Hacker News new | past | comments | ask | show | jobs | submit login
Sane RSS usage (marco.org)
150 points by sjs on Sept 4, 2011 | hide | past | web | favorite | 42 comments

What Jacqui did in RSS’ absence is always helpful: letting other people filter popular news sites for you.

I've been hearing this back from a lot of my subscribers (I run weekly JavaScript, Ruby, and HTML5 newsletters). They've gotten sick of the deluge that accompanies RSS feeds and the lack of discrimination that sometimes occurs on Twitter and they trust me to curate that info for them. Of course, the next step is to expand into some slightly bigger niches ;-)

Of course, there's the "why aren't you doing this with RSS" crowd too, but with 25k subscribers and growing, I'm finding it pays to focus on those who actually want the e-mails!

I've just subscribed. Where can I see an archive?

At /archive on the various domains :-) e.g. http://javascriptweekly.com/archive - they're scrappy but they work.

Why can't you do both?

To be honest, I do have RSS feeds of the newsletters that I don't promote (unless people specifically ask). I haven't put much effort into them for two reasons: 1) I'm trying to get the e-mail side of things 100% right rather than spread myself too thin, and 2) e-mail is a lot easier to monetize nowadays (think mid/high double digit CPMs rather than the $5 or so I make on my Web properties).

I can attest to the effectiveness of the Cooper mailings as they pretty much reflect the important stuff that is or could be very useful to the reader. It also helps to listen to the Javascript show as well as knowing where Mr. Cooper is getting some of his information from, as he in turn takes the time to vet and confirm something of interest.

Like the author, I too had dozens of feeds coming in and about a month ago I paired them down to under thirty or so and now I probably receive a couple dozen items a day versus hundreds previously, many of which you all know were duplicates or Techcrunch-like filler blurbs. I knew it was bad when people suggested that I needed infinite scroll for my reader, but there was always something about possessing information as it happened in real time* versus having the community discover it for you.

Now if I could only stop copying stuff to Everlater and actually do something with it.

I'm a subscriber to your javascript and html5 newsletters and I must admit you're doing a great job. every week i get the best of the news, carefully picked with short descriptions so I don't have to look any further if something doesn't interest me. i wished there was newsletters like yours for more of the topics that interest me!


These sound awesome! For those of you who are like me and hadn't heard of either before, here are the links:



I'm always open to suggestions! :-) There are a couple more on the way but it's reached a bit of a "crunch" time where I decide if there's a business in it or not. Interesting times.

Didn't know there was a HTML5 weekly too. Subscribed.

Already subscribed to Ruby and JS and absolutely love them. Thanks for the great job!

I mentioned this to Marco and it's why I built NewsBlur, http://www.newsblur.com, specifically for folks like me who follow a few dozen (even a hundred or more) blogs written by a single author. You want to engage with that author on their terms, which means their site. Not stripped away in an RSS feed reader with no context.

I built an Original view to solve this very issue. It's solving the crux of the RSS problem for people like this. You want to follow dozens of writers but you also want to go to their site. But since they may be infrequent, RSS is the only decent answer. NewsBlur's Original view couples the benefit of RSS with the writer's artistic intention in site design and flow.

I also added a trainable intelligence filter to hide stories from those constant-stream multiple-author blogs.

I use the intelligence filters to read the Hacker News firehose and only see stories with [`python`, `javascript`, `ask hn`, `poll`] in their titles.

> You want to engage with that author on their terms, which means their site.

Actually, that's one of the prime reason why I'm an ardent RSS reader user: It's a pretty neat way to just get the content of an article. And as we're talking about news and blog posts here, that's what it all should be about, not your great typekit-served webfont, not your twitter and facebook sidebars, not your blogroll.

My current approach basically has three tiers: First, I scroll through everything in Google Reader. Using keyboard shortcuts I skip the uninteresting parts, read the short blogs, and if there's a longer article that's interesting (or a crippled feed's abstract looks that way), I'll press "v" and get the article in a new tab (using a Firefox preference or a Chrome extension, this tab opens in the background).

After I'm "Inbox Zero" with my feed, I'll read those tabs. Quite often I have to use the print view or Readability on it to actually get to the basic content for those. And if those are too long or, while basically interesting, not very enticing or needed at the moment, I'll send them to Instapaper and read it at leisure later.

I think there's a Mac app that has something similar to your Original view, but personally, that would be a big step backwards regarding usability for me. Never mind that quite often it's not really the writer's "artistic intention" you're ignoring, but that of the news site's designers or the WordPress theme author (quite often not chosen for the readability of the body text). There aren't that many Jason Santa Marias out there who create unique styles for every blog post (and where style and substance actually exist in harmony).

