←back to thread

1222 points phantomathkg | 8 comments | | HN request time: 1.154s | source | bottom
1. poisonborz ◴[] No.44064404[source]
Saving links and offline reading is dime a dozen nowadays, but the one BIG feature was recommended articles, which basically opened the reading list of the world and you could read extremely interesting longform articles based on your list. It was the best magazine you could imagine. But I guess it required maintenance and moderation so they axed it - in the typical fashion of an acquired product I might say.

I still wish for something like this with privacy in mind and community maintained.

replies(3): >>44064490 #>>44064784 #>>44064911 #
2. aspenmayer ◴[] No.44064490[source]
You might be able to scrape archive sites or create search queries that you use to determine fresh or trending archives or domains?
replies(1): >>44064597 #
3. poisonborz ◴[] No.44064597[source]
I think it's really hard to replicate the "people save this article for later" pedigree without actually having this information.
replies(1): >>44064753 #
4. aspenmayer ◴[] No.44064753{3}[source]
True, there are a lot of bots that use Internet Archive, which is probably the easiest to scrape. Maybe ask Jason Scott of Archive Team if he has any ideas for how to use IA and other archives for this purpose, and for ideas about how else to get this data?

I think Instapaper was another solution in this space that may have the info you want.

Maybe ask on some data hoarder subreddits about how to find new content that’s relevant to your interests with existing social proof?

I can see how the data from Pocket would have made that a lot easier for you, but finding a quick solution may be difficult. I think Apple News has a bit of social components around surfacing popular content, but that is not the same as user generated content indicating interest in a specific site, which is your goal.

Are you familiar with MetaFilter? There a community that might have some insight into your question, as they’re like HN but somewhat more crunchy and broadly less technical, but very human. Asking around other communities, you might find some suggestions.

Please let me know if you find a solution because this is an interesting problem, and I would probably be just as interested in the solution.

5. zimpenfish ◴[] No.44064784[source]
> the one BIG feature was recommended articles

The TWO BIG features were recommended articles and integration with IFTTT. I think, of the suggested alternatives, only Instapaper has IFTTT integration (modulo setting up a local webhook).

Given I have multiple things per day (and have for 10+ years) going into Pocket, this is going to be a big pain in the arse to deal with.

replies(1): >>44064923 #
6. smallerize ◴[] No.44064911[source]
I'm not sure what you're missing from all the existing discover features in other products, but NewsBlur has a feature that sounds like that. https://blog.newsblur.com/2025/02/02/discover-related-storie...
7. poisonborz ◴[] No.44064923[source]
You could try Wallabag. It provides 3 separate RSS feeds per saved link status, that could trigger IFTT.
replies(1): >>44066076 #
8. zimpenfish ◴[] No.44066076{3}[source]
Ah, I'm talking about feeding -into- Pocket. I haven't found anything except Instapaper that has direct IFTTT actions for saving URLs or anything which natively supports a webhook kind of thing for saving (but I'm sure there must be something because it seems such an obvious thing to support!)