Are RSS guids actually expected to be _globally_ unique?

Viewed 1082

Just trying to clear up what level of uniqueness the <guid> element in an RSS feed is actually supposed to have. I understand that one of its main purposes is to be something that software can use to identify the item for such purposes as read/unread tracking. But am I right that:

  • once a guid has been used, it should never be used again, even if the last instance was removed from the feed ages ago?
  • it should be unique not only within a feed, but also across multiple feeds and even (to the extent it can be achieved) unrelated websites?
  • one of the reasons it's common to use URLs as guids is to help achieve the above?

Moreover, if a program does encounter the same guid twice in different feeds, what should happen?

  • it treats them as distinct RSS items, since they are in different feeds?
  • they are considered to be one and the same item, just published in multiple places (similar to Usenet crossposts)?
  • it depends on whether they're on the same site/domain?
  • the behaviour is undefined?
1 Answers
Related