Versioning Archives

In a collection of reference content, one of the key ideas is to avoid having changes occur between one retrieval and another of a given content item. Being a reference pretty much means being a relatively fixed standard of something. If the item is continually changing, it is too difficult to keep aiming at it as a reference.

In normal use, when a given item is uniquely identifiable, that identity often acts as a simple token of the reliability of the content. It’s no different than calling for the “right” person by their proper name.

Let’s continue that analogy… We know that a name acquires a reputation, and when we ask for someone by name we are usually counting on that reputation to be reconfirmed by our upcoming experience with them. But sometimes this is not going to be the way things work out.

Who amongst us has not observed the difference between the high interest in a famous athlete, and the discomfort of evaluating them when it is evident that they are past their prime or “over the hill”? In cases where the managers of a team have a such a player on the roster, it is often wise for the manager to keep the player but change the player’s role to something where they are still most likely to succeed. The type of success changes, but the idea that the player is valuable remains intact.

The player undergoes changes of two kinds, but they both amount to the same thing — a decreasing impact that certain kinds of skill have on the continuing incoming situations.

The player may actually have a degradation of the skill itself; or, the typical circumstances of their use may now be more complex or tangential compared to those for which the skill had been used in the past. The gap between the skill and the current requirement means that the decreasing relevance of the skill is threatening the value of the player.

In our analogy, skill is just a key characteristic or “property” of the player. When we look at content, the properties of the content are also conditions that may degrade in their relevance to ongoing needs.

To keep the content at a desirable degree of value, there are two main options. One is to “upgrade” the content so that it fills the same role better. The other is to find a better role for the current state of the content.

For example: an item of content might today be a good representation of instruction for best practice of something; at some point that same content may be outdated but it can still be considered for its accuracy and clarity as a historical explanation. Managing the content means knowing if and when the role of the content should change.

OR, that item of content might have begun as information understandable only by a technical reader; the same content item might be edited and expanded to make it usable by a non-technical reader as well. Managing the content means assuring that the non-technical readers are made aware of the improved suitability of the content for their use.

The word “Re-vision” is a great one in that it calls for us to “see something again”. The importance of the additional look is in noticing what value the item actually offers compared to what was earlier assumed.

As content collectors and providers, when we make or accept revised content, our objective is to be sure that the next user gets our “second look” first. We need the item’s current relevance to be more obvious than its prior reputation.

In the organization of a content collection, all content has a role that should be made apparent by the organization’s own design. In effect, the design of the collection’s organization is what gives value to the content within it.

More and more, digitization has come to mean that ideas and data found in one format of a content item are now portable to other items or even other formats.

This immediately brings up the question of whether a found item of content is the one that “should be” used.

More and more, digitization has come to mean that ideas and data found in one format of a content item are now portable to other items or even other formats.

This immediately brings up the question of whether a found item of content is the one that “should be” used.

In a designated repository, administrators have the responsibility to keep appropriate versions of content in appropriate places. The thinking behind doing that calls for an ability to track the lifecycle of the item from its birth (or first appearance) onward. This is typical content management practice, where the item that is the container of the content can be secured.

But digitization has introduced significantly more complexity to the challenge: the malleability of the content itself. When the contents can get out of the containers, the containers on their own can’t be the guarantee of delivering what is needed.

Reference Content

We normally think of “appropriate content” as being represented by a certain version of its container. The container is assumed to be under management by someone who is accepting or rejecting candidate data. The rules of acceptance are assumed to be in the best interest of the eventual content user. And for content intended to serve as a reliable reference to ideas, accuracy or truthfulness are assumed to be proven or at least provable on demand.

But digitization has taken the idea of appropriate back much closer to its real roots. Being “proper for a purpose” is something that can be tested, without preconceptions of whether the purpose is desirable, valuable or necessary.

If a user of content decides to apply the content to a purpose other than the intent of the original maker or current provider, the immediate effect is a change of what is accepted as relevant about the content, and the user need only make a good case for the new use being “effective”.

And here is the heightened challenge. In digital form, content items can be de-composed nearly as easily as they were composed or, in some circumstances, even more easily. Data making up the material of the content can be segmented and re-used, differently and certainly unpredictably.

The easy re-purposing of data points out a critically important issue in understanding what “content” is.

When data is composed, there may be multiple contributing data sources. This emphasizes the fact that “content” is just the material that winds up in a container. If the sources of the material are not themselves validated as appropriate contributors, then having the different material all wind up in one item of content can make the item suspect for its purpose.

A Name By Any Other Rose

We are no longer surprised when a big new film is released under the same title as a well-known book. For book users, the question quickly arises as to how “faithful” the film will be to the book. We might say that there is a “movie version” of the book, but often this turns out to be misleading. The container called “book” and the container called “film” turn out to have some data and ideas in common, but there can be very important differences, such as what they do not both include and how they respectively used what they included.

Such differences might render a non-fictional book into a fictional film. Yet the “story” in the film will be deemed a specific content idea itself and go on to have a life of its own. For example, the film story might be reproduced at another time with several kinds of modifications that make no attempt to refer back to the original non-fictional book. Meanwhile, within the film community, later film renderings of the story are said to refer back to the “original” film version…

Those observations bring an important point to the foreground: when it comes to content, the only reason that versioning is important is because it is defined for the benefit of the users.

The 3-D World of Versioning

What turns one version of content into a different version, or even, in effect, different content?

The short answer is the intent of the user. User intent comes up in at least three situations, each of which are made more likely due to digitization removing traditional obstacles.

With recomposition, the data making up an items’ content can easily produce a different new idea and impact than what made the earlier content item important. We need to be able to recognize that an earlier content item may actually “devolve” into being mainly a data source. But that new status may still be a strong enough reason to keep the item in a collection.

Without recomposition, users may still demonstrate that the initial content has a consistent, alternative meaning in a context other than the one originally addressed. Duplicating or sampling existing content in new situations is no longer unusual with or without planning. This may or may not be deemed production of “different” content. Often such expanded range of use is accomplished by including the earlier content within other additional content. The original container may or may not be brought along…

And without even adding or changing contexts, an earlier finished composition might be superceded by a later one that represents a “better” rendering of the original idea for the same purpose. That improvement can definitely include new decisions about what data will be included or excluded. Older data may be replaced or enhanced, which may or may not change the effects that the composition has overall.

Recomposing, recontextualizing and revising content generates new, flexible, or improved content in order to address what users want. Compared to an original item of content, each of those kinds of changes can be identified as a distinct addition to a content collection. And, even with the additions, there still may be an ongoing but new type of need for the original.

At any given time, the current value of each content item, whether the original item or later, should be the decisive factor in how it is saved and made available for reference use. Because digitized content is so difficult to secure, the ordinary challenge is instead to create and promote access points to managed collections having known curatorial attention to the content. A cooperative audience then has reduced likelihood of picking up inappropriate content from that collection.

The curatorial influence will cover all four of the changes discussed above as modifications of the original version. And finally, in the content collection, these new items derived from the original will be separately identified and located in a context appropriate for their likely type of effectiveness.

Revising Content for Quality

The Right Stuff