Note: PlainReader is no longer under development.
PlainReader (working title) is a clean, minimal web-based client for the excellent NewsBlur RSS reader. It's still in an early beta state with only the bare minimum of features to work as a functional news reader. PlainReader is designed to work on WebKit browsers only, such as Safari (including iPad) and Chrome.
Go try it out: the hosted version is now online!
- Simple list of unread items from NewsBlur's River of News
- One-click to send to Instapaper
- Pinboard integration ('popup with tags' & 'read later')
- Full article mode using Instapaper's text engine (just click the article title)
- 20px Georgia article font
- Keyboard shortcuts: up/down arrows to flip through articles, enter for full article mode (see above), spacebar to smoothly scroll articles
- Preview links and footnotes in a Lion-style popover
I started building PlainReader out of a desire for a simple, minimal RSS reading experience akin to that of Silvio Rizzi's Reeder,[^1] but in a web browser (and using NewsBlur instead of Google Reader). The goal is a single webapp that works equally well on my 11" MacBook Air and iPad 1. If you have a bigger or smaller screen, well, your mileage may vary. It's also not intended to be a replacement for the default NewsBlur interface -- PlainReader will never reach feature parity with NewsBlur's web UI, that's not the point.
- NewsBlur: web-based RSS reader (think Google Reader, but better), does all the actual work fetching feeds and such.
- Instapaper: text engine provides cleaned up article content to minimize loading times (and advertising).
- HTML5 Boilerplate: HTML/CSS/JS template.
- HTML5 Boilerplate Ant Build Script: minify, concatenate, and gzip.
- Iconic: free, minimal icons distributed (among other things) in OTF font format, allowing PlainReader's interface to get by without a single image.
- nginx: HTTP server that proxies NewsBlur.
The easiest way to use PlainReader is via the hosted version. However, if you want to run your own instance or make modifications, here are some instructions:
The hosted version is minified, gzipped, etc. by the HTML5 Boilerplate ant build script. To build the yourself site, you'll need Apache Ant. Then, just run
ant buildfrom within the
buildfolder. This is optional.
PlainReader depends upon the web server nginx to proxy NewsBlur and Instapaper. Unfortunately, not just any nginx will do -- you also need the Lua module. If you're on a Mac, use these instructions. Other platforms should be similar.
Once you have nginx installed, modify the
nginx.confat the root of this repo to match the path to PlainReader on your system (see lines 72 and 105). Then run nginx using that config file (
nginx -c nginx.conf). This will host a local instance of the PlainReader source at http://localhost:8000 and the compiled version at http://localhost:8001.
- interface additions/refinements:
- feed favicons & intelligence classifiers in story list
- reimplement 'mark all as read' button
- intelligence tag buttons in the article header
- offline storage
- mark as read/unread button
I don't plan on ever supporting:
- changing the Intelligence filter mode (only yellow and green items are displayed)
- non-WebKit browsers
- mobile phones
- reading individual feeds (instead of the River of News)
- social anything (unless you count pinboard)
- realtime updates (PubSubHubbub)
That said, I'll consider patches / pull requests on any of these if you figure out how to implement them without adding a bunch of complexity or cluttering up the interface.
[^1]: the UI is presently a brazen copy of Reeder for iPad, albeit with lots fewer features and no textures or gradients. File this under 'imitation is the sincerest form of flattery.'