Two recent posts (borisb on annotation and fgruber on attaching journal notes) have me thinking about integration between the library and journal apps.

Specifically, for most notes on what I'm reading, emulating paper is not what I want to do: not only is margin space typically limited, but I like to make the text larger, effectively eliminating margins. And I don't want what I'm writing to obscure what's beneath it. Instead, I want the behavior that fgruber describes: compose my notes on blank paper, but have them automatically associated with the text.

Here's how I could see it working: in the reader, the highlight tool (or something like it) allows you to highlight a piece of text. Doing so would then automatically split the screen, keeping the highlighted text (perhaps with some additional context) on top, while opening a blank page below for notes. On closing the page of notes, it is automatically saved and attached to the original.

There are some obvious variations on this theme-- display the highlighted region on the lcd instead of splitting the e-ink display. Or instead of having a blank page for each note, keep a single journal open, which is then synchronized with the text. There could even be a display mode where a thumbnail version of a note is displayed in the margin alongside the highlighted text.

But what this gets away from, along the lines of borisb's suggestion, is requiring a graphical overlay (annotations) to interoperate nicely with a textual display, particularly involving reflowed text. Let each medium perform the job it excels at, without requiring them to be displayed on the same page, which is what seems to be the difficult part.

Exporting annotations would be simpler, too, since it would be easier for external tools to deals with one type of information or the other (pen strokes vs. locations within text). This ought to simplify zooming and rotating, too, since there's less coupling required between the graphics and text modes, which I think is what prompted borisb's comment.