definition1Add my vote for this tag syndication1Add my vote for this tag create new tag
, view all tags

Rich Site Summary

RSS defines an XML document format that allows a client to grab details of new items from a website - now supported by TWiki and some other Wikis through WikiRssExtension, and implemented by many news sites (e.g. CNN), portals (e.g. My Netscape) and WebLogs (e.g. SlashDot). RSS can be generated from any type of data, including NewsGroups (now implemented...) and even business information such as supply chain updates.

RSS can be read by desktop-based tools, such as FeedReader, enabling a sort of InstantNotification of changes to TWiki sites, or can be used as content to go into a portal, e.g. SlashDot's use of 'slashboxes' (see WebLogs).


How RSS developed, and why RSS is not RDF

There's an excellent article on RSS at http://www.webreference.com/authoring/languages/xml/rss/1/ - page 2 explains how XML, DTDs, RDF and XML Schemas are used in defining RSS.

Some history that clarifies things a bit, based on this article:

  • The first RSS version, RSS 0.9, came from Netscape, was originally called 'RDF Site Summary' and was defined using XML and the standard RDF (ResourceDescriptionFramework).
  • A followup version, RSS 0.91, was renamed RichSiteSummary and switched to using XML DTDs (Document Type Descriptions) instead of RDF.
  • Then, UserLand released RSS 0.92, based on 0.91 but without bothering with a formal DTD... RSS 0.93 is also being defined by UserLand.
  • Finally, the RSS-Dev working group released RSS 1.0, basing this on the original RSS 0.9 - hence it uses RDF, not DTDs, and is not compatible with RSS 0.91/0.92/0.93. RSS 1.0 also uses XML Namespaces, which make it possible to define extension modules such as WikiRssExtension.

There are now two strands of RSS development, which are not compatible - RSS 0.9 and 1.0 (RDF-based) and RSS 0.91, 0.92 and 0.93 (DTD-based). Most feeds are apparently RSS 0.9 or 0.91, but the use of RSS 1.0 is growing (e.g. that's what TWiki's feed uses).

For a good overview of all this XML stuff as it is used by RSS, see http://www.webreference.com/authoring/languages/xml/rss/1/2.html

I've posted some links to RSS docs under WikiRssExtension. The webreference article also has a good set of links on the last page.

-- RichardDonkin - 11 Feb 2002


I've done some minor refactoring on various pages mentioned RDF and RSS, to reflect the above understanding of RSS and RDF. Have now deleted the old NetscapeRDF page, which was rather confusing and is now redundant. -- RD 14 June 2002

-- RichardDonkin - 19 Feb 2002

I believe that if each Twiki page has an RSS it would increase the immediacy of TWiki and increase user participation. I say this because when searching Google for wiki topics, SvgPlugin for example:

TWiki . Plugins . SvgPlugin
... Plugins.SvgPlugin moved from Codev.SvgPlugin on 19
Sep 2001 - 00:12 by MikeMannix - put it back. 
twiki.org/cgi-bin/view/Plugins/SvgPlugin - 6k - Cached - Similar pages

It shows an old date, I believe an RSS would reflect current status of the page, especially if it is automatically generated. OpenWiki has an RssAggregation site http://openwiki.com/ow.asp?RssPage%2FAggregation

-- DennisDaniels - 21 Mar 2002

Google only indexes sites periodically, based somehow on the frequency of site updates (or maybe even page updates). Having an RSS feed won't change this - and RSS is really for time-ordered data: try pulling a non-time-ordered RSS feed into most RSS readers, they will probably get a bit upset and not show new item alerts - at least FeedReader doesn't.

(Is-this-really-the-only-way-to-get-unbulleted-indents-in-twiki? -- no, you can use blockquote tags, like this -- you can even nest them for deeper indents -- only problem is they typically introduce a blank line that may not always be desired: -- rhk)

Part of the idea might be that Google (and other search engines -- what, you mean there are other search engines?) will have more options in how they time their indexing as things like RSS develop. E.g. a robot could poll a small file that tells them where changes have been made and to what pages, and then apply whatever heuristic to decide what to index now if anything. Then they could poll more sites more often. I guess it's PeerToPeer: just making use of the local computer to calculate and package some meta-information that Google otherwise has to figure out itself. -- JohnAbbe
  • Interesting, but I think that could be done through one of the RSS add-on modules which has information on update periods. Having an RSS feed per page seems like overkill - who is going to subscribe at that level of granularity? It might be useful to have more customisable RSS feeds that can include a subset of pages, e.g. pages you have commented on, providing ConversationTracking. -- RichardDonkin

TWiki does of course support generation of RSS feeds - see WikiRssExtension and TWikiSyndication, which are based on Dec 2001 core code. As for aggregation of RSS feeds, there is some code over on SlashdotPlugin that does this, but I'm not sure what state it's in.

RDF metadata is another matter - it would be good to enhance TWiki's generation of metadata, since we already have quite a lot such as author, time of last change, etc. Google does pick up some basic HTML metadata, such as description and contents - other tools or sites may pick up RDF metadata. There's some confusion about RDF vs. RSS - the latter is just one particular application of RDF (and in some versions doesn't even use RDF).

-- RichardDonkin - 21 Mar 2002

Added link to Ben Hammersley's weblog on RSS.

-- RichardDonkin - 31 May 2002

TWiki has now a RSS reader, Plugins.HeadlinesPlugin

-- PeterThoeny - 12 Jun 2002

Added pointer to RSS Autodiscovery - notes on how to implement for TWiki are at RSSAutoDiscovery.

