Write divs as nodes and rebuild css class pointers...

Most Web developers that peer into the source of a Lectora published HTML file shudder in horror. Moving forward, I'd be disappointed if the team didn't consider rewriting the publish core and methods to accomodate a cleaner and more efficient output. One that puts a little more control in the hands of the skilled developer (an HTML view) while still maintaining the layer of simplicity and all the other stuff that Lectora does well. I understand where the tool started out, but I'd hope that the tables have turned and more users are publishing to HTML output than EXE output.

I'd like to see the HTML output look more like a structured document and less like a pile of script. I think if it's done right a single style sheet, a node API with navigation logic and structure, and efficiently abstracted class files could cut the average page size for most deployments by half or more (as well as reducing the headaches associated with dealing with limitations of a presentation layer that's 100% semi-dynamic). This will also resolve the infamous page flash where visual elements appear to load again if there's a lot of script to execute on a page.

I'd like to see the user have the ability to determine the object name and have positive control over the node class (maybe 2 modes, auto / advanced). Dang... if I could do this I would have a lot more control and there would be a lot less hoop jumping even for the simpler activity screens.