Feature creep

by Cornel  

My last post provided some details about my current "work in progress". Things have been progressing quite nicely since then, and the new version of PaleoArchives is just about to hit a milestone, as the user interface is nearly feature complete. I completely reorganized the content sections, and in the process of doing so I made the underlying platform so generalized that it can accommodate any site type, Palaeolithic or otherwise. Assemblage types (e.g. lithic and fauna) are no longer "wired in" - they are now dynamic, and site managers can add new types (e.g. ceramic).

While more generalized in one sense, in another sense the new version of PaleoArchives is a lot more specific. Data on assemblages (and in the latest version, also individual samples) are presented and edited in tabular format, making the content a breeze to search. For the tabular data to be useful, every element has to be cross-linked (in the database itself, not just by textual reference) to stratigraphic components and entries in the excavation history. This is not a trivial task, since different researchers often report different (sometimes radically so) stratigraphies for the same site, but I got it figured out. My ultimate objective is to allow for details to go down all the way to individual sample XYZ in a coherent spatial framework. 

To make the new content breakdown more useful, the old "site list" table is now only one alternative to navigating the database; the other alternative is sample or analysis based navigation, which will show similar tables but for each assemblage type individually, in addition to generalized samples (e.g. C14) and features (e.g. hearth or house structure) tables.

In any case, there are two features that I plan on implementing before I make the new version public: a Harris Matrix composer, built right into the platform, and content delegation. Although both of these features will delay release (November probably), I think they are worth it. The first feature, the Harris Matrix composer, will be a bit of a challenge, since I plan on incorporating features not normally seen in Harris Matrices and, in any case, I'm not aware of -any- web based composer of this kind (perhaps I'll make this available as an independent program as well). I think it's imperative to build one into the infrastructure, however, since it seems to be the only sane way of presenting the stratigraphic data in a format that can easily be linked to other types of data, such as individual samples. The second feature, content delegation, is just cool - it will allow PaleoArchives to be updated with content from other databases. I won't reveal the details of how that'll work quite yet, but it's fairly straightforward to implement.


No feedback yet