Doesn't the fitter remove serendipity from the feed reading process. Machine learning algos can't ever replace human editing. The messiness of feeds if managed correctly will often surprise you with things you wouldn't have otherwise read.

I actually only use machine learning to present options for you to filter or highlight. Everything is explicit, which means you know exactly what you're missing and what's getting highlighted. (I refer to filter and highlight, but NewsBlur uses a traffic light of unread stories -- red is filtered, yellow is neither, and green is highlighted. You just flip between those three unread states with a slider control.)

typo in link. http://www.newsblur.com/ FTFY

Personally, I use RSS in Google Reader as a living bookmark system. I have about 500 subscriptions and don't pay any attention to the unread count. A couple times a week when I'm in the mood to check out stuff in a certain subject area, I'll go on google reader and click through things in that folder. For instance, if I'm looking for inspiration, I'll click through my "art and design" folder that has 50 to 100 feeds in it. Other times some business event will be in the news, and I'll go into my folder of business feeds.

Apparently using RSS in this way is somehow different than how most people use it. People complain about feeling guilty or somehow stressed out by unread counts. People complain about a psychological pressure to keep up with and read everything in a feed or collection of feeds.

However, I'm left wondering whether this is an issue with how the tools are designed, or perhaps with how RSS is framed. Perhaps we need to develop tools that deemphasize unread counts and instead further emphasize browsing or discontinuous digestion of the content streams.

For a time I thought topic based folders was the best approach, but over time I figured out that sorting folders in an 1 - 10 fashion, 1 being all the sites I really care about to 10 sites I would read if I have the time. By triaging my feeds like this I know then I can read the most important pieces in a very short amount of time. I don feel bad at all about emptying out the otter folders.

This seems like a great way to do things. Currently I have topical folders, and it isn't very optimized. I have things that I want to read (a few people's blogs, tech, security news, etc), and things that I might like to read (e.g. book reviews), and things that might be funny or a pleasant distraction (e.g. cat pictures) all mixed in together.

Moving to a priority model where folders are listed in order of general importance to me, sounds like a real improvement -- the ability to nuke lower priority folders without concern.

Thanks for the idea.

My browser pointed at Hacker News and sub-reddits is my "dip in to the info stream" tool. My RSS reader is for low-volume feeds where I'm fairly likely to care about each item.

That's both bases covered without compromise.

I am a fond user of RSS for the "large number of infrequently updated sites" use case, but I am still subscribed to some high-volume news sources. Relying on your friends to read them for you is good, but obviously not sustainable (if everyone did that...). I'm toying with the idea of doing naive Bayes filtering on my feeds to filter out the mass of stuff I'm not interested in.

> I'm toying with the idea of doing naive Bayes filtering on my feeds to filter out the mass of stuff I'm not interested in.

Some readers already have some machine learning stuff built in, like Google Reader's 'sort by magic' feature. (It works OK, not great.)

One thing that I use RSS for (well, I did before I switched computers...nothing really like Liferea for Mac) is reading webcomics. Irregular Webcomic, Darths & Droids, and The Adventures of Dr. McNinja, all in one place. (I would have done Freefall as well, but there's not an RSS feed for it.)

You may find http://www.diffbot.com/ handy for Freefall.

I have not played with it a lot, but its claim to fame seems to fit your need.

I think its just a case of user organization/finding good sources ... I was subscribed to techcrunch/engadget etc but I found that because of the high volume of stuff they pump out that I didn't read them at all ... so I got rid of all of them and started using the HN top 20/50 RSS feed that someone made a few weeks ago.

I also have my feeds organized into folders, so that the more frequently updating but news worthy stuff is in one folder and the others are other places ... depending on the time I have I go through them in a particular order and it works very well.

But yes, the truth is that sites that create RSS feeds need to be a bit more picky about what they publish ... especialy since they practice that same self restraint in publishing things to facebook, if they have a presence there

In my previous post I explain how to filter any RSS feed so you get only the most popular posts from the feed, thus making it "infrequently updated":


I love RSS and subscribe to about 150 sites or so. I do two things that make it a lot easier though. First, I have a "Read" folder/tag for blogs that I _need_ to read. There are probably ~5 blogs in there total. Second, I have a 'Mread' as in "Mark as read" folder for things that I don't need to read at all, but are available for a quick search as they contain useful information. Between these two I hardly feel like I need to read a ton, but I still get great usage out of RSS.

what the hell is with all these people telling me that I'm doing something wrong with RSS?! I think I'll feel free to consume things in any way I find relevant to my interests. If that means I want a firehose of updates, so be it. If that means I want to only read 6 new things a day like Marco, so be it. It will be my choice and my balance, not his or anyone else's.

