PDFTextOnline, our shiny-new AJAX-y PDF text extraction application, is a nifty tool, and we’re getting some decent feedback. However, many people have indicated (not so indirectly) that its user interface sucks. Yeah, OK, our bad.
This is a lesson that was learned about a decade ago, which we didn’t so much as recognize as stumble over. Here’s PDFTextOnline’s user interface currently (click to enlarge):
It doesn’t look too bad, right? Not so shabby for ‘beta’, whatever that means. Of course, using it is a wholly separate matter. The buttons in the toolbar in the upper right corner are entirely opaque as to their meaning — even though the icons use familiar visual metaphors (open folder for the ‘Open File’ action, a disk for the ‘Save As’ action, etc), they don’t seem to work within this environment. The split betwen the drawer on the left and the main text area doesn’t really work quite right, apparently regardless of whether you’re a Windows, Mac, or Linux person.
Users are flummoxed, and don’t see what the path is from pont A to point B.
Those are just a few of the comments we’ve received so far. The point being, of course, that we didn’t design the UI for the web, as we should have — we designed it to mimic a desktop PDF viewer (except PDFTextOnline’s stock in trade is text). Maybe if we redoubled our efforts, we could roll in a new widget set (perhaps those from Backbase, or something similar), tighten all of javascript that worked with the interactive bits to make those parts more snappy, and end up with something that felt more desktop-ish.
Of course, that’s a bad idea. This is not a desktop application, it’s a web application. Duh.
We’ll have something better in a few weeks, promise.
