Hi DJ -- > I'm currently working on separating the GUI from the core so that we > can support multiple HIDs (human interaction devices) simultaneously; > that should allows us to offer a range of interfaces in common code. > I won't make any promises, though, but it's coming along well at the > moment.
One of the problems with PCB's interface is not necessarily the widget set used for the GUI, but rather the actions you must take in order to drive the program. To cite an example of recent interest to me: I would like to edit text I have previously put on the silkscreen layer. Specifically, I want to change its size. How do I do that? Apparently, I need to destroy the current text, go into the preferences menu, change the size of text, and then re-lay the text I wish. What I want to do (and feels natural to anybody used to modern windowing systems) is to select the text, do right click -> edit, have a dialog box pop up, and then edit the text & change its size directly. Another example is the way to edit footprints: You copy the footprint into the past buffer, explode the paste buffer, place the exploded parts, edit them (in an form where pads are lines), then copy them into the paste buffer, create an element, and then paste it back. What I want (and is normal these days) is to select the part, do right click -> edit, and then have a footprint editor pop up window appear. My question is this: Will the new HID layer support easy writing of callbacks which will make the GUI's actions more natural? When I looked at the GTK-PCB source code, I was sufficiently confused about how it worked that I just gave up trying to understand it. I am hoping that the new PCB stuff will make it easy to modify not only the look and feel of the graphical widgets, but also the actions required to drive the program. Expectantly, Stuart
