Comment by JunkDNA | Hacker News Reader

Comment by JunkDNA | Hacker News Reader

I use RSS readers almost exclusively to consume content and would miss RSS if nothing else replaced it. I'm an information junkie and RSS readers have made it far easier for me to keep up with the torrent of information out there. That said, I can't escape the feeling that the concept of websites having feeds in a standard format is starting to wane. I also think google reader has sucked a lot of oxygen out of the RSS space. Every mobile or desktop reader pretty much has to sync with google reader which influences their design accordingly. While there are lots of attempts to make things look pretty, there's not a whole lot of major innovation in the RSS reader space. That's not to mention the fact that as far as I know, the reader API is unofficial so it could go bye, bye any time.

All that being said, RSS alone is not exactly the pinnacle of information delivery. What I really want is a better way to identify interesting and informative information and filter out all the junk. This is a very, very hard problem to solve in an automated way. Things like Flipboard are trying to tackle this, but I haven't been able to embrace them. I also don't want to rely on my social network, because I'm different from my network. I have my own interests and priorities (that change over time).

What I want is a feed of information that is what google is to search. Google nearly always shows me exactly what I'm looking for in the top hits. I want something that gives me the most important, useful, and interesting information in a prioritized list all the time. The only thing I've seen get close to this is Fever (http://feedafever.com/). That's a good start, but isn't quite there.

Not to overly self-promote, but you might want to give us a try: http://zite.com . I've personally spent years on the personalization technology behind it, so I'd love to hear what you think.

It's a real pity most (all?) services like this require the user to tell the service what their interests are.

This kind of relationship, where the user trades information about him- or herself for service is sadly all too common on today's internet, and amounts to voluntary participation in a vast spyware network, where you are ever more effectively watched and tracked no matter what you do.

Technologically, it should be quite feasible to do all the RSS feed gathering and filtering for this sort of service on the client's machine, using software installed locally on the user's machine rather than on the service's servers.

Another possibility is to do the processing on the service's servers, but to try to build in some sort of privacy and anonymity guarantees in to the service itself. This is inherently more problematic, since it would really require some sort of verification of those guarantees by a neutral, trusted third-party before it could be completely trusted. But it would still be a big step forward over what we have now.

Good luck with your business, but I'm not going to use it until and unless I can feel secure that I'm not giving you any information about myself, my tastes or my preferences by using it.

Prioritisation of information is a hard problem to solve.

And that's why I think Facebook is doing so well - it takes this problem and uses social means to solve it. You probably need the information most of your friends needed. And if you "like" some of what your friend post a lot, you will probably like what he posts next.

That's why RSS is dead and Facebook is thriving. People don't want linear flow of all possible information, especially today with too much information. You need to take the important stuff out - that's why newspapers have frontpages, that's what google news is doing, that's what slashdot/digg/reddit/hacker news is doing. Facebook do it socially.

Is it? There are a quite a few high-quality blogs I read (via RSS feeds, of course), but I usually discover them via aggregators such as HN, Reddit, Metafilter, etc.

I've found Facebook and Twitter to be a terrible way to identify high-quality content sources: they ironically lack the social-filter function. It's the cultural norms of a community like HN, combined with its also socially-mediated ranking mechanism, that enable aggregators to function so well as content filters. Facebook and Twitter are disjointed in that no one is taking to precisely the same group of people, and there's no evaluative mechanism at all: if someone you follow posts it, it shows up on your front page. Linear flow of all information is exactly what you get.

Facebook is excellent as a glorified contact list, but I still can't found a purpose for Twitter in my life. Reddit is the real winner: the subreddits are coherent communities, and they can be viewed and subscribed to independently of each other, so you're joining a community of communities, each with its own content-filtering mechanism. Of course, Reddit is essentially a modern version of USENET.

Thanks for the suggestion, I liked the philosophy behind Fever so much, I've just bought it (despite the crazy requirements). Might get me hooked on RSS again, who knows.

Reply on news.ycombinator.com