Show HN: A search engine for forgotten blog posts | Hacker News Reader

Show HN: A search engine for forgotten blog posts

lindylearn.io

11 pointsphgn6 comments

So you want to get away from social networks and still use twitter conversation to rank them?

I like your idea, but fear the hand picked blog approach is not going to be sustainable. Your UI is clean and crisp, search is fast. Good job.

I want to get away from social media noise while keeping the good recommendations you get sometimes. I'm experimenting with this blog search just to see if social comments are useful outside of the original platform.

I'm glad you like the idea generally. You're right about it not being sustainable, the search really doesn't work for "non Hacker News" topics.

My project is somewhat similar to yours. I have a search engine too: https://roastidio.us/search I also offer a web based RSS reader: https://airss.roastidio.us/ So my hope is to get people to use my RSS reader to subscribe to blogs, which can load feeds though my backend, and my backend can then index the blog posts on the side.

Ranking is hard; the problem with using social media popularity to rank is that it is a lottery driven by very strong positive feedback. My search engine so far has not done any ranking.

Interesting, do you see your RSS reader or the search as the main product? Have you thought about getting RSS feeds e.g. from the Common Crawl [0]?

Social media attention is highly asymmetric, but I'm not sure what else to use as filter. There's so too much information otherwise, especially if considering old blog posts. False negatives are the cost for getting more signal.

[0] https://commoncrawl.org

The RSS reader, the search and commenting platform are all parts of the idea that to make blogs more accessible and more discoverable while protecting the privacy of readers and writers.

commoncrawl has too much data and it is pretty hard to sift through. I will stick with the current sources for now.

BTW, if you want to implement user commenting, maybe you can leverage my platform.

Hey HN!

Sometimes I come across a gold mine of a blog, and don't know where to start reading. I wonder how many such blog posts I never hear about. Everyone has too much to read - but it's about things you don't want to miss, not which are useful. Conversely, finding thoughtful essays on a specific topic is hard. Social platforms will serve you the latest, not the most relevant information [1]. Your best bet is to ask good friends.

What this blog search does differently:

* Indexing only high-quality blog posts [2], to avoid SEO hacking and noisy news. It favours articles that are years old.

* Showing Twitter conversations for each article [3], and ranking search results based on them. I find that this gives a pretty good idea whether something is useful to spend time on.

The site is similar to the text-heavy search that got popular here month [4], but for thoughts and ideas rather than factual information.

The main challenge I see is that people rarely search for ideas on a specific topic. Random discovery of new things is appealing, otherwise you want quick answers. This experiment challenges that habit. I would love to hear what you think!

[1] I believe our entire obsession with new things online is unhealthy. An article from 10 years ago really isn't that old, and probably more true or obviously false than ever (Lindy Effect). See also https://news.ycombinator.com/item?id=28616134

[2] I manually added the most-upvoted blogs mentioned on HN, so it's very focused on startups and tech.

[3] Hacker News comments are more thoughtful, but take too long to skim and are often off-topic.

[4] https://news.ycombinator.com/item?id=28550764

Reply on news.ycombinator.com