Save RSS and Atom!
By Marvin Gülker, 2018-07-28.
This website is dedicated to the spreading and promotion of the RSS and Atom news syndication formats. It is primaryly aimed at website owners who publish news items, blog posts, or other content with date information on their websites and want to distribute the information of their new articles to possible readers. It strikes a point for not only relying on social media, but tries to promote the inclusion of RSS or Atom as well. Since the arguments for using these formats are mostly made from a user’s perspective, this document is interesting for ordinary web users as well.
What are RSS and Atom?
RSS and Atom are two well-known and functioning standards for subscription to websites that change over time. The most prominent, albeit not only, example for websites that are a primary target for these two formats are blogs. But really, everything that publishes new content every now and then can have a news feed.
A resource on the web written in RSS or Atom is called a news feed, or often simply “RSS feed” even if it’s the atom format since the RSS format is the historically elder one. A user can point a client software of his choice, called a feed reader or feed aggregator, to this resource and the software will in regular intervals read the newest publications from the news feed. If a news item is new, the software may alert the user about the new item or simply mark the item as new the next time the user opens the feed reader.
RSS (which stands for “Really Simple Syndication”) and Atom are proven standards that have been around since the early days of the web. In particular, RSS in its earliest form dates back as far as 1999 [Netscape99], and Atom was planned since 2003 [Trott03] with its formal standard being published via the IETF in 2005 [RFC4287]. The RSS format specification is maintained since 2003 by a dedicated entity known as the RSS Advisory Board [RSSBoard03]. Over time, RSS saw a number of new versions, with the last and current one being 2.0 from 2009 [RSSBoard09]. The Atom format hasn’t seen any further changes after it was standardised.
Why do RSS and Atom need to be saved?
The World Wide Web is built on certain principles, most notably the principle of decentralisation, which Berners-Lee himself once mentioned as part of the “life and breath of Internet” [Lee13]. RSS and Atom tell this story of the open and decentralised web, in which there are no central entities that control the user’s subscriptions or the content author’s writings. In particular, RSS and Atom are threatened by Facebook and Twitter as more and more web sites either entirely disappear in favour of social network pages, or website owners choose to only notify their readers via those social networks [Crichton18]. If you choose to widthstand the pressure to give up your own website in favour of a pure social network presence, then you already support an open web and it is natural and important to do the next step and liberate your news headline distribution from the social networks as well.
In an open web, nobody is forced to use a particular software, or have an account with a particular service. Each user decides on his own which is the correct way to consume his news. RSS and Atom give the user this power of choice, by choosing between a wealth of available clients that can process these formats (see below). Social network providers on the other hand force the user to consume the content the way they want, which usually means to include the user in their advertisement strategy. Since the social network provider knows all subscriptions of his user, he can compile from them the user’s list of (possible) interests and sell this information to paying customers, who can use this information to show more targetted advertisements.
With a feed reader software, there is nobody else than the user himself who has the knowledge about all of his subscriptions. In other words, using a feed reader gives the user back control about this part of his privacy. Simply resigning and dismissing this as a required evil because today’s web is based on advertising and most users don’t care about privacy [Crichton18] is only making things worse. With that argument, there can never be an escape from the current violations of users’ privacy. Whether most users really do not care is actually disputed [OX16], but in any case, there is a minority that does care. Even if one takes the disputed majority’s attitude as granted, the right to privacy can (and maybe should) be seen as a means of minority protection. As a compromise between user’s privacy and site owner’s need for advertisement money, it is worthwhile to explore whether you can limit your news feed to excerpts rather than including the full article content, thereby forcing users to visit your website.
Moreover, a good feed aggregator software empowers the user to better navigate in the wealth of news. Instead of a single stream flowing down the timeline, the user himself decides which websites to tag with which keywords and narrow down his current browsing experience accordingly. Archives of news articles can easily be searched on the local computer without even the need of an online research.
You as a website owner have a responsibility against your users. You are free to publish your content on social networks, or link to it on social networks, but do not use social networks as the only means of distribution for your news items or blog posts. Give your readers the choice to retain their privacy and use their own tools. It surely is debatable whether the standards should be improved to allow e.g. for a better branding or to cope with sites publishing dozens of articles in a single day [Crichton18], but not using it at all is not going to cause the incentive necessary to improve the standards.
