Forum Posts Feed
ConnectedText versus Ndxcards
< Next Topic | Back to topic list | Previous Topic >
Posted by Stephen R. Diamond
Oct 23, 2007 at 04:53 AM
Manfred wrote:
>>
>Well, not necessarily. You can connect everything to everything, but that
>would make the connections meaningless. It would be the equivalent of having every
>sub-topic in a branch of the outline, appear in every other sub-topic.
>
>In fact, you
>impart meaning on the information by selectively connecting things. Novices always
>ask for “automatic linking” or the idea that every topic refers to every topic in the
>whole database.
Yes, I understand that much. What I’m getting at is that it’s _hard_ to be selective with undirected connections. But then, the connections aren’t really undirected, you say:
>
>
>I don’t understand what “undirected link” means either. In the
>example there is a direction from Flowers to roses (or, in this case, from general to
>particular). You can also link back from rose to Flowers (or you could assign the
>category “flower” to the topic of roses, primroses, etc., and have these topics
>automatically listed in the topic flower).
So you’re saying that a distinction exists between connecting flowers to rose versus connecting roses to flowers? A connection from roses to flowers looks different from a connection from flowers to roses? Then I must have misconceived the essence of wikis. You hear about flowers and roses being “linked,” which seems to imply the indiscriminability of direction. Wikis are often lauded because they do not require that you choose which node is superordinate and subordinate. But if connections are directed, such a choice _is_ required (although, as in an outline with cloning, a circular connection is at least formally possible). The item linked from, then, is effectively subordinate to the item linked to. One can assign a semantics other than that of subordination, although I’m not sure what else the semantics (of one-way connections) could consistently comprise.
I interpret a link from rose to flowers as a hierarchy. Rose is designated as subordinate to flowers. (Or the other way around. It really doesn’t matter, the choice being conventional.) A link going the other way, from rose to flowers, makes flowers subordinate to rose, as when—per your example—instances of ‘rose” includes the flower rose and the woman named ‘Rose.’
Am I correctly understanding now that a “link” is actually an arrow, not a line?
Posted by Stephen R. Diamond
Oct 23, 2007 at 05:18 PM
Stephen Zeoli wrote:
>Having dabbled with various wiki-style PIMS (the concept of which I like, but have yet
>to get over the hurdle of adapting to all the conventions required to use them
>effectively), it seems to me that there are some functional differences between
>wiki-style linking and cloning in an outline.
As is no doubt obvious, I have not dabbled. My habit—for better or worse—is to obsess good and long before actually trying something new.
In the wiki the connections are
>instantly visible as you are reviewing the content of the note, whereas you have to
>look for and search an outline to understand the context. I would also suggest, though
>this is just my own theory, that the connections in a wiki are more organic—that is,
>they arise more naturally as the content develops—than in an outline (whether or not
>this is a benefit probably depends upon your purpose).
One question would concern the extent to which the differences you observe are inherent to Wikis or arise because of arbitrary or adventitious correlations between the basic approach and the features you value. Bracketing wikis for the moment, because of my fundamental ignorance, consider the difference between hierarchical key words and outlines. As often conceived, it boils down to a more formalism: do you drag children to parents, as in outlines, or do you parents to children, as in key words. No doubt some people prefer one or the other, but the distinction doesn’t seem to deserve to be called fundamental.
There is, I think, a distinction worth making between the real cores of the two approaches. In an outline you can subordinate to items; with a categorical approach (which seems more apt than ‘keyword,’ considering what the distinction actually amounts to) items you can subordinate to are disinguished from he basic ‘note’ unit. This lends itself to the category tree but doesn’t require it. Programs outputting an apparent quasi-outline can be categorical in nature. Black Hold Organizer comes to mind.
In theory, you could drag from parent to child and have an outline strucure. A program could accomplish this if the outline turned any categorized item into a category. This would suit someone who like to drag to subordinates instead of to parent but wants the greater flexibility of an quasi-outliner as opposed to system of relatively fixed categories.
All I’m getting at is that the clusters of features that existing programs present do not necessarily reflect what’s possible or what’s responsible for valued features.
>Finally, a wiki allows for
>connections between notes without imposing a judgment about which is the superior
>thought or idea. That isn’t always benefit, as it is important to be able to create that
>kind of organization, especially when writing—which, I assume, is why CT has now
>added the outlining feature.
>
>Steve Z.