Everybody thinks everyone else uses the internet^H^H^H^H^H^H^H^H RSS just like they do.

Everybody believes that everyone else should use the internet^H^H^H^H^H^H^H^H RSS in the way they do it.

Granted, most RSS reader applications get RSS horribly wrong, and erroneously graft the email “inbox” pattern as the UI focus, instead of a river to be skimmed (with power to dive in deep, too). Google Reader is the best in a sorry lot, but mainly because Google sports an inherent advantage (even in the foolhardy sense they’ve implemented it thus far) with its ominous search capability. Read and unread item counts matter not a scintilla to me, as I care just for nimbleness in scanning “what’s new” in the subject matters of my choosing in a chronological manner.

I enjoy reading RSS on Google Reader. It has totally supplanted the time I used to allot to “reading the newspaper”. I know I’ve shared this before, but I keep pace with 2,600 subscriptions. No, I certainly do not read every item and probably only click through less than 10-20% of items. Not true for all sites, as the frequently updated sites get clicked at a 2-3% rate whereas treasured, infrequently updated sites have all their items read. But I don’t fret over unread items and even if I miss reading for a day or two, I feel no obligation to “catch up”, and instead, if I want to review items of interest I may have missed, I use the “Search” feature.

Oh, additionally, all of the mobile and/or tablet RSS reader applications are colossal failures, except for maybe Flipboard, which comes at RSS in a different tact, mainly via Twitter. The whole point of RSS is accelerating the pace at which web content is perused. Doing RSS with a subscription set count of less than a hundred is not much of an efficiency improvement. And the mobile and/or tablet offerings simply choke and sputter on a larger dataset (unless there are new, or updated, offerings I am unaware of). Also, Google Reader (as well as all the Google web products) suck massively on the mobile platforms — it’s why my iPad mostly collects dust and the MacBook Air (using Chrome/Chromium in full screen mode) shines — for the superior Google Reader experience.

I find a very great deal of agreement in what you have said here. A good example of one of my firehose feeds is HackerNews. I don't actually know what proportion of the posts I read, but I would guess it's significantly less than 10%. Sure I'm wasting some time scrolling away most of the headlines, but since one of the primary functions of a human brain is to discard irrelevant inputs, I'm able to do this pretty quickly. The reward is that I get to read articles that I would never otherwise have seen if I just followed 6 people I already agree with.

I see very little reason I shouldn't do everything in RSS. I agree with Marco that this presents a danger of being committed to too much, of being afraid to trash things; the solution is to clearly isolate in your feed reader the feeds you do, and do not expect to fully read.

I use RSS primarily on intranets, to help the guys in the field, who spend maybe 10 minutes a day in front of a computer, keep up on what is going on in our industry. I control which feeds are displayed - they just need to open a browser and read the news page.

My Labor Day side project is basically Reddit style voting and commenting + RSS Feed aggregation, in the interest of solving basically this same problem.

I'm hoping to have it on the server by tonight, I will submit it here tomorrow for you guys to beat up and critique :)

RSS become my bookmarks folder in the cloud.

lifehacker gets lots of posts a day, I really need to filter that

Doesn't it also publish only a summary instead of the full article? If I see a site like that in my feed it quickly gets removed.

I'm not saying I'm particularly against driving hits to a website to show ads (although I really am), but to me it defies the purpose of RSS feeds. (Opinion)

I started using a Yahoo Pipe that just gives me a feed with the week's top Lifehacker posts: http://pipes.yahoo.com/pipes/pipe.info?_id=d750264bc19b3a7e5...

The author suggests you just browse the site manually...

Why do that when you can use an RSS reader that will only lists the articles? Instead of memorizing where the articles are on each website, just use RSS when you want to browse manually.

i read hackernews via rss.. hundreds of posts per day - i simply don't go through all of them. whenever there's nothing interesting to read, i open it up and check what titles are interesting and start reading

The way I catch up with HN posts after a few days away is hckr news (http://hckrnews.com/). It's stress free, and I can shorten the list to 10 or 20 top items per day. Then if I have time I'll expand to the top 50% list and scan for anything interesting.

http://hckrnews.com/ is an awesome way to keep up with hacker news without missing things if you take a few days off.

I don't know if you use Twitter but there are some great (unofficial) HN related accounts on there. @newsyc100 has all items that get at least 100 votes, for example. I also like @hnfirehose for the real time experience.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact