Sit down and really write out your requirements.

Two things will emerge from that exercise: 1. there probably is in fact a reader out there that meets it, you just haven't found it yet or 2. your requirements require solving the AI-complete problem of perfectly determining what you are interested in.

Generally I find most people slot into #2. I am not kidding. I've been listening to this complaint for about 8 years now and every time I've worked with someone, it's one of the two above, and usually #2.

A third possibility is that you are just trying to read too much, period, if even an AI-complete filter would still leave you with too much. The suck may not lie with the software, it may in the requirements.

I think I created a very nice feed reading experience with NewsBlur: http://www.newsblur.com.

It shows the original site, allows you to read as you normally would, but keeps track of the stories you're scrolling past.

It also allows you to filter stories based on what you like and dislike about them: words/phrases in the title, tags and categories, authors, and the publisher themselves. There is a slider that allows you to show/hide stories based on this filter. It's very fast, too.

I am writing an iPhone app so you can use NewsBlur everywhere. It's just a hobby project, and people have so far been impressed. But I would love for NewsBlur to become a useful tool that people choose to use.

I wrote it because I was also dissatisfied with readers, especially Google Reader. I also knew Python (Django!), JavaScript, and wanted to put them together to test my abilities.

I haven't tested it, but I'm intrigued by Shuan Inman's Fever reader (http://feedafever.com/). Basically the idea is that you separate your feeds into two distinct types: essential feeds that you read completely, and supplemental, high-volume feeds (aggregator-type blogs, like Engadget). Then the essentials are cross-referenced with the supplementals to see what you're likely to be interested in.

I have a similar strategy, but instead of trying to manage the "supplementals" in my RSS reader, I try to a) identify them, and b) banish them back to the browser where they belong. Identifying them can be tricky, at least for me, because I tend to think that feeds I've been reading for a long time are essential, when often they're not. Removing a lot of politics blogs from my reader I swear has made me less stressed, and I don't feel any less informed.

I think the higher volume of content means we have to become better at selection — not only in choosing what we read, but also in identifying and not reading what is noisy and lacking in substance, even when tempting (ahem, HuffPo). Edit: and I'm not sure algorithms are the solution; I think it's something our generation will have to learn — to be our own conscious curators.

I use and love Google Reader. I use the List mode (as opposed to Expanded) and my work flow is the same whether I'm at my desktop, iPhone, iPad, or Blackberry.

A handful of feeds I insist on reading everything, which means I hit J-J-J-J to get through them all. But most I merely skim the unread titles for things that look interesting, and when I'm done I hit Mark As Read to clean up the rest.

I like how I can do this at feed-level, category-level, or with all unread items at once.

And if I'm mobile and I want to see something from my desktop, I star it.

Admittedly, the last time I considered using a thick-client reader was way, way back in 2004 when I was still Windows-only, and I wasn't happy with anything I saw. I've used Google Reader since it was released in 2005, and I've had no reason to use anything else.

I wondered this myself until I hit upon the simple solution of deleting feeds with too low a signal to noise ratio (for my personal tastes). Expecting RSS readers to tease out relevance means having to give them an idea of what's relevant to you. This seems to be a self-inflicted problem on the part of people who use feed readers. I've tended to think I "needed" to follow more than I could possibly consume and retain. Pruning my feed list not only helped with overall comprehension of feeds I'm interested in, but eased a somewhat bothersome feeling that I was falling behind on a keeping up to date.

Personally I ditched NetNewsWire after Reeder for the iPad came out. I have Reeder on my both my iPhone/iPad and just use Google Reader for those times I check feeds on my notebook. The user experience (note that I didn't say user interface) it provides is astounding. Navigating feeds, marking read/favorite, and the article typesetting/layouts are all the best I've seen in an RSS client to date. Combine something like Reeder with a carefully curated feed list and I don't really see there being much of a problem.

The most obvious thing is that they don't do enough to help you filter the wheat from the chaff. The sheer volume of news that's available is too much for anybody to consume... it is almost literally like trying to drink from a firehose.

So our readers need to do a better job of helping us figure out which stories are relevant/interesting to us, give us ways to skim and summarize without reading the whole article, filter out "dupes", etc.

A friend recently launched http://feedingo.com/ which basically tries to make the user experience as friendly as possible.

