I've been using this since the beta launch, and I really like it. They're spot on about news being broken.
That said, I do think the service could be improved. Often the summary is a very short blurb that forces me to go to one of the original sites for the content, and hopefully land on one that is not obnoxious to use, which kind of defeats the purpose. The event timeline sounds interesting, but when it essentially shows 2 or 3 events that are obvious from the context, it's not so useful in practice. I always skip the "Quick questions" section, since it reads like an elementary school report, and the questions are really basic. How about letting me ask the questions I want?
Also:
> We don’t scrape content from websites. Instead, we use publicly available RSS feeds that publishers choose to provide.
I think this is a mistake. Most publishers are hostile to RSS and often don't offer it. Scraping is, unfortunately, a requirement if you want to consume public content on your own terms, which is the entire point of this service. Besides, scraping is how all search engines generate their index, so as long as the bot is well behaved and doesn't hammer the site, follows robots.txt or perhaps even bends the rules a bit, it should be fine. I would rather Kagi wasn't so respectful of publishers' wishes, if that would allow them to offer a better service. I understand if they want to avoid getting in trouble with publishers, but the alternative would be better for their users.
That said, I do think the service could be improved. Often the summary is a very short blurb that forces me to go to one of the original sites for the content, and hopefully land on one that is not obnoxious to use, which kind of defeats the purpose. The event timeline sounds interesting, but when it essentially shows 2 or 3 events that are obvious from the context, it's not so useful in practice. I always skip the "Quick questions" section, since it reads like an elementary school report, and the questions are really basic. How about letting me ask the questions I want?
Also:
> We don’t scrape content from websites. Instead, we use publicly available RSS feeds that publishers choose to provide.
I think this is a mistake. Most publishers are hostile to RSS and often don't offer it. Scraping is, unfortunately, a requirement if you want to consume public content on your own terms, which is the entire point of this service. Besides, scraping is how all search engines generate their index, so as long as the bot is well behaved and doesn't hammer the site, follows robots.txt or perhaps even bends the rules a bit, it should be fine. I would rather Kagi wasn't so respectful of publishers' wishes, if that would allow them to offer a better service. I understand if they want to avoid getting in trouble with publishers, but the alternative would be better for their users.