Isn’t e-mail an alternative?
Some website owners choose to send newsletters to their users via e-mail instead of providing a news feed. Such practice does no good. The list server needs to possess a list of e-mail subscribers and therefore exposes the reader’s identity to the operator, whereas RSS and Atom are pseudonymic formats that only reveal the readers’ current IP address (which may easily be obfuscated for example by means of using Tor, see https://www.torproject.org/).
Even worse, at least if your site targets EU users, your e-mail list server needs to comply with EU data protection law (Regulation EU/2016/679, GDPR) as the e-mail addresses are pieces of personal data. You can circumvent this by offering your news as an Atom or RSS feed: in that case, you will only see the reader’s current IP address on your webserver when they pull the latest news. While the IP address is considered to be personal data as well [CJEU16], this is covered already by your normal website’s compliance actions and privacy statement. There is no additional work for you.
Finally, e-mail is not an easily machine-readable format. A user who wants to process his entire collection of news articles faces difficulties if some news come in as e-mail newsletters. For instance, filtering his list of news for any websites the he has tagged “astronomy” becomes difficult. The requirement of having to deal with multiple media of news distribution alone is problematic as there is no easy way to unify RSS/Atom and e-mail news into a single base of news.
Again, you may choose to provide both RSS/Atom and e-mail to your readers, but do not force them to leave their privacy if there are technical possibilities that lift this requirement.
Should I choose RSS or Atom?
Today, there are not too many compelling arguments for or against one of these formats anymore. Both have developed into a mature and reliable state. For the purpose of this document, it is of no meaning which one of the two you use, as long as you do use one of them. The Atom format is a more refined standard that’s a little cleaner to read and easier to extend than RSS, so if in doubt, Atom is probably the better choice. If however you decide to use RSS, you should stick to the most current version, which is 2.0 as of this writing.
A notable exception applies to podcasts. Apple’s “iTunes” software, which is a very popular podcast reader, supports only a format that is based on RSS [Parman17]. In order to not exclude iTunes users, RSS should be used for podcasts rather than Atom.
A longer discussion is available on the Jekyll Bugtracker [Jekyll15]. For podcasters, an overview of best practises exists [Kögl15].
How do I provide an RSS or Atom feed?
Many popular blogging programmes support the creation of news feeds, so you should make use of their facilities. Following is an incomplete list of possibilities.
It is also possible to roll your own implementation. Both RSS and Atom are XML-based formats and there are XML generators available for all popular programming languages. Sometimes, you can even get around with not using a real XML generator at all. In Ruby for example you can just use the “erb” library provided in the standard library (see https://ruby-doc.org/stdlib-2.5.1/libdoc/erb/rdoc/ERB.html).
The technical reference for RSS is [RSSBoard09b]. For Atom, it is [RFC4287]. For both of them you should be able to find example implementations on the web.
Where do I get the feed icon?
There exists a
<link> tag that allows
browsers to autodetect the existance of a news
but Browsers have stopped to detect this
Thus, the existance of an RSS or Atom feed needs to be signalled
explicitely now to your users. Since 2005, there is a
traditional and widespread icon
() that has been
in use with many RSS-related
and is thus understood by most users of news feeds. It is
from feedicons.com in
various colours and sizes, including scalable SVG graphics. If
the standard orange icon does not fit your website’s style,
you can thus easily opt for a differently coloured variant.
The icons available from the website mentioned above can also be used for Atom feeds. From the user’s perspective, either of the RSS and Atom formats is valid input for his feed reader, so there is no need to differenciate the two at an icon level.
Available Feed Readers
A large amount of RSS and Atom feed reader software exists. A comprehensive listing would be outside the scope of this page, but without not at least mentioning the most popular ones it would not be complete. Hence, following is a curated list of still actively maintained feed reader programmes. Keep in mind, however, that the list is incomplete and your ideal feed reader may well be an entirely different one not listed here.
|Livemarks||Browser||Any (Firefox Extension)|
|Newsboat||ncurses||Linux, MacOS, Windows|
|QuiteRSS||Qt||Linux, MacOS, Windows|
|RSSOwl||Java||Linux, MacOS, Windows|
|Tiny Tiny RSS||Browser||Any (web-based)|
|Thunderbird||XUL||Linux, MacOS, Windows|
If you have ideas on how to expand this page, please let me know at email@example.com.
