I did a semi-hack to come up with something that's mostly good enough. It consists of:
- A two-frame frameset. The top frame has a header form, the bottom is computed to display a content form.
- Multiple content forms for the resolutions I know users will be using. They are named, for example, Content_1024x768 with an alias that is just the resolution.
- Some Win32 API code to get the current screen resolutions.
- Code in the header form's QueryOpen to get the current resolution.
- An interesting technique I learned from Kevin Pettit via SuperNTF for caching the resolution.
The real magic happens in LS.BE.Win32API script library:
The Content form is computed as follows:
When the frameset opens the Header form's QueryOpen grabs the screen resolution, then the content frame pulls that out and uses it to calculate the form to load. What this gives me is a way to create resolution-specific content and dynamically put it into a frame. Unfortunately this is a Windows-only approach, but I'm sure something similar could be done using Java.
provided by Julian Robichaux at nsftools.com.