Show HN: Newscombinator
newscombinator.com
newscombinator.com
The tool here is "just a frontend" I've build as an API-sideproject of mine. So, what does it do? In one sentence: Newscombinator gathers news from different sources, downloads it, makes a screenshot, cleans it, natural language process it, and push it through a MySQL DB into a Solr index.
The Solr-select handler is passed through a REST handler. Why that, why not directly the solr select handler? Because the REST handler also exposes endpoints to the MySQL Database. Some things are just not feasible in Solr alone. For example its hard to build aggregation queries in solr (think: SELECT AVG(points), SUM(points) FROM ... WHERE date_news_found ...), I've made that available in MySQL. And its easy to shift all that to a MariaDB cluster, or to an Amazon RDS, just in case...
What is available here is a RESTful interface to a Solr-Index with all Hacker-News articles, selected Reddits, etc, starting from Oct 2014. That is around 200k of articles. Including searchable content. I think that is a good base to start with for me.
Thank you in the meantime HN, I've learned something.
Also, making the panels on the right of the landing page opaque would aid readability.
Regarding the non-JS users. I am wondering how you can survive in the modern web. That shouldn't be meant as offensive statement, rather than I have the gut feeling that most websites are more like apps currently, running heavily on JS. That really makes me wonder...
To clarify, I have no problem with you using JS. I know I may have to enable it go gain functionality sometimes. My gripe is that on your sites I get zero functionality without it.
I don't mean to cause offense either. Thought it may be something you hadn't considered is all.
I don't want to get into a huge usability discussion right now, but it's important, and you should definitely look into it.
The site itself actually looks like something I would read and share with everyone if the UI/UX issues get worked out.