Pricing is the only problem, but after getting feedback he plans on trying to address that promptly (like $5/yr or something, for a non-free version)

I think you just need to think about what you want, and think about what sources you read from. You say you have 68 sources in your feed, whereas I have ~100.

Personally, I like Google Reader, but I don't use that to read my RSS (just manage it). Instead, I use Google/ig. I have several different categories setup and I don't necessarily read everything. One of the categories I have setup is mRead (short for mark as read) and I don't really read those articles (often), but I still subscribe incase I see a headline that screams my name, or if I want to use Google Reader to do a search within my subscriptions for something specific.

If you have too much noise in your RSS versus signal, than you may want to double check what you've subscribed to and cut back, or add a "mRead" category of your own.

I think there's an issue with people's expectations of their Feed Readers. If you add 100+ feeds then expect the reader to make it easy to read every single post, that will never happen.

Feed Readers work best, and are much more realistic when you treat them like Twitter. You can follow 100+ people, but you don't expect to read every single tweet that they send out. If you miss a post, oh well, it's yesterday's news.

Just log into your reader of choice when you have free time, read the latest news, then go about your day guilt free.

plug: http://feedingo.com

Free tip if you want to write a better RSS reader: research Usenet newsreaders like trn or tin. Many of the problems with RSS that everyone is grappling with now were solved (to varying degrees of success) decades ago. Usenet news and RSS aren't 100% analogous of course, but they are similar enough that there's a lot of man-years of thought which can be reused.

I think it's not the RSS readers' fault. It's the whole reading on internet that sucks.

P.S. - http://goodnoows.com did look promising. Found on HN recently.

I'm not sure if the question is aimed at geeks (ie HN crowd) or in general, but my own observations from previously working at a leading news website indicates that people find the lack of visual aesthetic disappointing in News Readers.

Images are often not included in RSS feeds, and if they are are then they are only shown once you are viewing an item.

When we are triaging which stories to read on a news/blog site, imagery plays an important as we scroll and select - and I don't think I've seen any RSS readers which show you a feed item's images at the story selection level.

The lack of site aesthetic/look and feel maybe less important to geeks but end-consumers also miss that when they are viewing news stories in a reader. In other-words, when I read CNN's feed in my reader I don't get any of CNN's site aesthetic.

Finally, don't forget comments and also the serendipity of suggested related content -- both of which are also missed out when we read stories in a news reader.

Check out Yahoo Pipes. This under-appreciated tool can help you be more selective in the news you actually look at.

http://pipes.yahoo.com/pipes/

If you're willing to forgo graphics, I find newsbeuter (http://www.newsbeuter.org) to be a great reader.

It has excellent mutt-esque key bindings, amazing customizability, etc.

I use Feed on Feeds, a PHP script that runs on my home machine so I can access it from anywhere. It pulls in feeds once an hour and I view them in a big list. It's got a bookmarklet that makes it easy to add a new feed when I'm looking at a site I like.

My workflow is to scroll down the list of unread feeds. If something looks interesting, I either read it within the browser window, or I open the article in a new tab and continue scanning. When I hit the end of the list, I "mark all as read". If I don't finish the list I can mark everything up to the one that's visible as read. Then I go through the tabs I've already opened.

I follow about 200 feeds, most of which are infrequent posters, so I'm going through about 100-200 items a day. Out of those I find I open up maybe 5 or 10. Since I'm scanning everything in a row I can go through the list in a few minutes if I check a couple of times a day.

In most cases I've found that high volume feeds have the good stuff duplicated elsewhere, so I can unsubscribe from the high volume one (though for some reason I still subscribe to Slashdot, even though I learned about all the good stuff a few days prior from other feeds)

The Feed On Feeds interface is also so sparse it doesn't get in the way.

I wrote myself an RSS reader that does exactly what I want...which is aggregate links for me.

It's here: http://newsyndicated.com if anybody is interested.

The only real "feature" of it is the ability to make a sort of "meta rss" feed so that you can share links that you have "liked" with your friends.

I think you need to categorize your feeds by importance. There are some sites you will surely want to catch everything from. There are others where you do not. Do not categorize your feeds by topic but by importance.. that way you can read the "Must Read" folder every day and hit mark all as read on the "Less Important" (or similar) folder if you haven't got the time. That's how I manage my Google Reader and it works a treat!

