> I like it for extensibility and mapping to complex sql schema's, but > I was thinking if I could map all columns of a treeview and an extra > column for the path. The path could then be applied to create the
in the first, non public, release of sqlkit I used a 1:1 mapping between liststore (I don't currently use treestores) and it turned out to be slower. Ythe reason is that in that case, you have to distribute the data you get from the class to the model for all values of the model, while I let the column's cell_data_function to do it just for the visible cells. > I am trying to determine if this might be a better > fit for small apps. It will keep the data storage in the same "shape" > that the widget uses, and keep it simple so I do not get confused :) I personally don't find it's simpler, the class that sqlalchemy returns it's a pretty easy object to work with, it's normally in the session and you don't have to retrieve the data from the TreeModel to check if it's changed, you just check session.dirty, maybe session.is_modified(obj). I'd really suggest you check why you consider it more natural, you're using an ORM that is exactly the way to use the *record* in an object oriented way, not in a tabular way... sandro *:-) -- Sandro Dentella *:-) http://sqlkit.argolinux.org SQLkit home page - PyGTK/python/sqlalchemy _______________________________________________ pygtk mailing list [email protected] http://www.daa.com.au/mailman/listinfo/pygtk Read the PyGTK FAQ: http://faq.pygtk.org/
