It’s a bit scary when I make a change/edit/?improvement to our Feed2JS (Feed to JavaScript) code, as any glitches or gotchas might show up on the hundreds? thousands? 3? of sites using it (I am basing some large numbers on the 20,000 files that built up in the cache directory).

My apologies for hiccups the last few hours as I tried to redo/undo some recent changes, detailed on the Feed2JS updates. I will ley readers know that I test all updates on a different site/host (actually running on my laptop as a server), but often there are things that pop up which are feed dependent, such as:

* it seems like feeds are not or slowly responding. Magpie (the RSS parser) is reporting timeouts, but it shows up as a no feed. Please verify the URLs are working (I am still researching).
* A new Magpie function added to make the date/time stamp of items easier to access apparently does not work for RSS 2.0 feeds (I reverted to some older, clumsy code to grab the date/time stamps.
* An attempt to improve output by preserving linebreaks actually wrecked other parts of the code, but I think the changes just made should have addressed them. That was my fault for not verifying the impact of modifying a code function and looking at the places where it was used.

Also this morning the server this is hosted in slowed down to a painful crawl (CPU usage topping out), so a restart was necessary (and the kick in the CPU seem to have helped).

I am no pro at this public software code, so bear with me as I learn along the way– if you email me with a problem/issue, I try and look at it ASAP given the large number of people counting on the feeds coming in.

The post "Feed Changes that Send Ripples" was originally slapped on the butt by a cigar smoking doctor yelling "It's a post!" at CogDogBlog ( on January 19, 2005.

Comments are closed.