That aside, relating to filtering the news, my site at http://coder.io/ is doing exactly that but with only coder/programming related news. Still in alpha and very early days but I'm hoping it will resolve the problems you raise for one tiny niche.

I would say it's a combination of a number of reasons, some -- if not all -- relate to the technologies encapsulated in RSS.

First, because RSS by design is meant to just be metadata, and because it's really hard to render HTML (especially those that are not well-formed) the readers typically have to jump through so many hoops to get things just working.

Second, I think it's also because the RSS readers are typically developed by non-designers. Although Google Reader keeps it simple, it's utilitarian and mostly not very pleasant to look at. I've tried others (which are also free) but typically it doesn't come close to making it enticing to read.

Maybe someday someone will do a feed reader that renders articles in a sane and readable way.

Does anybody know if there's an RSS reader that allows regular expression tagging across feed content? Eg, show me all RSS posts that contain "iPhone" or that match the regex "(?:iPhone|iPad|iPod|Apple|Android)" etc, as a simple example.

Try Fever: http://feedafever.com/

Is it the RSS readers that suck or the content itself? If you can avoid these sites that just aggregate content or provide second and third hand accounts of the same items it's easy to trim down your reading list.

I know it sounds like an ad - because it really is - but http://earlyedd.com wants to be the solution for the RSS problem (http://blog.earlyedd.com/post/739958248/the-problem).

EarlyEdd is the RSS reader that shapes on your interests. It discovers people reading pretty much the same stuff as you do (the fellows), and it compiles a relevant news edition especially for you. Give it a try ;)

Most RSS readers suck because they a) are too slow and b) assume you want to read every article i every feed.

I've not tried all the suggestions here (Reeder and Fever sound interesting) but having used GReader and Feedly I keep coming back to using Firefox Live Bookmarks with the LiveClick plugin. I can very quickly see and scan what's new. And with XMarks the feeds can be synced between your computers. Horses for courses I guess.

I've tried several Firefox plugins, Outlook, and a few others that I can't remember and finally settled with Google Reader.

Not sure why you think it's horrible. It's always available from wherever I am, mobile or desktop. Adding feeds is as easy or easier than most other readers I tried, and it even creates feeds for sites that don't have them. Also, the sharing and commenting features are nice.

Agreed. I put up a side project at http://beta.crunch3.com, which was an RSS-on-twitter concept. The interface is "slidey", similar to Pulse, but it gets relevant feeds from twitter, and takes screenshots like Google Fast Flip. I didn't continue development on it, so it's a little slow & the server will probably melt.

I recently launch http://www.gmbhnews.com as a mobile rss feed reader, give it a try and let me know if it solves your problem, all you need is a webkit broser for a better rendering.

For the desktop, you could try http://www.mcsquare.me , it should be fine as well.

Some readers cannot detect duplicated posts.

I'm reading this in NetNewsWire for the iPad, a beautiful, slick, polished app that syncs with Google Reader. I have no complaints.

It sounds like you're subscribing to too many feeds. 2500+ items? You need to do an audit, cut that figure in half, then see if you're still unhappy with your RSS app.

After trying a lot of choices TwitterTim.es is my favorite. It creates a personal newspaper from links popular among friends-of-friends of my Twitter account. It produces very very relevant results for me.

I have a quite good experience with feedly. It's a firefox extension which integrates with google reader. The integration seems to have some bugs with new "shared items", other than that it's perfect.

Wow - I love mine (FeedDemon: http://www.feeddemon.com/). So great - does almost everything I've ever wanted

undefined

Different people use RSS in very different ways.

Reeder on the iPad is spectacularly good.

IMHO it's the lack of obvious manual control over the special sorting algorithm that causes readers to suck.

Do we need an open source techmeme?

Because they are (mostly) web apps.

Www.Netvibes.com is my favorite.

get better at being ignorant

RSS readers are fine, Google Reader is better. The problem is in user experience and how they use it. If you do not prioritize and constantly worry about missed or unread items, you will only end up blaming your reader. The trick is in organizing, prioritizing and ignoring the unread/missed items. With these three strategies, you can effectively absorb unbelievable amount of data on daily basis.

I have A, B, and C tags ordered by "importance" to me - i.e. mostly rss feeds where I want to read or skim every entry.

The rest are ordered by category/tag.

Works fine for me!

Also I use Postrank to turn some of the high-volume into low-volume high-importance feeds.

I use bloglines.