- 2019-01-14: Add link for generating RSS with Joomla!
- 2019-01-14: Added QuiteRSS client. Now each of the popular UI toolkits has an RSS reader mentioned.
- 2018-11-15: Added Vienna RSS client and Microsoft Outlook’s RSS functionality.
- 2018-08-02: Fixed layout of tables on this page
- 2018-08-02: Updated link to Liferea website
- 2018-08-02: Added Livemarks Firefox extension (replacement for Firefox’ dynamic bookmarks)
- 2018-07-29: Added Winds RSS client.
- 2018-07-29: Replaced recommendation for newsbeuter with newsboat, as the latter is actively maintained.
- Brett, Matt: The New Standard Feed Icon, 2005-12-19, https://mattbrett.com/blog/design/2005/the-new-standard-feed-icon/ (accessed on 2018-07-28).
- Id.: Standard Feed Icon (Vector Reproduction), 2006-08-10, http://www.feedicons.com/download/readme.pdf, (accessed on 2018-07-28).
- Court of Justice of the European Union: Judgment in case C-582/14 (Breyer/Germany), ECLI:EU:C:2016:779.
- Crichton, Danny: RSS is undead, 2018-04-07, https://techcrunch.com/2018/04/07/rss-is-undead/ (accessed on 2018-07-28).
- Balter, Ben et al.: What feed format is best?, 2015-05-12, Ticket #2 in the “jekyll-feed” bugtracker, https://github.com/jekyll/jekyll-feed/issues/2 (accessed on 2018-07-28).
- Kögl, Stefan: gPodder Podcast Best Practice, 2015-05-10, https://github.com/gpodder/podcast-feed-best-practice/blob/master/podcast-feed-best-practice.md (accessed 2018-07-28).
- Berners-Lee, Tim: Principles of Design, 2013-03-04, https://www.w3.org/DesignIssues/Principles.html (accessed on 2018-07-28).
- Mills, Chris David et al.:Syndicating content with RSS, 2018-04-23, https://developer.mozilla.org/en-US/docs/Web/RSS/Getting_Started/Syndicating#Adding_the_.3Clink.3E (accessed on 2018-07-28).
- Mozilla Foundation: Remove feed reader and live bookmarks support from Firefox, 2018-07-23, https://bugzilla.mozilla.org/show_bug.cgi?id=1477667 (accessed on 2018-07-28).
- Open-Xchange AG: Consumer Openness Index 2016, https://www.open-xchange.com/fileadmin/user_upload/open-xchange/document/report/open-xchange_coi_report_2016.pdf (accessed on 2018-07-28).
- Parman, Ryan: Spec: iTunes Podcast RSS, 2017-08-27, https://github.com/simplepie/simplepie-ng/wiki/Spec:-iTunes-Podcast-RSS (accessed on 2018-07-28).
- RSS Advisory Board: RSS Advisory Board Members, 2003, http://www.rssboard.org/advisory-board (accessed on 2018-07-28).
- RSS Advisory Board: RSS Specification History, 2009-03-30, http://www.rssboard.org/rss-history (accessed on 2018-07-28).
- RSS Advisory Board: RSS 2.0 Specification, 2009-03-30, http://www.rssboard.org/rss-specification (accessed on 2018-07-28).
- Netscape Communications Corp.:, My Netscape Network Help, copy by the Internet Archive, 1999, https://web.archive.org/web/20001208063100/http://my.netscape.com/publish/help/quickstart.html (accessed on 2018-07-28).
- Nottingham/Sayre (editors): RFC 4287, The Atom Syndication Format, standard published by the Internet Engineering Task Force (IETF), 2005.
- Trott, Benjamin: Why We Need Echo, copy by the Internet Archive, 2003-06-28, https://web.archive.org/web/20080216234454/http://www.sixapart.com/about/news/2003/06/why_we_need_ech.html (accessed on 2018-07-28).
Copyright © 2018 Marvin Gülker
Some Rights Reserved.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. When distributing it, you are asked to include a link to the canonical website of this work, but this is not required.