-- RichardDonkin - 16 Jun 2002

Very interesting article by Jon Udell about the emerging RSS network - individuals' WebLogs generate RSS feeds that are then aggregated and referenced by other weblogs. RSS is moving from a centralised to a distributed (PeerToPeer) model, and according to this article and others I have read recently is 'near the tipping point' at which it will really take off. Doesn't mention RSSAutoDiscovery, but this will probably be important to ensure that any Net-accessible resource (news site, WebLog, Wiki, NewsGroups, mailing lists) can be easily accessed in RSS format.

Wireless blogging (already a reality in press conferences, with news events blogged as soon as they happen), and even WirelessTWiki, will help as well.

-- RichardDonkin - 18 Jun 2002

Added a link to Mark Nottingham's RSS tutorial.

-- ChrisCroome - 6 Mar 2003

I'd like to be able to get the recent changes from within the Evolution email client as it supports news feeds on its Summary page. I get an "Error Downloading RDF" ... Is there an url that will work? I've tried a few around here like the WebRss page with no luck. -- EvanLanglois

  • Evan - I just tried with Evolution_1_4_5 and it worked fine - RSS -- SvenDowideit - 26 Jan 2004

There are various links on WikiRssExtension, give those a try. Note that TWiki generates RSS 1.0, which may or may not be supported by Evolution. If not, it would be possible to edit the WebRss and WebRssBase pages to do another version, as RSS 1.0 is the most complex one - no programming required, only knowledge of the RSS format and TWiki's FormattedSearch.

-- RichardDonkin - 21 May 2003

There's also RSS 2.0 ! : RSS 2.0 Specification

Developped by UserLand, based on 0.9x and incompatible with 1.0 (2.0 is not RDF based).

-- NicolasDuboc - 31 Oct 2003

my RSS clients are not showing the dc:contributor information while it is embedded in the rdf:value tag. I'd like to move it out. Does anyone have one that it works for? (I'm using Liferea on Linux)

-- SvenDowideit - 26 Jan 2004

For validation of our RSS feeds there are a number of resources

  1. http://aggregator.userland.com/validator?url=http%3A%2F%2Ftwiki.org%2Fcgi-bin%2Fview%2FTWiki%2FWebRss%3Fskin%3Drss
  2. http://www.w3.org/2000/06/webdata/xslt?xmlfile=http%3A%2F%2Ftwiki.org%2Fcgi-bin%2Fview%2FTWiki%2FWebRss%3Fskin%3Drss&xslfile=http%3A%2F%2Fhome.freeuk.com%2Fleigh.dodds%2Frss_validator%2F1.0%2Fw3c_validator.xsl
  3. http://www.w3.org/2000/06/webdata/xslt?xmlfile=http%3A%2F%2Ftwiki.org%2Fcgi-bin%2Fview%2FTWiki%2FWebRss%3Fskin%3Drss&xslfile=http%3A%2F%2Fhome.freeuk.com%2Fleigh.dodds%2Frss_validator%2F1.0%2Fexp_w3c_validator.xsl
  4. http://feedvalidator.org/check?url=http%3A%2F%2Ftwiki.org%2Fcgi-bin%2Fview%2FTWiki%2FWebRss%3Fskin%3Drss <<<<----- there's a rogue paragraph marker in the template
  5. http://www.redland.opensource.ac.uk/rss?uri=http%3A%2F%2Ftwiki.org%2Fcgi-bin%2Fview%2FTWiki%2FWebRss%3Fskin%3Drss&Go=Go&box=no <<<----- more errors
  6. http://www.ukoln.ac.uk/cgi-bin/dcdot.pl <<<<----- DC editor - not the results you would expect..

There's a new syndication format called Atom that aims to solve some of the problems with RSS. Let's discuss this on AtomSyndication.

-- RichardDonkin - 12 Feb 2004

TWiki has now an ATOM feed, see AtomSyndication.

-- PeterThoeny - 23 Jan 2006

I'm running a public TWiki site which also has a restricted web for members, containing things like an Agenda, News, meeting minutes, etc. It would be nice if the members could get an RSS or Atom feed off this web, so they don't have to log in everytime to check for updates. Of course they could use e-mail notification, but a newsfeed is much preferred.

I don't suppose it's possible to get an RSS/Atom feed off a view-restricted web? My RSS readers complain about it not being an RSS feed, obviously because they get redirected to an "access denied" OOPS:


Is there a way to authenticate by using URL parameters or something (not very secure, I know..)? I've tried Thunderbird and RSSOwl. Are there other readers that can do it?

-- LevienVanZon - 30 Jan 2006

Eureka! Finally managed to get it working by using NatSkinPlugin's bin/natlogon script instead of bin/view, and adding username and password URL parameters. So you get something like: http://my.twiki.url.here/bin/natlogon/MYWEB/WebAtom?username=myuser&password=mypassword

-- LevienVanZon - 30 Jan 2006

Another option is to store the feed as a static XML page, updated by cron using:
wget http://example.com/bin/view/Myweb/WebRss > /www/html/feeds/MywebWebRss.xml

Here is a test of a Codev feed (not updated): http://twiki.org/~pthoeny/tmp/CodevWebRss.xml

-- PeterThoeny - 30 Jan 2006

Edit | Attach | Watch | Print version | History: r31 < r30 < r29 < r28 < r27 | Backlinks | Raw View | Raw edit | More topic actions
Topic revision: r31 - 2006-01-30 - PeterThoeny
  • Learn about TWiki  
  • Download TWiki
This site is powered by the TWiki collaboration platform Powered by Perl Hosted by OICcam.com Ideas, requests, problems regarding TWiki? Send feedback. Ask community in the support forum.
Copyright © 1999-2018 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.