I have long wanted to give DeepaMehta (see my last post) another try. During my vacation, I have now finally managed to import the hundreds of notes from the personal wiki system I chose last year, and I documented the successful (smaller) part of the procedures.
To summarize: I will not use DeepaMehta as my main notes tool, but only occasionally, as “the cherry on the cake”, when a category of my private WordPress has become too populated, to help gaining structure.
The eligibility of WordPress was quite a surprise for me. Although I know it from our corporate environment, I did not think of the features that would solve my main problem with the wiki (reported last year):
- on one hand, trying to conserve as much context as possible,
- on the other hand, desiring granular backlinks.
With the WordPress category view, I have the needed context. And when I link to another item, the pingback function automatically generates a link back. And while the forward link is available from the category context page, the backlink still points precisely to the individual entry (unlike wiki links from one paragraph to another, where the source position is lost in the backlink). Even better, the links as comments can be tracked via RSS and automatically processed further).
Another surprise was, how little I miss the large visual network, and how much I am content with the hierarchical categories, as long as I can rest on visual submaps for the few extreme cases where I need them.
In principle, DeepaMehta’s “on demand” link revelation allows for arbitrary submap composition. But at the present state, it is very cumbersome to reveal more than 7 items, to move a collection of icons, and to even adminster more than 100 items. And it was not possible to import any link not previously exported. But I could get an idea how the revelation would ideally work on the basis of individual items: Still unsufficient, because some aggregation, or zooming, level would be necessary to cope with the plentitude of items. I would need a link connecting to a submap cloud, say, to Paris, which would not specify which suburb within the cloud is actually linked to, say, Orly, until I expand this submap cloud.
Okay, so I ended up without integration of text and visual connections that I dreamt of, with WordPress for the text portion and manually synched powerpoint visualizations. The powerpoint submaps collection is considerably hierarchical. For the bulk of items, this is surprisingly sufficient, as long as I have for special cases…
… my “cherry on the cake” DeepaMehta. If a category is very large or unstructured, I leverage DM’s fine-tuned affordance descibed in the last post. And again surprisingly, also this fully network capable, non-hierarchical process ends in most cases with hierarchical groups rather than arbitrary network links. But, during the process the arbitrary connections are very important. They determine which groups should be constructed, and even if many of the links dissolve in category groups, their optimal handling and manipulation was then essential for this organizing process.