As development on Arctic seems to have halted, Blorp seems like a promising alternative for its Piefed support, but could use additional quality of life features. Among the most important of these from Arctic is additional granularity in post filtering, ensuring that posts one doesn't want to see can be excluded without impacting the visibility of other posts.
Beyond just filtering keywords, options to apply distinct keyword filtering for post titles, content, URL, OP username, and community name would allow users to preemptively exclude content they aren't interested in. Furthermore, via regex expressions, terms can be combined to increase efficiency and maximize the scope of the filters.
Here's an example of how Arctic handles it, with several of my own regex filters.
While I think that approach would be beneficial to new Lemmy and Piefed users who are likely to be overwhelmed by a deluge of unfiltered content, I think it would also be helpful to have an "expert mode" built into the app to construct such filters. As a comparison, while uBlock Origin's filter lists are extremely helpful for ad blocking, the ability to write additional filters oneself locally adds a degree of personalization to the process. If the process were confined to filter lists, I'd personally write my own given the meticulous detail to which my existing filters go.
Regardless of which approach is taken, however, I definitely think that there is merit in supporting the five aforementioned filter types; identifying filters as applying to either the title or post content enables the blocking of specific reoccurring posts without accidentally catching others in the filter, blocking URLs enables filtering out fake or biased news sources (per one's own standards), or otherwise unwanted content, without blocking posts discussing those sources (i.e. blocking links to YouTube without blocking headlines mentioning YouTube), blocking community names enables pre-filtering communities that a user isn't interested ahead of time in without blocking them individually (i.e. any community with 'meme' in the title), and username filters enable the blocking of posts by certain users without blocking the visibility of their comments on other posts.
Lastly, I also think that even if json filters were used, that regex support should be maintained, so as to minimize the number of individual filter entries needed and apply the filters as broadly or narrowly as intended by the filter's creator.
That’s good! But as a counter point, only having filter lists would force you to publish your filters which would benefit everyone.
If this is something people are interested in, I would be willing to build some debug tools to test the filter lists. Maybe a setting in the app to highlight filtered posts in red instead of hiding them while you’re tuning your list.
I could even build a site to browse the filter lists people publish. Or browse from within the app.
I definitely agree that the ability to share filter lists would be a major improvement to how Arctic handles it, particularly when Lemmy and Piefed need all the help they can get in streamlining the experience for new users.
An official repository would definitely be useful, likely increasing user adoption of filters.
I love Lemmy, and I don’t wanna tune out of politics, but the fire hose of US politics is really bad for my mental health. I need better filters if I’m gonna keep using Lemmy. I just need a few people to validate my idea before I sink time into building it.