Wiki Inside Out
Previous post says to edit wiki pages directly if one's particular personal wiki stores each page as a text file on the filesystem.
Now look at this from the inside out: Given a whole bunch of text files, wrap a wiki around them that has some smarts to recognise/create WikiWords from the content of said text files.
One difficulty I've had with full-text indexing of my existing content is that, somehow, I've never made much good use of the indices. Perhaps it just does not fit with how I process information. Wikifying (instead of, or in addition to, full-text indexing) all those text files is an approach worth investigating.
Sounds like a product idea.
In many ways this seems a useful approach. The current development version of my Wiki implementation has the ability to include random in-situ text files as part of the wiki. However, the main problem that presents itself is that of metadata.
In general, external text files store only content, and you can't mess with the format. This implies that information such as WikiName/filename mapping, who/when edited etc. have to be stored elsewhere. My approach stores a "stub" page in the main Wiki page repository which contains the metadata and the location of the content file, but this is obviously not the only possible solution.
I'm looking at this from the perspective of a personal wiki. These are just some high-level considerations driven by this design goal: How to wikify a bunch of existing, unstructured, text files.
- Each text file is a wiki page. The name of the file is the page title.
- Who/when-edited etc. is per filesystem metadata for the given text file.
- I own the text files, so I am free to modify them as I wish.
- Since these aren't already wiki pages, need some smarts (possibly keyword-driven initially, UI-driven subsequently) to turn words/phrases into WikiWords, to link the pages up.
There is no trackback.