Avatar
Rif'at Ahdi R
2b67e480b7f99d2835684a8f7276d86edbe8e318ea55cf77ccfd559c5f24f645
Special family-friendly relay with filter settings (Language, Safe For Work, Hate speech, Sentiment, Topic, etc) for Global Feed: https://github.com/atrifat/nostr-filter-relay/blob/main/USAGE.md wss://nfrelay.app Indonesian. Learning and interested in PHP, JS, Go, DevOps, Android, and Machine Learning

I think Japanese users have been quite affected with limited Air-reply and other interactions ๐Ÿ˜…

Are you okay, Takumi-san? Hopefully you are in good way ๐Ÿซ‚

Yes, it sits like a proxy relay in front of its backend relay. Let me know your experieces self-host it ๐Ÿ‘€

Yes, it depends on client whether they offer all the features to freely customized their feed settings. This relay can complement any filtering features that were probably not exist yet (language, sentiment, topic, user type, untagged non sfw content, etc) in the client.

Hopefully, all the features that were offered in nostr-filter-relay fully implemented in the client.

Yes, it can be integrated freely with any backend relay (nostream, nostr-rs-relay, and other relays). We just need to setup relay url config in env variable.

This repo only integrate several modules which can be replaced and customized individually.

nostr-filter-relay v0.3.0 update

"Browse global. Know that mostr exists. Follow lots of people. Nostr is NOT an echo chamber, if you use it right." (Gigi)

Thank you to our users and OpenSats for supporting this project ๐Ÿค—.

This release bring new major features, modules updates, several code improvements, security fix, better documentation, and mark the completion of nostr-filter-relay features as follows:

- (New ๐Ÿš€) Sentiment analysis filtering (Check USAGE.md)

- (New ๐Ÿš€) Topic classification filtering (Check USAGE.md)

- Language filter

- NSFW/SFW image content filter

- User type filtering (Nostr native user/non bridged user, activitypub/mostr bridged user)

- Toxic comment filter

Feel free to check USAGE.md documentation to see more Examples on how to use nostr-filter-relay.

All filtering features can be easily disabled (opt-out) with parameters content=all :

wss://nfrelay.app/?content=all

One last major issue that will become next development target:

- NIP-32 Label Event Compatibility

What's Changed?

- Topic classification filtering support

- Sentiment filtering support

- Update nostr-monitoring-tool and nostr-filter dependencies

- Documentation of nostr-filter-relay new features

- All detail update for nostr-filter, nostr-monitoring-tool, and other modules were listed in release note link below

#NostrFilterRelay Release Note v0.3.0:

https://github.com/atrifat/nostr-filter-relay/releases/tag/v0.3.0

It runs on NB, nostr:npub138s5hey76qrnm2pmv7p8nnffhfddsm8sqzm285dyc0wy4f8a6qkqtzx624 will need to update api call and itโ€™ll work ๐Ÿถ๐Ÿพ๐Ÿซก

Glad to hear. ๐Ÿซ‚

Oh nice, is it already running for nostr.build and nostrcheck.me? ๐Ÿ‘€ .

Hopefully it can help both of your services.

We could work on the multimodal thingy together then. nostr:npub138s5hey76qrnm2pmv7p8nnffhfddsm8sqzm285dyc0wy4f8a6qkqtzx624 and I have been experimenting with general detection of safe/unsafe media. I can share my repo (still closed source for now until I am sure the code is safe and clean) and take it from there. Interested? ๐Ÿถ๐Ÿพ๐Ÿค”

Yes, i'm interested. I think i can probably join helping 1 month later. I need to finish some of pending work in nfrelay.app for the moment.

I have some idea to improve and adjust it to be able detect CSAM. Is it ok to send through to your DM (Regular DM or Gift-Wrapped DM) Fishcake?

Really hope that LLM research will be more matured and big players (Meta, etc.) fully open their dataset source (not just pre-trained weight) of training data (publicly available) thus other researchers and users can really judge "how much unbiased" it was. Research in Computer Vision for example has shown many research that were developed based on fully open source dataset (ImageNet, COCO, Visual Object Tracking, etc). Reproducible for other researchers.

Yes, it is really stable. Damus relay and nos.lol (top relays) use it. Most likely the maintainer was occupied with other things previously.

If you want to match top relays storage i think you can probably need around 500GB. Assuming you want to store all previous events from 2022-2024. You can still probably have half of the capacity left.