Web pages are sometimes divided into frames the content of which is drawn from different sources. Each of these frames may have its own publication date which may have to be checked. In an archiving system, for instance, it may happen that one frame contains the archived information with an old publishing date whereas other frames contain commercials generated at the time of retrieval. The examiner should ensure that he uses the right publication date, i.e. that the cited publication date refers to the intended content.
When a document retrieved from the Internet Archive contains links, there is no guarantee that the links point to documents archived on the same date. It may even happen that the link does not point to an archived page at all but to the current version of the web page. This may in particular be the case for linked images, which are often not archived. It may also happen that archived links do not work at all.
Some internet addresses (URLs) are not persistent, i.e. they are designed to work only during a single session. Long URLs with seemingly random numbers and letters are indicative of these. The presence of such a URL does not prevent the disclosure being used as prior art, but it does mean that the URL will not work for other people (e.g. the applicant when he receives the search report). For non-persistent URLs, or if, for other reasons, it is considered prudent, the examiner should indicate how he arrived at that specific URL from the main home page of the respective website (i.e. which links were followed, or which search terms were used).