wouldn't carry over properly between clients. Lemmy 1.0 has backend support for this feature.
I'd like to survey our users on this to see how they feel about what they read being tracked in this way.
Some time ago I floated the idea of somehow highlighting/prioritizing comments by people who had clicked the link to the article. People who mouth off without reading the article are a pain in the ass so I wanted to somehow de-emphasize those. The feedback I got at the time was pretty strongly "no I don't want you knowing what I read". It's too easy to do big data stuff on it to build profiles on people, I guess. It wasn't a proper poll, just people commenting so it's hard to know what the silent majority thinks.
So I'm a bit wary about it now.
Any suggestions for how to word the survey question? How it's framed will make a big difference to the result.
If there was a readAt date for the post, we'd be able to highlight comments created since that date. I don't think tracking a readAt date is much less private than tracking a read boolean, which you're already doing. It could always be made opt-out, if people are concerned about it. Having the date that someone read a post isn't as exploitable as the link click tracking you mention, I don't think.
Currently, we are missing API support for this. Lemmy 1.0 (the upcoming version of Lemmy) includes this feature, so we'll add support for it when Lemmy 1.0 is released :)
Mlem for Lemmy
Official community for Mlem, a free and open-source iOS Lemmy client.
Rules
- Keep it civil.
- This is a forum for discussion about Mlem. We welcome a degree of general chatter, but anything not related to Mlem may be removed at moderator discretion. This is not a forum for iPhone/Android debate. Posts and comments saying nothing but "iOS bad/I use Android" will be removed as off-topic.
- We welcome constructive criticism, but ask that it be both precise and polite.
FAQ
- When will insert feature here be implemented?
- Check our issue board--if there isn't an issue open for the feature you want, feel free to open an issue or make post! Just remember that devs are people too--we're doing this for free in our spare time, and building a quality app takes a lot of patient work.
- Is Mlem available for Android?
- No. Mlem is written using SwiftUI, which is not currently supported on Android. If such support becomes available, we will look into bringing Mlem to our Android friends.
- How do I join the beta?
- How do I join the dev team?
- Head over to our recruitment channel, or go straight to our GitHub and read CONTRIBUTING.md to get started.