The Internet isn’t our modern version of the Library of Alexandria, Brewster Kahle’s Internet Archive is closer in spirit to that ancient centre of knowledge.

link rot
Over time, especially Internet time, web links break. It’s a part of the natural changing order of things. But the Portable Consultant is particularly sorry to see knowledge, freely offered by a company as part of its marketing, removed completely from the web after a merger or acquisition.

The Oracle Corporation is good at this – or bad, depending on your point of view.

It happened when Oracle acquired Stellent, a well-known ECM company.

The Stellent site remained for a time, but links to free resources like white papers broke quickly and those to training programs followed soon after. It was reasonable, perhaps, in the case of training as that function came under the control of Oracle University. It was sad in the case of the other online ECM resources and white papers that disappeared. Some of these still exist but have been buried without the Stellent name, no doubt because the branding was inconsistent. If the Google search that lists any of these has a broken link you might try looking on Oracle.com under Fusion Middleware (say, What?).

re-branding trumps Search Engine Optimization
Such re-branding seems foolish in one important sense: a Google search on “Stellent training” does not offer any canonical links to Oracle training services on the first page of hits, but if you search the Oracle site you may find it buried in Oracle University under the incomprehensible name of Fusion Middleware. Right, as if all those who read about Stellent in (pre-acquisition) articles and reviews will recognize that Stellent is today “Fusion Middleware”.

No doubt this made sense to someone Oracle management.

Hyperion articles lost
It happened again when Oracle acquired Hyperion.

This time it was more personal. The Portable Consultant had written here about an article on the (former) Hyperion site, Unstructured Text and Structured Data. It took its theme from the George Lakoff book Women, Fire, and Dangerous Things (what categories reveal about the mind). The Hyperion article links categorization with data mining and Enterprise Content Management, describing several data mining approaches from various companies active in the space.

Sadly, that article is nowhere to be found on the Net. If it does still exist it is buried so deeply within the Oracle site that none of the unique keywords can retrieve it.

However, with some patience, I was able to retrieve a copy through The WayBack Machine at the Internet Archive.

What a shame that such articles should be lost for no apparent reason or for the sake of branding or a change in marketing approach. Thank goodness someone recognized the fragility of our Internet and is attempting to build our new Library of Alexandria.

The Open Text Index and the Internet Archive
I first heard of Brewster Kahle in connection with the intriguing WAIS (Wide Area Information Servers) distributed data search/retrieval system. WAIS indexed databases on Internet sites (not web sites, for this was before Berners-Lee created the world-wide web). As I recall, a central index led searchers to indexes, often at universities and research establishments – foreshadowing today’s Internet search engines.

Later, when Open Text Corporation’s Tim Bray had established the Open Text Index search engine, it became for a short time the back-end search engine for Yahoo! (circa 1995). As technical administrator of the OTI I regularly created a small pile of backup tapes for Open Text’s CEO, Tom Jenkins, to hold up at presentations as: “all the information on the world wide web”. For a time this was probably a fair approximation. Later, these tapes were passed on to Brewster Kahle to incorporate into the Internet Archive.

Today the WayBack Machine provides one of the few links (pun intended) to our Internet past. Whole sites are harvested at regular intervals to provide access, if sometimes spotty, to past web site versions and the information they held. I was able to find my missing article in a collection of pages from the old Hyperion site of 2006.

That I can retrieve Unstructured Text and Structured Data from the Internet Archive is a testament to the foresight of Internet pioneers like Kahle. That I cannot find it through an Internet search engine or retrieve it from an Oracle archive is an example of (understandable) short term commercial interests that do not practice knowledge management… including those companies that would be happy to tell you all about their KM software offerings! 🙂

Cheers,
-pmh