Lately the Slate Moneybox feed http://www.slate.com/blogs/moneybox.f… has not been displaying correctly in Newsblur. It’s a full text feed but many Newsblur entries in Feed view only display the photo caption, and sometimes even the Text view doesn’t display the whole entry either. This only started happening a few days ago.
It’s showing the content instead of the content. For example, the item with http://www.slate.com/blogs/moneybox/2… shows “These white guys aren’t op-ed writers even though they’re directly in front of a newspaper building. Just some angry postal workers.”, which is in the of the xml file, while the starts “<p>Trolling the universe this morning,…”
This is unfortunately a result of my upgrading to a newer feed parser, which is more correct than I could ever be in getting the right fields out. I’m going to have to defer to Slate’s feed, which probably has an issue.
Checking it in the feed validator, it seems it does in fact have an issue: http://validator.w3.org/feed/check.cg…
Those errors in the Slate feed shouldn’t cause your parser to think that the element is the description for the whole item. The only error is “media:group must have multiple media:content children”, which should probably make your parser ignore the whole , while the problem is that your parser is using a piece of the group in an inappropriate way.
http://feeds.guardian.co.uk/theguardi…, mentioned in https://getsatisfaction.com/newsblur/…, also validates but has the same problem in NewsBlur.
Thanks, works perfectly now.
Ok, I took a good look at this and came out with a great resolution. While I maintain that the feed is behaving incorrectly (using content and summary, where the summary is longer than the truncated content!), I went ahead and made a sanity check to grab the longer of the two. Feed works great now. Insta-fetch the Guardian feed to correct that one as well.