Only 1 of the 9 items listed on kernel.org RSS feed are picked up in Drupal6. I have verified this with 3 other drupal installations, Does anyone have some insight as to the cause?

Feed URL: http://www.kernel.org/kdist/rss.xml
Number of items: 9

CommentFileSizeAuthor
#6 rss.txt9.09 KBlhtown
Support from Acquia helps fund testing for Drupal Acquia logo

Comments

lhtown’s picture

I have the same problem with Drupal 6.14. Only the first item of the feed is listed regardless of aggregator settings.
http://www.kernel.org/kdist/rss.xml

I checked the w3 validator and the feed checks as valid with no issues.

No errors show up in Drupal's log.

lhtown’s picture

One thing about that feed.

Here are the first two articles:

next-20091120: linux-next
11/20/2009 07:32 AM

linux-next:
next-20091120
11/20/2009 08:01 AM

Both are on the same date, but the second one was timestamped 29 minutes AFTER the first one.

Perhaps the aggregator is choking on the date and refusing to go any further.

If that theory is true, when a new article is added, it should display along with the one that is currently displaying.

Rewted’s picture

Very odd. I've submitted a bug report for the module (as you saw), what is our course of action?

lhtown’s picture

We need to reproduce the bug by setting up a Drupal or other feed and verify that it is the date causing the problems. If we can reproduce this consistently, it might be relatively easy to find and fix.

lhtown’s picture

I tried backdating an article in a Drupal blog and the aggregator picks it up just fine.

This bug may still revolve around this issue though since there are other factors at work.

Does anyone know how the rss feeds are generated at kernel.org?

I am going to try making a text file identical to the feed in question and change the date to see if Drupal will pick it up.

lhtown’s picture

FileSize
9.09 KB

OK, perhaps the date theory was wrong. kernel.org updated and changed their feed so that while there are still some goofy things going on with the dates (note the last two entries), the date problem with the second post is no longer an issue, but the Drupal aggregator still only picks up the first item from the feed.

I am attaching a current copy of the feed source code.

Rewted’s picture

Any progress on this?

gmicsko’s picture

We have the same problem with Drupal 5.20:

http://hup.hu

(right sidebar, "__define__ kernel" block)

Rewted’s picture

An obvious issue, wonder why this ticket is still "Unassigned"?

Rewted’s picture

Tried an even older version of Drupal (v4) and the issue remains. Assuming it has something to do with the way Drupal parses the feed.

escoles’s picture

Seeing a similar problem on 6.14, except: All feeds in question are Atom feeds; some feeds load only the first item, others load none, with no errors.

For the following Atom feed, only the first line is aggregated:

http://locusmag.blogspot.com/atom.xml

For the following Atom feeds, no items are aggregated:

http://www.locusmag.com/Roundtable/atom.xml
http://www.mobileread.com/feeds/locusblinks_rss20.xml
http://www.locusmag.com/News/atom.xml

Also, this may be a duplicate of: http://drupal.org/node/424774

... which would mean that the issue goes back at least to D6.9.

dddave’s picture

Marked #424774: Aggregator doesn't add all items a dup because this issue here has more life. ;)

Can anyone confirm this bug with 6.16 or at least 6.15?

escoles’s picture

Version: 6.13 » 6.16

Sorry, yes, I mis-reported earlier: Site where I've observed this is at 6.16, not 6.14. Updating accordingly.

escoles’s picture

Here's another feed that doesn't work (atom):

http://groups.google.com/group/r-spec/feed/atom_v1_0_msgs.xml

Compare to a feed that does (RSS 2.0):

http://groups.google.com/group/r-spec/feed/rss_v2_0_msgs.xml

Also, of the feeds posted above, the following now do work (at least at the moment):

http://www.mobileread.com/feeds/locusblinks_rss20.xml
http://www.locusmag.com/News/atom.xml

Observations:

That the feed works sometimes and not others suggests that it's a transient issue -- related possibly to the formatting of specific items.

For what it's worth, have not yet gotten a Google Atom feed (Google Groups, Blogspot) to work.

EDIT: The Google Groups feed above now seems to work. So, something transient about the feed was stopping it from working for the first 24 hours or so.

EDIT [2]: To anticipate a potential question, in every case where there was a "bad" feed (i.e., either loaded only first item or no items at all), I looked at the feed to ensure there were additional items. However, I did not validate them. If I get a chance today I'll go back and validate any feeds that are still failing.

dddave’s picture