Show HN: Feedbunch, a simple and elegant feed reader
feedbunch.com
feedbunch.com
I've released Feedbunch, a web app to read RSS and Atom feeds similar to the disappeared google reader. Signups are currently open and it will be totally free for at least as long as it is in the beta phase.
It aims to be simple to use, hiding unnecessary complexity from the user as much as possible (feed autodiscovery, detecting when a feed has changed its URL, etc). It is responsive and supports being installed to the home screen in mobile devices, behaving almost like a native app.
Feedbunch is opensource under the MIT license. The backend is a json REST API built with Ruby on Rails, and the frontend is a mix of Bootstrap and AngularJS.
I'm very interested in hearing your opinion and I hope you find it useful.
email: demo@feedbunch.com
password: feedbunch-demo
looks very sleek, the only thing that would be a deal-breaker for me are keyboard shortcuts for navigation (or maybe I just didn't figure them out?)
It's an often requested feature, though, so I'll prioritize it :)
The thing about using any hosted RSS reader is (which basically means you are getting used to), that you cannot tell if the service will exist tomorrow or not. Feedly is a piece of crap today as it was the same yesterday. They just had a buzz going for them since they provided relatively faster service compared to any other around the time of GReader's demise. Compared to Feedly, Inoreader has lots of bells and whistles ... but ultimately, the notice is on the wall anyway. You are not going to be making a lot of money selling this as a product. There is hardly a chance that any RSS feed reader could become a hot property for any big company to grab. So not sure why my individual OPML be of any value to you.
Hope you follow.
I understand you being worried that a service you use may disappear. But realistically that can happen with absolutely any SaaS. Being backed by a big company is no guarantee in this regard, as Google has shown with google reader and other services. My intention is to keep Feedbunch running for as long as possible, but I was also burned by the GReader shutdown so I understand your concern. The best I can offer is this: Feedbunch has an OPML export feature, so any user can take his subscriptions to any other feed reader that supports OPML whenever he wants. That's the best way to avoid vendor lock-in: to work with interoperable standards so that you can take your data elsewhere if necessary. And of course being opensource you can self-host it.
In case you're interested I originally built Feedbunch for myself when GReader closed down, because I was unhappy with the alternatives. I've been using it for my feed reading needs for some time. Offering it as a SaaS is an extension of that, but the project started as a way for me to have a reader I liked. I'm not aiming to catch the eye of a big company here, but to run a good service.
That was a bit all over the place but I hope I addressed your questions.
Note to everybody else: It's open-source ;)
Anyway, the "About Feedbunch" of the landing page says that it is opensource (MIT license). Maybe it should be more prominent.
* Immediate responsiveness, especially displaying article content
* Information-dense, well-designed UI, so my eyes can skim headlines quickly.
* Search and sort
* Deduplication
* Automatic grouping: This feature would reduce the headlines I need to review by 50% or more, I would guess. Group articles regarding the same topic, similar to what Google News does (but with my feeds and a more information-dense UI). For example, automatically group articles about the new Samsung wearable announcement, so I can choose the best one to read or simply delete them all.
* Filtering: Based on headers or full-text, especially to highlight my favorite feeds and authors.
* Analytics: Which feeds, authors, etc. do I read the most? From which do I open the full article the most (I download all summaries, for speed)? It would help me weed out less productive feeds.
* Efficient management of feeds: Auto-discover changed feed URLs, etc.
* Security: No external connections from within articles without user confirmation. That is, no image downloads, beacons, etc. No JavaScript.
I'd happily pay > $100 for all that.
Nice and simple, seems fast enough. I like that it's a responsive design.
I need an option for oldest first. (I can't use a feed reader without this option.)
Needs j/k for next/previous entry.
Top bar should scroll out of the way when reading entries so it doesn't take up so much space.
Default font should be something more readable than Helvetica/Arial. (My suggestion: Source Sans Pro.)
(Disclosures: I am a happy paid user of FeedHQ, but I like trying new feed readers.)
Now that I've tried it on my phone, I'd say the default font size is a too small. (I tried Firefox on Android.) On desktop, I can use browser page zoom, but on mobile I can't.
Non-Helvetica/Arial could be an option rather than default, for people who don't want to load an external font.
What do the rest of you think about the tipography? Too small, hard to read, good the way it is?
Helvetica/Arial is common enough, and sites like the BBC use it for body text, but there are so many free fonts available that are much readable for running text.
I'm getting 14 px text for article bodies. Usually something closer to 17 px is optimal, and I think zooming the page 120% looks best (14 px · 1.2 = 16.8 px).
In any case, your client does not say anything regarding the matter. My apologies in case this is something that can't be controlled.
EDIT: Another, small, thing: probably would be a better idea to make put the bar that, after registration, alerts to "check the email in order to verify the account" on top of the screen rather than on the bottom; or at least chat it's color to something more eyecatching. I don't know if it's only me, but I almost missed it.
What Feedbunch does, however, is limit how many entries it remembers for each feed. The current limit is 500 entries for a single feed; once a feed has more than 500 entries the oldest ones are removed until there are 500 left. The number (500) could be subject to change during the beta phase if users' feedback indicates it is too high or too low.
> The number (500) could be subject to change during the beta phase if users' feedback indicates it is too high or too low.
I don't think if this would be of any use for you but actually, there are websites that in a month produce way more than 500 entries. LifeHacker for example produces something from 730 to 800 entries per month (with very few of them being useful at all... and this is the reason why I decided to unsubscribe). So everything depends on how (and how much) a user checks the reader.
Not affiliated, just a happy user.
- twitter (@feedbunch)
- github issue tracker (github.com/amatriain/feedbunch)
- email (admin@feedbunch.com)
They are indicated in the landing page and in the signup, login and other pages.
I don't understand your question about the HN feed. HN has an RSS feed (https://news.ycombinator.com/rss) which as far as I can tell includes every link posted to the front page, but not necessarily in the same order. Feedbunch does not create this feed, HN does, and I guess only someone from HN could explain what gets added to the feed. Feedbunch just shows whatever is in the feed.
Regarding HN's feed, all I know is, if I reset Feedly and Feedbunch (i.e., mark everything as read), after x minutes Feedly will show more unread articles than Feedbunch. Over the course of eight hours (when I'm asleep, e.g.), Feedbunch will show maybe a dozen new articles and Feedly will show, like, 40. I don't see this issue with my other feeds.