It’s just data

Intertwingled paths

Shelley Powers: I appreciate Sam's effort, though I think it's important to note that the RSS 0.9x and the RSS 1.0 efforts are following two separate and not necessarily parallel paths.  Why?  It seems to me that there is a common core that they both share.  It seems to me that if we can get past the squabbling over what the outermost element name is and whether items have a life independent of channels, it would seem to me that everybody would be free to innovate however they like in the safe and comfy confines of their own namespaces.

yeah, it's a shame that squabbling derailed things.

unfortunately, i have two beefs with rss 1.0 that prevent me from just switching (since it already supports the one feature i really want -- dates on items -- via the dublin core module).

1. the redundant <items><rdf:Seq> junk. for rss 0.9x, you can just iterate over a list of items and spit them out. as far as i can tell, this <items> thing forces you to iterate over your list twice or build up things in memory before you can output them. silly.

2. it requires a <title> on every <item>. so if you have a weblog without titles, you have to do something dumb like generate fake ones.

oh, and according to syndic8, i appear to be the major source of 0.93 feeds out there (http://www.syndic8.com/feedlist.php?ShowRSSVersion=0.93&ShowStatus=all). 67 of 68 are mine.

Posted by jim winstead at

1. "the redundant <items><rdf:Seq> junk" is redundant for some people, vital for others. The design of a public format is by necessity a balancing job of meeting the interests of as many use-cases as possible without support for one getting in the way of others too much. The lack of proper rdf:Seq rules the other formats with RSS in their name out for some of my uses. Supporting it requires a few extra lines.

2. Use an empty title element!

Posted by Jon Hanna at

Add your comment