Re: [syndication] pubDate and dc:date variance

["Bill Kearney" <ml_yahoo@ideaspace.net>]

> Does anyone have any empirical data on what values have been seen in these
> fields from different feeds?  I'm not only looking to make the logic parse
> dates that are in spec, but also to be lenient where dates are not so
> compliant.

Save yourself the hassle.  Do not be lenient.  The slope gets pretty slippery.
The date formats are /really/ pretty simple.  Outside of something making the
unexpected mistake there's really just NO excuse for getting these wrong.

That and when I've contacted feed makers with bad data they've all fixed the
bugs.  If you come across one that's broken, STOP and throw it back as an error.
It's only by doing this, again and again, will the bugs get exposed and fixed.
Most of the time they're trivial fixes.

I could sift through Syndic8's most recent polling but to what end?  We don't
directly track this so it'd be a bit of work.  Right now, anyone with a feed can
verify that it validates.  Run your feed through the validator:

http://feeds.archive.org/validator/

It's pretty strict and will usually catch most errors.

-Bill Kearney